S [X S [ G MA

A Complete Step-by-Step Guide

A Complete Training & Reference Guide for

White Retts, Yettow Retts, Green Retts, and Rtack Retts
 StXStGMA:
A COMPLETE STEP-BY-STEP GU!DE

i
© 2018 The Council for Six Sigma Certification. All rights reserved.

Harmony Living, LLC, 412 N. Main St, Suite 100, Buffalo, WY 82834

July 2018 Edition

Disclaimer: The information provided within this book is for general informational purposes
only. While we try to keep the information up-to-date and correct, there are no
representations or warranties, express or implied, about the completeness, accuracy,
reliability, suitability or availability with respect to the information, products, services, or
related graphics contained in this eBook for any purpose. Any use of this information is at
your own risk.

The author does not assume and hereby disclaims any liability to any party for any loss,
damage, or disruption caused by errors or omissions, whether such errors or omissions
result from accident, negligence, or any other cause.

ii
 SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

Using the most recent edition in your workplace?

Because we continually attempt to keep our Handbook up to date with the latest industry
developments, be sure to check our website often for the most recent edition at
www.sixsigmacouncil.org.

iii
 TABLE OF CONTENTS
TABLE OF CONTENTS )V

UNIT 1: INTRODUCTION TO SIX SIGMA 8

CHAPTER 1: WHAT IS SIX SIGMA?

CHAPTER 2: SIX SIGMA HISTORY AND APPLICATION 21
CHAPTER 3: OTHER PROCESS IMPROVEMENT AND QUALITY METHODS 33
CHAPTER 4: LEAN CONCEPTS 43
CHAPTER 5: BASIC SIX SIGMA CONCEPTS 60
CHAPTER 6: APPROACHING THE PROBLEM 84

UNIT 2: PROJECTS AND PROCESSES 94

CHAPTER 7: WHAT IS A PROCESS? 94

CHAPTER 8: QUALITY 113
CHAPTER 9: SELECTINGTHE RIGHT PROJECTS 126
CHAPTER 10: BASIC SIX SIGMA TEAM MANAGEMENT 142
CHAPTER 11: INTRODUCTION TO DMAIC AND DMADV 156

UNIT3: ADVANCED DMAIC 169

CHAPTER12: DEFINE 169

CHAPTER 13: MEASURE 182

CHAPTER 14: ANALYZE 206

CHAPTER 15: IMPROVE 224

CHAPTER 16: CONTROL 234

UNIT 4: BEGINNER STATISTICS 243

CHAPTER 17: INTERMEDIATE GRAPHICALANALYSIS 243

CHAPTER 18: NORMAL PROBABILITY DISTRIBUTIONS 274
CHAPTER 19: CORRELATION AND REGRESSION 310

UNIT 5: INTERMEDIATE STATISTICS 335

iv
SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

v
 CHAPTER41: SIXSIGMAINMARKETINGANDSALES 735
CHAPTER 42: SIX SIGMA IN CUSTOMER SERVICE 742
CHAPTER 43: SIX SIGMA IN CALLCENTERS 749
CHAPTER 44: SIX SIGMA IN THE RETAIL INDUSTRY 755
CHAPTER 45: SIX SIGMA IN ECOMMERCE 762

UNIT 11: SIX SIGMA IN NON CUSTOMER-FACING FIELDS 770

CHAPTER 46: SIX SIGMA IN WAREHOUSE 770

CHAPTER 47: SIX SIGMA IN DISTRIBUTION 778
CHAPTER 48: SIX SIGMA IN MANUFACTURING 786
CHAPTER 49: SIX SIGMA IN FOODSERVICE 794
CHAPTER 50: SIX SIGMA IN CONSTRUCTION 801
CHAPTER 51: SIX SIGMA IN MILITARY/DEFENSE 809
CHAPTER 52: SIX SIGMA IN GOVERNMENT 816

vi
SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

vii
Six Sigma, or 6 , is both a methodology for process improvement and a statistical concept
that seeks to define the variation inherent in any process. The overarching premise of Six
Sigma is that variation in a process leads to opportunities for error; opportunities for error
then lead to risks for product defects. Product defects whether in a tangible process or a
service lead to poor customer satisfaction. By working to reduce variation and
opportunities for error, the Six Sigma method ultimately reduces process costs and
increases customer satisfaction.

In applying Six Sigma, organizations, teams, and project managers seek to implement
strategies that are based on measurement and metrics. Historically, many business leaders
made decisions based on intuition or experience. Despite some common beliefs in various

negate the importance of intuition in any process. Instead, Six Sigma works alongside other
skills, experience, and knowledge to provide a mathematical and statistical foundation for
statistics prove that to be
true. Intuition might guide a project manager to believe a certain change could improve
output; Six Sigma tools help organizations validate those assumptions.

8
 SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

Without proper measurement and

analysis, decision making processes in What is beta testing?
an organization might proceed as
Beta testing is the act of implementing
follows:
a new idea, system, or product with a
Someone with clout in the select group of people or processes in
organization has a good idea as controlled an environment as
or takes interest in someone possible. After beta testers identify
potential problems and those
Based on past experience or problems are corrected, the idea,
knowledge, decision makers
system, or product can be rolled out to
within an organization
the entire population of customers,
believe the idea will be
successful.
employees, or processes. The purpose
The idea is implemented; of beta testing is to reduce the risks
sometimes it is implemented and costs inherent in launching an
in beta mode so expenses unproven product or system to a
and risks are minimized. widespread audience.
The success of the idea is
weighed after implementation; problems are addressed after they impact products
or processes in some way in the present or the future.

Beta testing is sometimes used in a Six Sigma approach, but the idea or change in question
goes through rigorous analysis and data testing first. The disadvantage of launching ideas
into beta or to an entire population--without going through a Six Sigma methodology is
that organizations can experience unintended consequences from changes, spend money

through trial-and-error periods rife with opportunities for error. In many cases,
organi
true gain or loss associated with the change. Some improvements may appear to work on
the surface without actually impacting customer satisfaction or profit in a positive way.

The Six Sigma method lets organizations identify problems, validate assumptions,
brainstorm solutions, and plan for implementation to avoid unintended consequences. By
applying tools such as statistical analysis and process mapping to problems and solutions,
9
 WHAT IS SIX SIGMA?

teams can visualize and predict outcomes with a high-level of accuracy, letting leadership
make decisions with less financial risk.

use
of the tools described in this book, problems can arise for teams as they implement and

implement changes, they can control processes for a fraction of the cost of traditional
quality methods by continuing the use of Six Sigma tools and statistics.

Six Sigma as a methodology for process improvement involves a vast library of tools and
knowledge, which will be covered throughou
the statistical concept represented by 6

At the most basic definition, 6 is a statistical representation for what many experts call a
nly 3.4 defects per million
opportunities. In percentages, that means 99.99966 percent of the products from a Six
Sigma process are without defect. At just one sigma level below 5 , or 99.97 percent
accuracy--processes experience 233 errors per million opportunities. In simpler terms, there
are going to be many more unsatisfied customers.

According to the National Oceanic and Atmospheric Administration, air traffic controllers in
the United States handle 28,537 commercial flights daily.1 In a year, that is approximately
10.416 million flights. Based on a Five Sigma air traffic control process, errors of some type
occur in the process for handling approximately 2,426 flights every year. With a Six Sigma
process, that risk drops to 35.41 errors.

The CDC reports that approximately 51.4 million surgeries are performed in the United
States in a given year.2 Based on a 99.97 accuracy rate, doctors would make errors in 11,976
surgeries each year, or 230 surgeries a week. At Six Sigma, that drops to approximately 174
errors a year for the entire country, or just over 3 errors each week. At Five Sigma, patients
are 68 times more likely to experience an error at the hands of medical providers.

http://sos.noaa.gov/Datasets/dataset.php?id=44
2
rol and Prevention.
http://www.cdc.gov/nchs/fastats/inpatient-surgery.htm
10
 SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

While most people accept a 99.9 percent accuracy rate in even the most critical services on
a daily basis, the above examples highlight how wide the gap between Six Sigma and Five

associated with each error.

Consider an example based on Amazon shipments. On Cyber Monday in 2013, Amazon

processed a whopping 36.8 million orders.3
company an average of $35 (a very conservative number, considering that costs might
include return shipping, labor to answer customer phone calls or emails, and labor and
shipping to right a wrong order).

Cost of Amazon Order Errors, 5

Total Orders Errors Average Cost per Total Cost of
Error Errors

36.8 million 8574.4 $35 $300,104.00

Cost of Amazon Order Errors, 6

Total Orders Errors Average Cost per Total Cost of
Error Errors

36.8 million 125.12 $35 $4,379.20

For this example, the cost difference in sigma levels is still over $295,000 for the Cyber
Monday business.

For most organizations, Six Sigma processes are a constant target. Achieving and

improvement. But even advancing from lower levels of sigma to a Four or Five Sigma
process has
Cyber Monday example at other levels of sigma.

Boy Genius Reports, Dec. 26, 2013. http://bgr.com/2013/12/26/amazon-holiday-season-

sales-2013/
11
 WHAT IS SIX SIGMA?

Sigma Level Defects per Million Estimated Cyber Total Cost (at $35
Opportunities Monday Defects estimate per
error)

One Sigma 690,000 25,392,000 $888,720,000

Two Sigma 308,000 11,334,400 $396,704,000

Three Sigma 66,800 2,458,240 $86,038,400

Four Sigma 6,200 228,160 $7,985,600

Five Sigma 233 8,574.4 $300,104

Six Sigma 3.4 125.12 $4,379

At very low levels of sigma, any process is unlikely to be profitable. The higher the sigma
level, the better the bottom line is likely to be.

Organizations and teams can calculate the sigma level of a product or process using the
equation below:

Consider a process in a marketing department that distributes letters to customers or

prospects. For the purposes of the example, imagine that the process inserts 30,000 letters
in preaddressed envelopes each day. In a given business week, the process outputs 150,000
letters.

The marketing department begins receiving complaints that people are receiving letters in
envelopes that are addressed to them, but the letters inside are addressed to or relevant to
someone else. The marketing department randomly selects 1,000 letters from the next
of them have errors. Applying that to the total amount, they

12
 SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

estimate that as many as 750 letters could have errors. (Sampling and extrapolation are
covered in depth in the advanced chapters on statistics.)

The letter process has 150,000 opportunities for error each week and an estimated 750
defects.

((150,000 750) / 150,000) * 100 = a yield of 99.5

Look up a yield of 99.5 in the abridged Sigma table below

described above is currently between 4 and 4.1 sigma.

Yield % DPMO Sigma Level

99.7450 2,550 4.3

99.6540 3,460 4.2

99.5340 4,550 4.1

99.3790 6,210 4.0

99.1810 8,190 3.9

Sigma levels provide organization with a high-level look at how a process is performing, but

process an organization should improve first. Leadership should also consider costs,
resources, and the estimated impact of improvements.

13
 WHAT IS SIX SIGMA?

For example, consider these processes that might be found in a food processing plant:

Process Performance Metric(s) Current Sigma Level

Attaching a decorative Decorative touch is 2.2

element to food item centered on food
product and stable so it

Packing product Product is sealed for 3.1

freshness

Shipping of product Product reaches the right 4.3

customer in a timely
manner

A glance at sigma levels indicates that the process that attaches the decorative element is in
most need of improvement. While that process has the highest rate of defects, leadership
within the plant would have to ask themselves: How much does that matter to the
customer, and what is the hit to the bottom line?

reaches the right location. Since bad product has to be thrown away, the most expensive
errors might be associated with improper sealing during packing. The plant is likely to use
resources to improve the packing process before addressing the decorative element issue.

After the packing process is improved, the plant might then consider whether to improve
the decorating process or the shipping process. As part of that consideration, the company
might conduct customer surveys to reveal that some customers have stopped buying the
product because of the decorative element issue. An analyst estimates that the loss of sales
related to that issue are costing the company $1,000 a week. Shipping issues are costing the
company $500 a week.

Should the company address the costlier issue first? What if you were told that the shipping
process could be improved with staff training sessions, while the decorative element issue
required an expensive machinery update? Sometimes, organizations have to consider the
expense of an improvement. A

14
 SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

lucrative since those improvements take time and money. A Six Sigma culture is about
continuous improvement, which means teams consider all options before embarking on the
most lucrative improvement measures.

Organizations can impact their sigma level by integrating core principles from the Six Sigma
methodology into leadership styles, process management, and improvement endeavors.
The principles of Six Sigma, and the tools used to achieve them, are covered in detail in
various sections of this book, but some common ideas are introduced below.

improvements for the sake of driving up sigma levels. A primary principle of the
methodology is a focus on the customer. In
Customer (VoC) and ways for establishing what the customer really wants from a product or
process. By combining that knowledge with measurements, statistics, and process
improvement methods, organizations increase customer satisfaction, ultimately bolstering
profits, customer retention, and loyalty.

A detailed understanding of the customer and customer desires not only lets businesses
customize product offerings and services, but it also lets organizations:

Offer additional features customers want and are willing to pay for
Prioritize product development to meet current needs
Develop new ideas based on customer feedback
Understand changing trends in the market
Identify areas of concern
Prioritize work on challenges based on how customers perceive various problems or
issues
Test solutions and ideas before investing time and money in them

Value Streams The value stream is the sequence of all items, events, and people required to
produce an end result. For example, the value stream for serving a hotdog with ketchup to
someone would include:

A hotdog supplier
A bun supplier

15
 WHAT IS SIX SIGMA?

A ketchup supplier
Hotdogs
Buns
Ketchup
A cooking procedure for the hotdog
A pot
Tongs
Someone to do the cooking
A plate
Someone to put the hotdog into the bun
Someone to put the ketchup on the hotdog
Someone to put the completed hotdog onto a plate
Someone to serve the hotdog to another

If you combine all of the above processes into a pictorial representation of exactly how
these elements become the served hotdog, then you have a value stream map.

The purpose for determining a value stream for a process is that you can identify areas of
concern, waste, and improvement. In the above process, are there four different people
putting the hotdog together and serving it, or is one person doing all four of those tasks? Is
the supplier a single grocery store, or are you shopping for items at various stores and why?
Do you get savings benefits to offset the added time spent working with multiple suppliers?
These are some examples of the questions you can reveal and answer during value stream
mapping.

Inherent in the Six Sigma method is continuous process improvement. An organization that
completely adopts a Six Sigma methodology never stops improving. It identifies and
prioritizes areas of opportunity on a continuous basis. Once one area is improved upon, the
organization moves on to improving another area. If a process is improved from 4 Sigma to
4.4 Sigma, the organization considers ways to move the sigma level up further. The goal is

organization while maintaining other goals and requirements, such as financial stability, as
quickly as possible.

16
 SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

One of the ways to continuously improve a process is to reduce the variation in the process.
Every process contains inherent variation: in a call center with 20 employees, variation will
exist in each phone call even if the calls are scripted. Inflection, accents, environmental
concerns, and caller moods are just some things that lead to variation in this circumstance.
By providing employees with a script or suggested comments for common scenarios, the
call center reduces variation to some degree.

Consider another example: A pizzeria. The employees are instructed to use certain amounts
of ingredients for each size of pizza. A small gets one cup of cheese; a large gets two cups.
The pizzeria owner notes a great deal of variation in how much cheese is on each pizza, and
he fears it will lead to inconsistent customer experiences. To reduce variation, he provides
employees with two measuring cups: a 1-cup container for small pizzas and a 2-cup
container for large pizzas.

The variation is reduced, but it is still present. Some employees pour cheese into the cups
and some scoop it. Some fill the cups just to the rim; others let the cheese create a mound
above the rim. The owner acts to reduce variation again: he trains all employees to fill the
cup over the rim and use a flat spatula to scrape excess cheese off. While variation will still
exist due to factors such as air pockets or how cheese settles in the cup, it is greatly
reduced, and customers experience more consistent pizzas.

Remember the hotdog example for value streams? We asked the question: do four different
people act to place the hotdog in the bun, put the ketchup on the hotdog, plate the hotdog,
and serve it? If so, does the process take more time because the product has to be
transferred between four people? Would it be faster to have one person perform all those
ac in the process in this case, waste of
conveyance.

Removing waste items, actions, or people that are unnecessary to the outcome of a
process reduces processing time, opportunities for errors, and overall costs. While waste
is a major concern in the Six Sigma methodology, the concept of waste comes from a
methodology known as Lean Process Management..

17
 WHAT IS SIX SIGMA?

Implementing improved processes is a temporary measure unless organizations equip their

employees working with processes to monitor and maintain improvements. In most
organizations, process improvement includes a two-pronged approach. First, a process
improvement team comprised of project management, methodology experts, and subject-
matter experts define, plan, and implement an improvement. That team then equips the
employees who work directly with the process daily to control and manage the process in
its improved state.

Often, Six Sigma improvements address processes that are out of control. Out of control
processes meet specific statistical requirements. The goal of improvement is to bring a
process back within a state of statistical control. Then, after improvements are
implemented, measurements, statistics, and other Six Sigma tools are used to ensure the
process remains in control. Part of any continuous improvement process is ensuring such
controls are put in place and that the employees who are hands-on with the process on a
regular basis know how to use the controls.

Six Sigma is not without its own challenges. As an expansive method that requires
commitment to continuous improvement, Six Sigma is often viewed as an expensive or
unnecessary process, especially for small or mid-sized organizations. Leadership at Ideal
Aerosmith, a manufacturing and engineering company in Minnesota, was skeptical of Six
Sigma ideas and the costs associated with implementing them. Despite reservations, the
company waded into Six Sigma implementations, eventually seeing worthwhile results after
only 18 months. Those results included a production improvement of 25 percent, a 5
percent improvement in profits within the first year, and a 30 percent improvement in
timely deliverables.4

Some obstacles and challenges that often stand in the way of positive results from Six Sigma
include lack of support, resources, or knowledge, poor execution of projects, inconsistent
access to valid statistical data, and concerns about using the methodology in new industries.

http://www.qualitydigest.com/april05/articles/02_article.shtml
18
 SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

Six Sigma requires support and buy-in at all levels of an organization. Leaders and
executives must be willing to back initiatives with resources financial and labor related.
Subject-matter experts must be open to sharing information about their processes with
project teams, and employees at all levels must embrace the idea of change and
improvement and participate in training. Common barriers to support include:

Leaders willing to pursue improvements initially but who lose interest in overseeing
and championing projects before they are completed
Staff that is fearful of change, especially in an environment when change has
historically caused negative consequences for employees
Employees who are resistant to change because they believe improvements might
make them obsolete, drastically change their jobs, or make their jobs harder
Department heads or employees who constantly champion their own processes and
needs and are unwilling to enter into big-picture thinking

Lack of resources can be

barrier. Lack of knowledge about how to use and implement Six Sigma is one of the first
issues small- and mid-
dedicated resources to handle continuous process improvement, but the availability of
resources and Six Sigma training makes it increasingly possible for organizations to use
some of the tools without an expert or to send in-house staff to be certified in Six Sigma.

Companies implementing Six Sigma for the first time, especially in a project environment,
often turn away from the entire methodology if the first project or improvement falls flat.
Proponents of Six Sigma within any organization really have to hit it out of the ballpark with
the first project if leadership and others are on the fence about the methodology. Teams
can help avoid poor project performance by taking extreme care to execute every phase of
the project correctly. By choosing low-risk, high-reward improvements, teams can also stack
the deck in their favor with first-time projects. The only disadvantage with such a tactic is
that it can be hard to duplicate the wow factor with subsequent improvements, making it
important to remember that long-term implementation and commitment is vital in Six
Sigma.
19
 WHAT IS SIX SIGMA?

Data and analytics issues are a common challenge for organizations of all sizes. Gaining
access to consistent and accurate data streams and applying statistical analysis to that
data in an appropriate manner is difficult. Some data-related challenges include:

Discovering that an important process metric is not being captured

The use of manual data processes in many processes
Automated data processes that capture enormous amounts and create scope
challenges
Data that is skewed due to assumptions, human interaction in the process, or
incorrect capture
Lengthy times between raw data capture and access
Industry or company compliance rules that make it difficult to gain access to
necessary data

Six Sigma originated in the manufacturing industry and many of the concepts and tools of
the methodology are still taught in the context of a factory or industrial environment.
Because of this, organizations often discount the methods or believe they will be too
difficult to implement in other industries. In reality, Six Sigma can be customized to any
industry.

20
While the roots of Six Sigma are commonly attributed to companies such as Toyota and
Motorola, the methodology is actually grounded in concepts that date as far back as the
19th rtant to understand the
difference between traditional quality programs, such as Total Quality Management, and
continuous process improvement methods, such as Six Sigma.

Most modern quality and improvement programs can be traced back to the same roots.
Both quality programs and continuous process improvement methods look to achieve goals
such as reducing errors and defects, making processes more efficient, improving customer
satisfaction, and boosting profits. But quality programs are concerned with achieving a
specific goal. The program either runs forever, constantly working toward the same goal, or
it achieves the end goal and must be reset for a new goal.

Six Sigma seeks to instill a culture of continuous improvement and quality that optimizes
perfo
Sigma that lets organizations enact both small and sweeping improvements that drastically
impact efficiencies and costs. Six Sigma does work toward individual goals with regard to
each project, but the projects are part of the overall culture of improvement that, in
practice, is never done. Six Sigma creates safeguards and tactics so that, even after a project
is considered complete, controls are in place to ensure progress continues and it is
impossible to revert to old ways.

Six Sigma applies statistics to define, measure, analyze, verify, and control processes. In fact,
Six Sigma teams usually use methodologies known as DMAIC or DMADV to accomplish
improvements and develop controls for processes. DMAIC stands for Define, Measure,
Analyze, Improve, and Control. These are the five phases of a Six Sigma project to improve a
process that already exists. When developing a new process, teams use DMADV, which

21
 SIX SIGMA HISTORY AND APPLICATION

stands for Define, Measure, Analyze, Design, and Verify. Both methods are discussed in
Chapter 11, and Unit 3 provides in-depth information about each phase of DMAIC.

The roots of statistical process control, which provide a backbone for Six Sigma methods,
began with the development of the normal curve by Carl Friedrich Gauss in the 19th
century. We know today that the normal curve is just one of several possible probability
distribution models. It is perhaps the most widely used model, and the other models
developed from the normal curve. Probability distribution models are discussed in later
chapters on statistics

In the early part of the 20th century, statistical process control received another big boost
thanks to contributions from an engineer and scholar named Walter Shewhart. Shewhart's
contributions to quality are many, but two specific ideas stand out. First, Shewhart was the
first person to closely relate sigma level and quality. He defined a process in need of
correction as one that is performing at three sigma. If you look back to Chapter 1 and the
theoretical Amazon example, the cost difference between four sigma and three sigma is
over $78 million; in comparison, the difference between five and four sigma is only
approximately $7.6 million. Because errors and costs exponentially increase as sigma level

as a method seeks to move ever toward less than 3.4 defects per million opportunities

22
 SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

(dpmo), it is also true that if the quality of a process decreases, as it approaches three
sigma, the costs associated with errors increase substantially.

Second, Shewhart is considered the father of control charts. Control charts, which are
covered in depth in the chapters on advanced statistics, are a critical component of
statistical process control that lets organizations maintain improved performance after a Six
Sigma initiative. At a time when scholars were writing about the theoretical application of
statistics in a growing number of fields, Shewhart developed ways to apply these concepts
to manufacturing and industrial processes specifically.

During the same time period, W. Edwards Deming was working for the U.S. Department of
Agriculture. A physicist and mathematician, Deming was in charge of teaching courses at

Deming brought Shewhart's statistical concepts to the United States Census Bureau,
applying his theories outside of an industrial or manufacturing environment for possibly the
first time.

-do-check-act cycle. The idea is that

improvement comes when you recognize there is a need for change and make a plan to
create improvement. Next, you do something by testing your ideas. Using the results of the
test, you check or verify that your improvements are working. Then you act, bringing your
improvements to a production environment or scaling improvements outside of the test
environment. The fact that PDCA is a cycle means it never ends; there are always
improvements to be made. This is a core tenet of Six Sigma.
23
 SIX SIGMA HISTORY AND APPLICATION

Following World War II, Deming worked in Japan on behalf of the United States government
in several capacities. While in post-war Japan, Deming befriended statisticians and
convinced at least one notable engineer that statistical process control was relevant to
Japan's need to drastically drive economic and production performance to overcome
damage from the war. In the end, Deming became a valued teacher and consultant to
manufacturing companies in Japan, planting the ideas and concepts that would soon
become the Toyota Production System, or Lean Six Sigma.

What is Jidoka?
Jidoka is a principle that creates
control of defects inside a
Deming's teachings and the need for Japanese industry business process. Instead of
to make a successful comeback following a identifying defects at the end of
catastrophic war combined to bear fruit for Toyota. the production line and
attempting to trace errors back
prior to WWII, but improved performance and to a source, jidoka demands that
efficiency became a more critical goal given the nature a process stop as soon as errors
of Japan's economy and resources in the 1940s and are detected so improvements
50s. Taking manufacturing ideas attributed to Henry or troubleshooting can happen
Ford, Toyota leaders applied statistics and new quality immediately.
concepts to create a system they felt would increase
For jidoka to work properly,
production and allow for variable products while
machines are often equipped to
reducing costs and ensuring quality.
recognize bad outputs from
Several individuals were instrumental in the ultimate good outputs; the machines are
development of the Toyota Production System. They also equipped with a notification
infused the process with automated machinery, quality of some type to spark human
controls to keep defects from occurring, and efficiency interaction in the process when
tools that had not yet been applied with such detail things go awry.
and consistency. One man, Kiichiro Toyoda, had
previous factory experience. In his previous jobs, he

24
 SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

added efficiencies to processes in textile mills through conveyor and other automated
systems. Toyoda introduced the same concepts on certain lines in the Toyota
manufacturing process. Later, Eiji Toyoda and Taiichi Ohno introduced concepts known as
Just-in-Time and jidoka, which are the pillars of the Toyota Production System.

The principles driving Toyota's system, and later, the foundation of Lean Process
Management or Lean Six Sigma, include:

Defining customer values

Identifying the value stream for customer needs and desires
Identifying waste in the process
Creation of a continuous process flow
Continually working to reduce the number of steps and time it takes to reach
customer satisfaction

Lean management is highly concerned with removing waste from any process. Waste
increases costs and time spent on a process, making it undesirable in any form.

Though the basis for Six Sigma was laid in the late 19th and early 20th centuries, it wasn't
until the mid-1980s that these concepts saw large-scale success in the United States.
Decades after Toyota developed its system, engineers at Motorola began to question how
effective their quality management programs were. Those questions first arose after a
Japanese company took over a Motorola television manufacturing plant. By applying Lean
concepts, the new company began creating televisions that demonstrated 1/20 th the
amount of

At the time, departments across Motorola measured defects as a ratio of a thousand

opportunities. Bob Galvin, the CEO of Motorola, issued a challenge to his team. He wanted
to see an improvement in quality and production not just any improvement; he wanted a
ten-fold improvement in half a decade. Engineer Bill Smith and a new addition to the
Motorola team Dr. Mikel Harry began to work on the problem.

The team realized that measuring errors against a thousand opportunities didn't provide the
level of detail needed for true statistical process control. Instead, the engineers wanted to
measure defects against a million opportunities. We know that sigma levels were already
defined and the idea of using sigma levels as a measure of quality began with Shewhart. It

25
 SIX SIGMA HISTORY AND APPLICATION

wasn't a long jump for the Motorola engineers to make from their desire for more accurate
data to the basic concepts of Six Sigma as both a goal and a methodology.

Throughout the next two decades, Motorola worked to perfect its Six Sigma methodology,
seeing positive results along the way. In addition to statistical tools, the team created a
step-by-step process by which any team--in almost any industry--could make gains and
improvements. For the first time, this type of statistical process control was taken out of the
manufacturing environment on a large scale company-wide. Motorola applied the method
to customer service, engineering, and technical support. It used the process to create a
collaborative environment between stakeholders inside and outside of the organization. It
was highly successful; according to Motorola, the company saved more than $16 billion as a
result of continuous process improvement initiatives within 12 years.5

Motorola did more than improve its own systems and products, though. Galvin directed his
team to share Six Sigma with the world. Motorola and its team published articles and books
on the Six Sigma method and implemented efforts to train others. In this way, they created
a methodology based on statistics that could be taught and implemented within any
organization or industry.

After leaving Motorola, Dr. Harry joined Asea Brown Boveri. At ABB, Harry worked with
Richard Schroeder, who would also become a champion for Six Sigma. In fact, the two men
later cofounded the Six Sigma Academy. At ABB, Harry came to realize a key idea in the
evolution of Six Sigma: business, or profits, in some ways came before quality. Quality, in

Because the individuals with the ability to decide in favor of Six Sigma initiatives were highly
motivated by dollars, Harry incorporated financial tactics into the Six Sigma methodology.
For the first time, the method was focused on the bottom-line as a primary goal with other
concerns and goals stemming from financially-led goals.

In 1993, both Schroeder and Harry changed jobs, joining the team at Allied Signal. Allied

executives and other high-level leaders experienced knowledge barriers while attempting to
interact and collaborate with analysts, process engineers, and Six Sigma experts. Bossidy

5
-to-six-
sigma/history/history-six-sigma/
26
 SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

suggested that leadership at a company had to be well-versed in Six Sigma to pick the right
projects for success and support those projects on a company-wide basis to ensure success.

Harry, who is sometimes referred to as the father of Six Sigma, created a system for
educating executive leaders. In conjunction with others at Allied Signal, he developed
systems that allowed Six Sigma to be effectively deployed by leadership throughout an
organization in its entirety.

Around the same time, GE CEO Jack Welch entered into the Six Sigma arena. Prior to
learning about Six Sigma, Welch had stated he was not a proponent of quality measures.
ticized quality programs as heavy-handed approaches that did little to
deliver results. Welch invited Larry Bossidy to speak at a GE corporate meeting in 1995. He
also requested an analysis regarding the benefits of implementing Six Sigma at GE. At that
time, GE was performing at between three and four sigma. The potential savings should the
company rise to six sigma were enormous; estimates were $7 to $10 billion.6

Welch is known as a champion of Six Sigma not because he contributed in major ways to the
development of statistical process controls or the Six Sigma toolsets, but because he
demonstrated exactly how leaders should approach Six Sigma. He also made GE a
historically successful Six Sigma organization by tying Six Sigma goals to employee reward
structures. Employees were no longer only compensated based on financial performance
factors; they were also evaluated based on Six Sigma performance. Suddenly, employees at
every level had a personal reason to become involved in continuous process improvement,
and employees and managers were supplied with the Six Sigma training to succeed.

Following the success of corporations such as GE and Motorola, companies across the
country rushed to implement Six Sigma. Unfortunately, in the rush to implement the
process, many organizations executed improvements poorly or failed to gain an adequate
understanding of statistical process control before moving forward with improvements.
Although Six Sigma methods have been used by organizations to gain millions even
billions in savings and efficiencies, some companies walked away with a bad taste for the
process. That bad taste has resulted in the following misconceptions and myths that are still
prevalent today in many industries:

http://www.pqa.net/ProdServices/sixsigma/W06002009.html
27
 SIX SIGMA HISTORY AND APPLICATION

Six Sigma is solely concerned with metrics and ignores common sense. The opposite
is actually true: Six Sigma often starts with traditional common sense ideas, often
arrived at through brainstorming, and validates those assumptions with data. The
reason for this myth is twofold. First, managers and others who are used to making
calls without being questioned are suddenly questioned in a Six Sigma environment.
Not only are they questioned, but hard data sometimes proves them wrong.
Second, in some cases data is improperly used to support conclusions that are
against common sense or tradition. When those conclusions turn out to be faulty,

of the statistical theories involved.

Six Sigma is too expensive. While enterprise-wide adoption of Six Sigma can be
costly at first, due in part to training needs, slowly integrating the concepts into a
company often costs very little in the long run. Organizations have to balance how
they adopt Six Sigma with budgetary concerns but when implemented correctly,
Six Sigma generally leads to savings that more than cover its initial investment.
Six Sigma can fix anything. Opposite the nay-sayers are Six Sigma cheerleaders who
believe they can apply the method like a salve to any problem. While Six Sigma can

culture or people. If morale or other human resource problems are at the root of an
issue, statis
work with or is performing poorly, Six Sigma can be used to improve the process,
thereby improving morale.

Six Sigma is applied via a controlled project selection and management process. Once areas
of concern are identified, leaders usually turn to analysts, Six Sigma experts, and subject-
matter-experts for cost-benefit analyses. Six Sigma teams attempt to quantify how broken a
process is (by calculating sigma level, costs of defects, downtime, and other metrics) and
how much it might cost to address the problem. Problems are then prioritized according to
e. Teams begin working
through the priority list, returning to the analysis from time to time to ensure the list has
not changed. The majority of this book covers the methods by which teams identify and
address problems using Six Sigma.

28
 SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

Possessing a Six Sigma certification proves that an individual has demonstrated practical
applications and knowledge of Six Sigma. Some organizations offer in-house certification
processes. Most people seek certification by enrolling in online or onsite Six Sigma training
course. Most organizations that offer Six Sigma education also offer a path to certification.
You can take courses for certification at various levels; Six Sigma levels are differentiated by
belt level.

A certified Six Sigma White belt is familiar with the basic tenets of the Six Sigma

White belt training is a good introduction to Six Sigma for auxiliary staff members within an
organization and can provide the information necessary for understanding why project
teams do what they do. The training lets employees review project processes, understand
information presented in milestone meetings, and better participate in project selection
processes. White belt training can also be used across all levels of employees when
organizations are attempting to implement a Six Sigma culture. It is worth noting that White
Belt training usually only provides a very basic introduction and overview of Six Sigma, so
much so that not all Six Sigma professionals recognize it as a true Six Sigma certification.

A yellow belt certification is a step above white belt: it is still considered a basic introduction
to the concepts of Six Sigma, but a yellow belt learns basic information about the DMAIC
method often used to improve processes. The following concepts are often included in Six
Sigma yellow belt training:

Six Sigma roles

Team development and management
Basic quality tools such as Pareto charts, run charts, scatter diagrams and
histograms
Common Six Sigma metrics
Data collection
Measurement system analysis
Root cause analysis
An introduction to hypothesis testing

29
 SIX SIGMA HISTORY AND APPLICATION

At the yellow belt level, training is often geared toward understanding of the overall

hypothesis testing, but they must understand the language of hypothesis testing and the
conclusions that are drawn from such tests. Yellow belts are often employees who need to
know about the overall process and why it is being implemented.

Certified green belts work within Six Sigma teams, usually under the supervision of a black
belt or master black belt. In some cases, green belts might lead or handle smaller projects
on their own. Green belts are generally equipped with intermediate statistical analysis
capabilities; they might address data and analysis concerns, help Black Belts apply Six Sigma
tools to a project, or teach others within an organization about the overall Six Sigma
methodology.

Green Belts can be middle managers, business analysts, project managers, and others who
have a reason to be involved regularly with process improvement initiatives but who might
not be a full-time Six Sigma expert within an organization. Sometimes, Green Belts are
considered the worker bees of the Six Sigma methodology because they undertake most of
the statistical data collection and analysis under the supervision of certified Black Belts.

The following concepts are often included in Green Belt training:

All of the information listed for yellow belt certification

Failure mode and effects analysis
Project and team management
Probability and the Central Limit Theorem
Statistical distributions
Descriptive statistics
How to perform basic hypothesis testing
Waste elimination and Kaizen
Basic control charts

A certified Six Sigma Black Belt usually works as the project leader on process improvement
projects. They might also work within management, analyst, or planning roles throughout a

30
 SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

company. Common minimum requirements for black belt certification include everything
listed for yellow and green belts in addition to:

Advanced project and team management skills

Knowledge of the expansive list of Six Sigma brainstorming and project tools
Intermediate to advanced statistics
An understanding of other process improvement and quality programs, including
Lean and Total Quality Management
An ability to design processes
Advanced capabilities for diagraming processes, including flow charts and value
stream maps
Use of software to conduct analysis, such as Excel or Minitab

A Master Black Belt is the highest certification level achievable for Six Sigma. Within a
business organization, Master Black Belts usually manage Black Belts and Green Belts,
consult on especially difficult project concerns, offer advice and education about
challenging statistical concepts, and train others in Six Sigma methodology.

Most certification programs require individuals to pass an exam for certification; some
require that green and black belt candidates also demonstrate their knowledge in the form
of Six Sigma project experience.

If an exam is required for white or yellow belt certification, it is usually fairly short and
covers basic concepts about the methodology. Green belt exams are longer and might
include questions about statistics and some basic calculations. Black belt exams often take
up to four hours to complete; they test for understanding and application. Exams might
include difficult statistical problems or questions about how a project leader might handle
various situations. While exams differ by organization, this book is designed based on The
published body-of-knowledge requirements.

Note: For those that are utilizing this textbook in preparation for one of the certification
exams administered directly by the Council for Six Sigma Certification
(www.sixsigmacouncil.org), the following material should be reviewed as follows in
preparation for the open-book examination(s):

White Belt Certification or Lean White Belt Certification: Chapter 1 thru Chapter 3

31
 SIX SIGMA HISTORY AND APPLICATION

Yellow Belt Certification or Lean Yellow Belt Certification: Chapter 1 thru Chapter 11
Green Belt Certification or Lean Green Belt Certification: Chapter 1 thru Chapter 24
Black Belt Certification or Lean Black Belt Certification: Chapter 1 thru Chapter 33
Master Black Belt Certification or Lean Master Black Belt Certification: Chapter 1 thru
Chapter 33

32
related to a number of other quality-driven initiatives developed over the past century. This
is true in part because all successful businesses ultimately seek to do the same thing: serve
a customer a product or service they need while making as much profit as possible.

While Six Sigma encompasses all the tools you need to approach virtually any problem of
process, familiarity with other types of process improvement and quality methods is
important. Some of these methods, such as Lean and JumpStart, add value within a Six
Sigma approach. Others might be used by outside resources alongside a Six Sigma project.
se or work with some of these programs, you will need to communicate
with leadership and business partners who are more familiar with other methods. The
ability to frame Six Sigma concepts in a more global quality management approach can help
you win support for your own projects.

Lean principles often go hand-in-hand with Six Sigma principles. While Lean originally
developed as a concept for reducing waste in a manufacturing environment, the ideas of
Lean Process Management can be applied to any process that involves the movement or
creation of goods or services. This is true even if those services are virtual or digital, such as
in a computerized workflow process.

One of the ways that Lean is similar to Six Sigma is that it is concerned with continuous
improvements; like Six Sigma, Lean provides waste-removal tools so daily control and
improvements can be made to processes. In fact, one of

33
 OTHER PROCESS IMPROVEMENT AND QUALITY METHODS

The purpose of every change in a Kaizen environment is to eliminate waste and/or create
more value for the customer on a continuous basis.

Lean Process Management can be deployed within a project environment or in daily

production. Like Six Sigma, Lean is more about an overall culture of quality than a single
quality event. Many organizations use Lean principles to make improvements in processes.
By simply instituting some of the Lean principles, managers can drastically increase
production and reduce costs for their departments.

Because Lean principles are so effective and fit so well with Six Sigma principles, for the
purpose of this book, we will often treat Lean as a part of the Six Sigma methodology.

Total Quality Management, or TQM, is a phrase well-known by anyone who worked in

business in the last quarter of the 20th century. The TQM approach to quality is one of the
first formal methods enacted in business environments in the United States. Originally

across the country until the 80s. At one point, TQM was so popular with executives and
other leaders that it actually became something of a joke among certain workforces who
believed that much effort and expense was expended on quality without an equal resulting
benefit. In fact, if you remember from the last chapter, Jack Welch at GE felt this way.

While Total Quality Management programs were often somewhat lackluster when it came
to results, the method was an essential stepping point to current improvement and quality
methods such as Six Sigma. TQM was not without its results: as with any method, results
depended highly on the way the program was implemented and the culture of the
organization. For this reason, TQM and its variations are still in play in many industries
today. Some requirements for a successful TQM program include:

A strict quality commitment at all levels of the organization, especially among

leaders
Empowered employees who can make quality decisions while working within the
process without constantly seeking leadership approval for those decisions
A reward and recognition structure to promote quality work so that employees
have a reason to make quality-making decisions
Strategic planning that takes quality and quality improvement goals into account
when making long-term decisions
Systems that let organizations make improvements and monitor quality

34
 SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

Successful TQM initiatives require eight key elements: ethics, integrity, trust, training,
teamwork, leadership, recognition, and communication. You can view these elements as if
they were part of the components needed to build a high-quality, lasting building. Ethics,
integrity, and trust become the foundation for quality. Training, teamwork, and leadership
are the bricks by which quality organizations are built. Honest, open, and concise
communication is the mortar that binds everything else together, and recognition is the
roof that covers everything, providing employees with a reason to seek and maintain
quality.

One of the biggest advantages of the TQM mentality is that it began to force organizations
to see themselves as one entity rather than a number of loosely related entities or
departments. Prior to the quality methods developed in the last half of the 20th century,
many organizations were run via heavily siloed departments. One department often did not
understand what another was doing, which caused a great deal of rework and waste. Each
department might seek higher quality levels or process improvements, but in the end, the
organization was only as strong as the weakest element.

TQM began to change departmental thinking on a massive scale: organizations began to

take enterprise approaches to decision making, quality, and customer service. Business
leaders started to look at companies as a series of linked processes operating toward a
single end goal. Within the bounds of TQM, the ideas for business process reengineering
began to develop.

Organizations using TQM often experienced benefits such as:

Improved employee engagement and morale

A reduction in production or product costs
Decreased cycle times
More satisfied customers

Six Sigma, Lean and TQM are all concerned with making continuous changes on both a large
and small scale that bring an organization ever closer to a model of perfection. In the case
of Lean, that model is a process that has zero waste; in Six Sigma, the model is statistically 6
sigma. In TQM, organizations often define their own version of perfection before working
toward it. Business Process Reengineering, or BPR, is less concerned with incremental

35
 OTHER PROCESS IMPROVEMENT AND QUALITY METHODS

quality wins and more concerned with a radical change across an entire organization or
process architecture.

Business process reengineering, which is also called business process redesign, is most often
concerned with the technical processes that occur throughout an organization. Those
processes might include systems, software, data storage, cloud and web processes, and
computer-based workflows operated and maintained by human users. Because of the
intense integration of automation and computer elements into processes with BPR,
organizations that enter BPR endeavors have to rely heavily on both inside and outside
technical resources. Inside resources provide programming, integration, and
troubleshooting services as processes are developed or redesigned. Outside resources can
be BPR consultants, contracted programmers and developers, or vendors bringing new
software products to the table.

As you can probably imagine, BPR initiatives can be costly, which is why they are often
deployed only when an organization expects exponential gain or has determined that
current processes are obsolete or badly broken.

there is with Six Sigma. Most projects go through planning, design, and implementation
phases. During planning, teams use process mapping and process architecture principles to
define enterprise-wide processes in their current state. Teams look for opportunities for
improvement and brainstorm new architectures for processes throughout the organization.

During the design phase, BPR teams use validation techniques 3 to ensure solutions they
are planning will work within the enterprise structure. They also begin to build tools and
programs to integrate the changes; technical teams might use the Scrum methods
described later in this chapter at this point in the process.

Finally, organizations implement the changes they have made. Since changes are often
programmatic in nature, implementation usually includes a rigorous change management
and testing procedure. Testing in technical environments includes steps such as:

Sandbox testing of basic functionality

Quality assurance testing by trained technical resources
Beta testing during which experienced subject matter experts vet all aspects of a
program in a limited live environment

36
 SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

A rollout of the program to the enterprise, often conducted in a phased approach

during which technical resources are on call to immediately resolve
troubleshooting issues
A conversion to regular function where technical resources are available in a
normal capacity to deal with occasional issues

As process improvement methods became increasingly popular in the 1980s and later,
individuals often took portions of one method or another and integrated it into new
improvement or quality programs. In this manner, companies outside of the manufacturing
industry began implementing bits and pieces of methods that incorporated Lean and Six
Sigma elements. One such program is known as Rummler-Brache.

Rummler-Brache was pioneered in the 80s by Geary Rummler and Alan Brache. They
developed what remains a proprietary program used by their own consulting firm, but
details of the method have been published and used by others. The method seeks to affect
positive change in processes and organizations by using a set of practical tools to address
business issues and process problems.

One of the foundational components of Rummler-Brache is known as the Nine Boxes

Model. The model is created by a matrix of three performance levels and three
performance dimensions. Performance levels are the performer, the process, and the
organization. Dimensions are management, design, and goal. When placed on a grid, the
levels and dimensions form nine boxes, as seen below.

37
 OTHER PROCESS IMPROVEMENT AND QUALITY METHODS

Management Design Goals

Performer Concerned with Concerned with Concerned with

feedback, the tools and performance
consequences, training needed metrics and
and rewards to do the job as requirements at
well as job an individual
documentation level

Process Concerned with Concerned with Concerned with

who owns the the design of the the
process and how process, work requirements of
they might space, or system the business and
improve it the customer

Organization Concerned with Concerned with Concerned with

overall leadership overall org charts operating plans
culture and the and process and top-level
requirements of architecture metrics
performance
evaluation

Rummler-Brache approaches improvement in six phases:

Improvement planning. During the first phase, leadership and subject-matter-

experts commit to making improvements and begin to identify opportunities for
change.
Definition. During the second phase, project goals and scopes are defined and
teams are formed to create improvements.
Analysis and Design. Teams use analysis to understand the current problem and to
define and validate workable solutions.
Implementation. Teams implement process changes. Depending on the type of
change, this might include programming changes, retraining staff, changes in
machinery or equipment, or policy changes.
Management of process. Teams monitor the process during and immediately
following the change to ensure improvements function as planned.
38
 SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

Processes are turned over to daily teams. Management of the process is turned
over to daily teams, often with some type of control in place to ensure continued
success.

Scrum is a project development method specific to Agile programming endeavors in

technical departments. Scrum is used when teams want to create new technical products or
integrate new developments on existing products within a short time frame. Commonly,
Scrum projects last between two and four weeks, which is traditionally a very tight timeline
for programming projects. Scrum was developed as programming and development teams
needed a way to meet continuous technical design and improvement needs from other
departments without substantially increasing programming, testing employee hours, or
hiring more technical staff. Scrum can also be used to drive faster times to production or
market for software and application products.

Scrum is a related concept to other process improvement initiatives discussed in the book
because many projects today call for some type of technical resource or change. While
project teams are working to validate and measure, technical departments often
simultaneously deploy Scrum concepts to meet development needs for the improvement
project by deadline.

Scrum projects feature three main phases:

The pregame. Development teams analyze available data and business

requirements. They use this information to come up with the concept for the new
product or upgrade. Often, this involves translating business and process concepts
into computer and technical concepts.
The game. Teams begin to develop the product via programming sprints. Sprints are
smaller phases of development that are completed in sequence, usually with a
review and validation of the work before moving on to the next sprint. By validating
work during development, teams are able to create working products faster.
The postgame. Even though validation occurs during development, teams still have
to follow quality assurance, testing, and change management procedures. Quality
preparation for product release is handled in the final phase.

39
 OTHER PROCESS IMPROVEMENT AND QUALITY METHODS

Like Rummler-Brache, the Customer Experience Management Method, or CEM Method,

was created by process improvement consultants to address needs in organizations outside
of manufacturing. CEM combines some process improvement tools with customer relations
management. It was developed in the 1990s by the Virgin Group and became popular
throughout the 90s and early part of the 21st century.

The CEM Method takes an outside-in approach to process improvement, focusing on what
the customer wants or needs and how each process in an organization serves that need.
The primary purpose of CEM is to align processes throughout an organization with customer
satisfaction goals. As such, even processes without a direct relation to customers are
defined in terms of customers.

to define how those processes can best serve customers. Shipments should arrive on time,
be accurate to orders, and shipping costs should be affordable.

In-house human resource processes are harder to link to customer-facing goals. However,
the morale and functionality of employees is directly related to how those employees can
serve customers. You can make a customer-facing statement about almost any process in
an organization in this manner. If organizations cannot link a process to the customer, then
they must ask whether the process is necessary or broken.

customer goals and the success of processes without collecting and analyzing customer
feedback. The advantage of CEM is that organizations are able to deploy customer-facing
tactics across the enterprise, which often results in enormous gains in customer
satisfaction, loyalty, and spending. A disadvantage of this method is that traditionally
inward-facing departments, such as human resources, legal, and accounting, often have a
difficult time implementing customer-focused cultural change.

JumpStart differs from the other programs and methods described in this chapter in that it
is a fast-paced method for identifying problems and solutions in a single session. JumpStart
can be used within almost all of the other methods described in this book as a way to spark
discussion regarding processes or to identify possible solutions. It can also be used as a
management tool for helping teams come to tenable solutions outside of project
environments or in the absence of project resources.
40
 SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

own, teams should not use this method to enact sweeping changes or attempt to improve
processes that could seriously impact customer experience or the bottom line. One
disadvantage of using JumpStart alone is that changes are sometimes made on a wait-and-
see mentality, which is safe for many inner-team changes but often dangerous for
department or enterprise-wide processes, or for making changes to processes that are
closely tied to regulatory or compliance rules.

JumpStart usually begins when leaders at some level identify an area of concern or
opportunity. The manager, supervisor, or other delegate identifies a team of employees
who they believe would offer appropriate insight on the issue at hand. In most cases,

they already know what is wrong. Instead, the group spends several hours brainstorming
root causes for the problem and coming up with possible solutions.

Six Sigma and other process improvement tools can be deployed during JumpStart sessions.
Fishbone diagrams and solutions selections matrixes, both covered in later chapters, can be
used to validate assumptions using only the knowledge of the people in the room and some
quick research.

The benefit of JumpStart is that it lets teams create and implement small-scale solutions
quickly, often providing problem resolution the same day. It also lets teams identify issues
that need to be addressed in a more comprehensive project environment.

Some organizations make use of various project improvement methods. As a Six Sigma
expert, you might have to champion your own method on occasion. Here are some reasons
to choose Six Sigma over other methods described in this chapter.

Six Sigma is designe

they only know
some metric is not performing as desired. For example, an organization might experience a
drop
can begin to seek the causes of the problem, prioritize them, and work toward solutions.

41
 OTHER PROCESS IMPROVEMENT AND QUALITY METHODS

Even when a problem is understood, if it is wide in scope and not well defined,
improvement projects that are not tightly managed can escalate in scope to a point that
they become unmanageable. In this situation, teams attempt to solve increasingly bigger
issues. As a result, no problem is ever completely solved. Six Sigma includes controls for
avoiding such scope creep so teams can make incremental improvements that steadily

If processes are complex and feature many variables, it is difficult to determine how to
approach a solution, much less define and measure success. Knowledge of statistical
analysis and process control lets teams approach problems that involve enormous amounts
of data and many variables. Through analysis and graphical representation, complex ideas
can be distilled to specific hypotheses, premises, and conclusions.

al process control component lets teams make more accurate

assumptions than almost any other method, it is very appropriate for situations that are
closely tied to revenue or cost. When a single tiny change can result in millions of dollars in
gains or losses, teams must validate assumptions with an extremely small margin of error.
Guesswork, basic research, and even years of experience cannot do that as accurately as
properly implemented Six Sigma methods.

42
 concepts in the previous three chapters because most Six Sigma
approaches today incorporate Lean concepts into problem solving and the control of a
process. In fact, organizations often use the term Lean Six Sigma when describing a process
improvement approach that incorporates tenants from both Six Sigma and Lean
methodologies. This is a popular approach because the greatest results usually come when
you improve a process so that both defects and waste are eliminated. That statement rings
-driven bottom-line or a customer-
satisfaction approach.

A Six Sigma defect

requirements in Chapter 8 when we define quality. For now, know that defects cost money
because businesses have to replace parts, equipment, or products that are not perfect.
Organizations also experience financial loss associated with defects when quality reputation
is so low that customers choose not to return or purchase from the company. From a
customer satisfaction standpoint, defects can increase the time it takes for a customer to
get what they want or can cause the customer to be unhappy with the end product or
service.

Waste costs money because it is unnecessary time, labor, or material in the process.

satisfactory outcome. In some cases, waste creates a customer satisfaction issue because it
holds up the process or introduces undesirable elements or defects in the end product.

touch on some Lean concepts for creating the most efficient processes.

Muda is a Japanese word that translates to waste. It describes a concept of being useless,
unnecessary, or idle. The concept that muda must be eliminated in a process is a driving
concept of the Toyota Production System and Lean manufacturing. Muda is a non-value-
added task (NVA) within a process. Some types of muda are easier to identify than others,
which is why Lean Six Sigma deploys tools such as value stream mapping. By understanding
43
 LEAN CONCEPTS

a process at all levels, teams are more likely to identify various forms of muda. According to
Taiicho Ohno, chief engineer for Toyota, there are seven muda, or resources that are
commonly misused and mismanaged: overproduction, correction, inventory, motion,
conveyance, over processing, and waiting.

Overproduction is one of the easiest forms of muda to spot, as it tends to result in what we
commonly think of as waste. Overproduction means a product, part, or service was
produced too fast, at the wrong time, or in too much quantity for the process. To
understand the idea of overproduction, consider a basic fast food restaurant that offers
hamburgers and French fries for lunch. The restaurant does not serve breakfast, and it
opens its doors at 11:00 a.m. for the lunch crowd.

If the cooks light up the grill at 11:00 a.m., then they might start the day behind, as it is
possible that several orders will be placed immediately. However, if the cooks start making
hamburger patties at 10:30 a.m., they might have patties that sit for some time before
being consumed, which leads to customer dissatisfaction or waste if the patties are thrown
out. Making 10 patties every 10 minutes starting at 10:30 a.m. is overproduction the
patties are being made too soon.

What if the restaurant owners have done some research and they know the average
number of orders between 11:00 and 11:15 a.m. on a Tuesday is 10 hamburgers? They
might instruct the cooks to begin making patties at 10:50 a.m. and to make 5 patties every
10 minutes. The goal is to align patty-making with customer orders so that wait times are
reduced but customers are still able to enjoy fresh patties.

By noon, the owners know orders tend to come in quickly, so they ask the cooks to make 15
patties every 10 minutes. By 2:00 p.m., however, the order traffic usually drops to 10
hamburgers an hour. If the cooks are still making 15 every 10 minutes, then the process
suffers from overproduction.

At some point, the traffic in the restaurant may call for made-to-order grilling a process
where the cook only makes hamburger patties as they are ordered to avoid wasting any
materials. By understanding the traffic trends in their restaurant, the owners are able to
estimate needs and create processes that reduce the amount of waste made in the kitchen
while still meeting the quality demands of customers.

Overproduction is most often associated with tangible outcomes from a process, though
the
44
 SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

business that prints business cards, stationary, invitations, and other documents. Perhaps
this business offers a printing and mailing service; to complete this service, the company
uses a machine-driven process that prints, folds, and stuffs mailings. The printer is capable
of delivering 1,000 pages an hour, but the folding machine is only able to fold 800 pages an
hour.

Even if a customer wants 1,000 pages printed and mailed, the printer is overproducing if the
first machine is set to operate at maximum speed. The process will take longer than one
hour because it is contingent upon the slowest machine. Since the overprodu
result in tangible waste the printed pages will eventually be folded and mailed the

temporary overproduction is detrimental to the process. Does the stacking of extra paper
before the folding process create an extra risk for error? Does operating the printing
machine at maximum capacity without necessity put unnecessary strain or wear and tear on
it? If the answer is yes to either question, then there exists waste that should be eliminated
from the process.

Overproduction can also exist with regard to reporting, digital assets, and preparation for
processes. Almost anyone working in a business environment is familiar with reporting
requirements just as almost anyone who has created reports knows the unfortunate truth
that the information often goes unread. Creating reports no one reads or creating highly
detailed reports when an overview would suffice is overproduction.

a process is also overproduction. Surgery centers

often prepare machines, equipment trays, and operating rooms before shifts begin. The
goal is to create efficient processes for any patient operation; staff must also be able to
access equipment quickly if issues arise during procedures. Preparing 20 trays of equipment
on a day when only 10 surgeries are scheduled might be considered overproduction if there
is no use for the extra trays.

In some situations, such as the medical example above, processes might call for slight
overproduction. If 10 surgeries are scheduled, staff might prepare 13 or 14 equipment
trays. This way, if an emergency surgery arrives, or if an issue comes up with an existing
tray, stand-by equipment is available. By understanding nuances within processes and
requirements, Six Sigma teams can better identify muda of overproduction versus
overproduction that might be required by regulation or problem-solving policies.

The key to eliminating overproduction is plan

each solution came when the process operator understood the needs inherent within the
45
 LEAN CONCEPTS

process. When the owners knew how many orders were likely, they were able to plan to
reduce waste. When the printer knows the capability of each machine in the process, he or
she can plan for the most efficient printing run.

Also known as muda of rework, this form of waste often plagues organizations that are keen
on traditional quality programs. In a desire to eliminate defects from the end product,
organizations institute in-process quality checks that route work with defects back for
correction. While rework might be necessary in some cases, especially if materials are
particularly valuable and work is worth saving rather than scrapping, it is still waste in the
process that should be identified and analyzed.

When rework occurs, it increases overall process time and uses additional labor and

might be found in the call center for a large automobile insurance company. Some of the
calls coming into the center are from individuals who have sustained damage to vehicles in
an accident. Consider the following process for handling such calls:

The caller reports damage to the car.

The representative records information in a computer form.
The representative makes a decision based on the information provided by the
caller:
o The claim is routed for immediate handling if it is deemed an emergency
o The claim is routed to team A if it is a single-car or act-of-God incident
o The claim is routed to team B if it is a multi-car accident

Now, imagine a claim that arrives in the work queue for an employee working on team B.
The information entered into the computer is incomplete, so the employee cannot
determine for sure if the claim is related to a multi-car accident. The employee might send
the claim back to the original queue, which means a call-center employee would have to call
the insured back to gather additional information. The claim would then be routed again,
according to the new information, and would wait in a work queue again.

One of the biggest problems with muda of correction is that a case can often be made that
the processing is necessary. Perhaps team B is not equipped to deal with single-car
incidents, so the work has to be rerouted if the claim is to be completed. But why not equip
down-line team members to handle any claim, or create a process for rerouting claims to
the correct team without sending the work back to the original queue?

46
 SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

Correction, or rework, can occur in any type of process. Manufacturing processes cull out
defective parts and products; sometimes materials are reworked for a better outcome and
sometimes they are scrapped also a form of waste. Call centers and digital work queues
l format. In some
cases, rework occurs not because of correction, but simply because departmental or worker
responsibilities overlap.

To eliminate rework or correction, organizations must use a twofold approach. First, the
root cause of the rework that which is causing the errors must be addressed. Is further
employee training required? Could a process be changed to make it more mistake proof? In
some cases, the principles discussed in later units on process control, including a strategy
called a poka yoke, can be deployed to make it more difficult to create defects than to not
create defects during a process. When defects are avoided, rework is also avoided.

In addition to addressing the root cause of errors, organizations should create quality steps
that reduce rework waste. In the example about the insurance company call center, we
noted that it would be more efficient for the down-line worker to reroute the claim to the
correct team than to return it to the original team. This method sometimes causes a
problem of culture, though; there is a feeling among leaders and staff that the first team or
first team member should be held accountable for the error. One way of seeking
accountability is to have that person correct his or her mistake. While reworking errors can

held accountable through goal-setting and metrics for the greater good of the organization
as a whole.

Muda of inventory is similar to muda of overproduction; in fact, overproduction can cause a

waste of inventory. Muda of inventory occurs when materials or inputs stack up before a
step in the process; this phenomenon is also called a bottleneck. Remember the printing
example for overproduction? If the first machine is set to the highest production level, it will
generate 1,000 pages an hour. The folding machine can only handle 800 pages an hour, so
you would have an inventory of 200 pages created every hour. If both machines ran at
highest production for four hours, the process would have developed an inventory of 800
pages. Until and unless the folding process becomes faster, then those extra pages are
wasted inventory and will continue to pile up.

47
 LEAN CONCEPTS

Inventory waste can occur when items are purchased or created before they are needed in
a manufacturing or service process. Inventory can also occur in work queues, digital data
queues, or even email inboxes. If you receive 300 emails a day and you are only handling 30
of them on a regular basis, you have a process problem with your communications.

While inventory waste can occur in any process, it is especially common in processes that
operate in batches. Traditional lean wisdom says to avoid batch process processes that
involve creating a certain number of products or outputs before pushing them down the
line. Reducing batch size lowers lead times the time it takes to deliver the end product. It
also reduces the amount of inventory that occurs before each step of the process.

While Lean mindsets usually push a batching-is-

and reduced lead times on individual outputs a
Consider a baker who is preparing an order of one dozen cupcakes. He could prepare the
order as a batch, or he could prepare each cupcake separately. Obviously, mixing the
ingredients and baking the cupcakes as a batch makes more sense. Decorating them as a
batch also works well the baker might frost all of the cupcakes, add piping to all the
cupcakes, and then top all the cupcakes with a candy. Batching works for the baker because
the first cupcake is not going to leave the bakery before the 12th is finished they all move
together because they are considered a unit.

You can also reduce waste of inventory by understanding a process and basing inventory
decisions on historic metrics. A shipping center that processes between 50,000 and 100,000

to make two.

Muda of motion has to do with how employees themselves move during a process. This
type of waste is often relevant to people-powered processes in manufacturing,
warehousing, shipping, delivery, or industrial fields, but waste of motion can even crop up
in processes that are computerized.

For example, if a data-entry employee has to click back and forth between screens when
entering information, this could be muda of motion. If the system is designed so that a
number is to be entered in one window and a second number entered in a different
window, the click between windows is wasting motion. Moving to another window involves
a mouse or keyboard action that could be eliminated if the data were entered on the same
window. It seems like an inconsequential detail, but imagine what happens when the data
48
 SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

employee completes this process hundreds or thousands of times a day. The waste can
absolutely add up:

If each data-entry employee completes 600 entries an hour on average, and there
are 10 employees on a team, the team is completing 48,000 entries each day.
If it takes only half a second more to toggle between the two windows, the team is
still spending a collective 400 minutes a day just over 6.5 hours toggling.
If the average hourly wage for a data worker is $10, the team is spending an extra
$325 each week almost $17,000 each year to cover the act of toggling between
windows.

Given those numbers, a programing change that includes both data elements on the same
window could save the company $17,000 in a single year. This same concept can be applied
to any form of muda when you can apply a time and dollar figure to the waste.

Other examples of muda of motion include a task that requires an individual to physically
move back and forth between work, extra motion that stems from a poor layout of work, or
movement that occurs when an employee leaves an area and has to return one or more
times because he or she forgot something. Public libraries have long employed a simple tool
to avoid muda of motion in the reshelving process: books are first ordered on a cart. The
cart allows an employee to carry many books without moving back and forth, and placing
the books in order on the cart lets the individual move through the stacks once. If books are
placed at random on the cart, the subsequent movement between shelves wastes motion
and time.

Streamlining company processes eliminates muda of motion, and data must be collected
and analyzed to identify unnecessary movement. A common tool used in manufacturing
and similar environments to track movement is known as a spaghetti diagram. Begin with a

doors, and walls. Observe an actual process, tracking any and all movements with a line on
the diagram.

When drawn correctly, the diagram looks like a string of spaghetti fell onto your page. Once
the process is complete, you can look at the diagram to see where the movements cross
paths multiple times or go out of the way unnecessarily. This helps you find opportunities
for streamlining the movement in a process

diagram only reveals a snapshot of movement in time; sometimes, it is worthwhile to

49
 LEAN CONCEPTS

complete more than one such diagram at different times of day or with different employees
to reveal an overall picture of a process and possible muda.

Muda of conveyance is similar to muda of movement except conveyance involves the

movement of outputs, products, or resources. It is sometimes also referred to as muda of
transportation. For example, in a doll-making facility, if the glue that binds doll eyes to doll
faces is kept in an inventory room and carried, as needed, to the process, there might be
muda of conveyance.

If an expense report is printed and then carried to a manager for approval who then routes
it in an inner-office envelope to a director, who then carries it to the accounting
department, the muda of conveyance is occurring. This is especially true because
appropriate technology used correctly and efficiently lets organizations handle expensive
reporting via computer with little conveyance at all.

If a plate is prepared by a chef and placed on a counter where a kitchen assistant moves the
plate to a different table where wait staff know to pick it up, conveyance is occurring. The
wait staff then carries the plate to the customer.

Conveyance can relate to physical movement of items or digital movement of data or

workflow. Email strings, which are present in many work environments, often present muda
of digital conveyance. A CEO might email a director with a request for data. The director
emails a manager, who emails a supervisor, who emails a subject-matter-expert. The SME
delivers the information to her supervisor, and the emails work their way back up the chain.
The same request and information was conveyed multiple times when it only may have
needed to be conveyed once. This allows for many opportunities for error.

Some might point to chain-of-command concerns, but for the purpose of this example, if
the data request is a repeated process, then it makes sense for the chain of command to
inform the CEO where this information comes from. To go even further, a truly Lean process
would require all waste be removed from this process. If the data requirement comes on a
weekly basis, then Lean ideals require that the SME automatically generate the data and
send it to the CEO weekly.

Physical conveyance is often easier to locate and address than digital conveyance. A
spaghetti diagram, process map, or value stream map can help you identify areas where
muda of conveyance might exist. Spaghetti diagrams work well in physical conveyance
situations, and process maps help you identify conveyance in digital settings.
50
 SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

Once you identify muda of conveyance, you can eliminate it by making changes in the

poor process design or work-area layout, it might be related to another form of muda.
Conveyance is often seen in processes that involve a lot of correction, because work is
transferred back and forth between staff or areas. By addressing the muda of correction,
you often also address the muda of conveyance.

Over-processing occurs when an employee or process inputs more resources into a product
or service than is valued by the customer. This could occur because of ignorance, a desire
for perfection, or even excitement. Sometimes over-processing occurs because an
had training on the most efficient way to handle a task. Other times, it
occurs because an employee or process is more thorough than is worthwhile. A goal of any
process should be to do just enough useful and necessary work to ensure that customer or
end-user expectations are met.

One example of over-processing often occurs in healthcare administrative offices during the

office attempts to verify that a person is covered by insurance for the services that are

policies, a staff member either checks benefits via a computer program or calls the
insurance company.

In most cases, the goal of insurance verification is simply to ensure that the insured is
covered by the plan for the date the services are to be rendered. Sometimes, however, an
office worker delves deeper into the verification, spending up to an hour on the phone with
an insurance company to receive detailed benefits or calling back to check with another
representative to ensure the original information provided was correct. While specific cases
exist that require in- re this level
of work. A staff member who is taking up to an hour to verify insurance coverage is

In consumer-centric processes, over-processing occurs when you put more into a product
than the value afforded by the customer. While product excellence is important, at some
point the work you put into a product exceeds that which is deemed necessary or useful by
the customer, and this is often tied to price point. A customer expects more out of a
premium, more expensive model, for example.

51
 LEAN CONCEPTS

An example of over-processing might be painting the bottom of an inside cabinet shelf

when no one will ever see it. In a restaurant, ironing table linens to an exacting standard
could be considered over-processing, especially if you remove the linens between meals for
washing. In most restaurants, much of the table linen drapes over the table; owners might
ams. In the
technical world, building an app with 100 features when 99 percent of people only want to
use 10 main functions can be considered both over-processing and over-producing.

A value stream map, covered in Chapter 35, is a good tool for identifying any points of over-
-

even more likely that over-processing is occurring.

In a true Lean process, every step of a process provides value, but it can be tricky to
determine when value is not occurring. Quality is important to both the success of the
process and the end customer, for example, but the cust
realize, that your process is imbued with quality checks throughout. They care instead that a
process takes 10 minutes longer because of those quality checks teams have to dig deep
into processes, metrics, and customer voice to determine if those 10 extra minutes are
providing enough added value to cover the annoyance or loss of customer because of the
added time. This knowledge is all ascertained through data collection and statistical analysis
discussed in detail throughout this book.

Muda of waiting refers to any idle time in a process, whether that idle time is for machinery
or people. In other words, an employee or machine is working below capacity or is not
working at all due to waiting on inputs from another part of the process. Waiting occurs
when steps in the process are not properly coordinated, when processes are unreliable,
when work is batched too large, during rework, and during long changeovers between staff
or machines.

r, another cashier takes

over at the station. To maintain financial integrity, almost all companies switch out cash
register drawers during the change between cashiers; if there is a mistake with the drawer,
the company knows who was running the register at that point in time, making it easier to
find trends or issues. However, when the drawer is being changed out, the register usually

52
 SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

To reduce the muda of waiting in this cash register example, most organizations use a
process that includes a series of register drawers that are prepped at the beginning of the
day or shift. The drawers are designed to be interchangeable in the registers, so the old
drawer can be pulled and the new drawer put in in less than 30 seconds, solving the waste
of waiting.

Waste of waiting is common in construction environments. Construction of roads, bridges,

the roadway of a bridge until the pillars are in place and steady. This concept is seen even
on a small scale in construction, causing paid employees to often stand around waiting for
others to complete a piece of the job before they can finish the assigned task.

Because construction is a field of specialists and certifications, one employee often cannot
do the work another can. In other words, employee tasks are not interchangeable. This
causes additional muda of waiting; one group of construction workers might have to wait
on the forklift operator to come and move some items. Meanwhile, the forklift operator is

to construction: some offices have policies requiring an official IT staff member to handle

need to switch out an underperforming mouse, regular office staff must send a support
ticket or make a phone call and wait for IT staff to solve the problem.

You can eliminate waste of waiting within many processes by balancing machinery, people,
and production. The process will only perform as fast as the slowest link; beefing up the
production of a single element does nothing for the whole, so teams must work to balance
and improve the entire process.

Sometimes, scheduling is a key component in eliminating waiting. In the construction

example, advanced planning and scheduling tactics could reduce the chance that one team
would be waiting for a forklift operator while he or she is busy elsewhere. Understanding
organizational, team, project, and process needs also helps leadership provide the right
amount of resources to reduce waiting. The construction site might benefit from two forklift
operators, for example. In the IT example, a company can reduce wait times by maximizing
IT staffing at high-volume times or implementing processes within the IT department to
create more efficient responses to help tickets. An auxiliary IT staff member can be hired to
handle less technical issues such as switching out an under-performing mouse.

53
 LEAN CONCEPTS

While Toyota originally defined seven common muda, even Taiichi Ohno agreed that other
types of waste exist. In some cases, what seems like other types of waste are just more

seem to fit neatly into one of the seven muda categories.

Talent can be
available. If a process calls for data to be entered, and the staff member slowest on 10-key
is assigned to the task, resources are being wasted. Hiring the wrong person, putting staff in

muda of talent. Wasted talent is more a concern for leadership and human resources than
for process improvement specialists, though Six Sigma experts should be aware that the
way personnel resources are handled can drastically impact the efficiency of a process.

Muda of ideas occurs when the thoughts and ideas of people are discounted, not sought
dership and project teams
can often overlook subject matter experts who have detailed insight into a process and, as
such, could offer first-hand knowledge and ideas. The result could be the design of a
process that works great in theory but falls flat on

One reason waste of ideas is such a concern for organizations is that people themselves

real to contribute, might feel like their

looking silly or ignorant.

Six Sigma Green or Black Belts in charge of projects can facilitate less muda of ideas by
encouraging subject matter experts to contribute and encouraging leaders to seek all ideas
before moving forward with change. Brainstorming tools, which are covered in later
chapters, are valuable for this purpose because they are designed to create a safe haven for
all ideas. By fostering all ideas in a safe environment, teams can foster valuable ideas and
avoid waste of ideas.

54
 SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

the right decision. Muda of capital or cash occurs when leadership decides not to invest in
upgrades or improvements that would create additional cash flow. This type of waste is very
similar to waiting, except the cash itself is waiting, often for a time when leadership feels
safe enough to spend it.

Six Sigma helps reduce muda of cash because statistical analysis helps point leaders to
decisions that involve the least risk or most gain. No business decision is 100 percent
guaranteed, but Six Sigma helps leaders hedge bets by using statistical data in the decision-
that could be used to drive gains in
efficiency, production, and profit.

All muda is waste that fails to add value to a product or process as defined by the customer
or end-user. All muda can also be divided into two overall types, which can help
organizations prioritize waste for project and improvement purposes. Muda can be referred
to as type I or type II.

Type I muda are non-value-added tasks that might actually be essential or required by
circumstances. Inspection of products during a process might be required if the process is

put quality controls in place. The act of the inspection, and the time and expense it adds to
a process --
addresses the cause of the defects within the process.

Sometimes, auxiliary processes within an organization are Type I muda. For example, the
external end

essential to the organization. Instead of removing the muda completely, teams might look
for ways to make essential muda as efficient and cost-effective as possible.

55
 LEAN CONCEPTS

Type II muda are non-value-added tasks that are not essential and can be immediately
removed from a process. For example, if a product is carried to and from several work

rearranging the workflow, teams might be able to reduce the muda by a substantial amount
without making any actual changes to how the product is put together.

5S is a Japanese Lean approach to organizing a workspace, so that by making a process

more effective and efficient, it will become easier to identify and expunge muda. 5S relies
on visual cues and a clean work area to enhance efficiencies, reduce accidents, and
standardize workflows to reduce defects. The method is based on five steps:

Sort (Seiri)
Straighten (Seiton)
Shine (Seiso)
Standardize (Seiketsu)
Sustain (Shitsuke)

During the sort phase, all items or materials in a workspace are reviewed, removing
unneeded items and keeping necessary resources.

Consider the copy room in an office: over the years, supplies, tools, and machinery have
piled up. When going through the room, teams might decide that the stapler and Scotch

necessary for the team, since no one ever performs paper-cutting duties. However,
someone notes that the team in shipping and warehousing has to manually cut pages down
sometimes, so the paper cutter is relocated to that department. The team decides to toss a
bin of miscellaneous loose paper and an old fax machine because a new copy machine
includes fax capability. By eliminating obstacles and unnecessary items, costs, time, and
employee frustration are also removed.

The sort step lets you take inventory of an area, discover unused or wasted resources, and
make room for reorganization. Sort can also be applied with computerized processes.

56
 SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

Once excess is removed from the work area, teams must provide a streamlined and easy-to-
use location for everything necessary to the workspace. During the straighten phase, every
item, tool, or material is given a home. To facilitate ongoing organization, the location of
resources should be labeled clearly. The idea is to create a workspace that anyone could
use: if someone from another area comes to your copy room, it should only take a few
seconds to locate the right size paper. Employees in a factory should be able to move from
station to station, finding equipment and tools with ease. The goal is to provide the visual
controls that allow for common-sense operation. Labeling a shelf for letter-sized paper,
arranging sockets in size order in a drawer labeled sockets, or parking the forklift in a
marked area of factory floor when not in use are all good examples. Labeling the stapler
with a label maker is an example of things going beyond common
label items that most people in the workplace would recognize on site.

The straighten phase also works well in a digital environment, especially when computers or
systems are used by a variety of people. In an office that has a shared workspace policy,
computer desktops might be pushed out by technical resources so that the same programs
are available to everyone. Not only are all programs the same, but the icons are in the same
ch for programs if they move to a new
workstation.

The third phase in the 5S methodology is targeted to keeping the workplace clean and neat.

space by cleaning it, maintaining equipment, and returning items to the proper place after
use. In a computerized environment, the shine phase can be accomplished by naming files
in a manner that makes them easy to locate, keeping folder structures intact, and deleting
or archiving files that are no longer necessary.

Shine can be applied to any environment, physical or digital.

The standardize phase is used to maintain the progress achieved in all previous phases. By
keeping high standards of orderliness in place, the benefits of the 5S methodology can be
long-term. The stress and speed of a daily workday can make it hard to keep up with the 5S
standards. If everyone is committed to working together, the benefits can be ongoing.

57
 LEAN CONCEPTS

5S only works if everyone on the team or within the organization commits to the process.
Employees must follow the rules that are set up for standardizing and sustaining the
organization. Otherwise, the team enters a cycle of cleaning up after a period of failing to
keep up with the standards of 5S.

The overall benefits the 5S method includes:

Reduced risks of accidents and safety issues

Increased compliance with regulations from organizations such as OSHA
A foundation that makes additional improvements easier to implement
Waste is easier to identify and eliminate
Production and quality are generally improved

All of these benefits translate to increased profits and customer satisfaction, which are the
overall goals of the Six Sigma methodology.

Just-in-Time manufacturing, or JIT, is another Lean concept that originated with Toyota.
Originally, JIT took a literal meaning. The goal of JIT manufacturing was to produce an

process that required the output; sometimes, that meant the end customers, and, other
times, the customer was a different employee or process within the organization.

In a JIT processing situation, one machine might produce a part required by another
machine. JIT manufacturing means that the first machine supplies only the amount of parts
that the second machine can process. If the second machine can process one part per

from the sections on muda of inventory and overproduction.

-in-time for the end customer, but

most modern companies do try to come as close as possible. Using predictive analytics,
companies attempt to estimate how many of each product will sell before they produce
those products. In some cases, such as with book publishing, companies run a smaller
number of items first. If those items sell well or sell out, the company orders bigger and
bigger runs of the product.

58
 SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

Some entrepreneurs have applied JIT manufacturing to DVD sales. These individuals and
companies buy the rights to films that have never been released on DVD or Blu-ray before.
They usually go after low-cost films that are likely to have a small cult or niche following.

fashion. However, the companies make a stable profit by selling the DVDs through online
retailers such as Amazon.com and printing the DVDs to order. Modern technology lets this
process occur with minimal expense and waiting.

Today, JIT mentalities are less about the literal idea of providing the product just in time;
rather, it has become a more general concept of Lean manufacturing that helps
organizations eliminate waste in the process.

repeat throughout the rest of this book. While Six Sigma is concerned with improving
processes and reducing defects, eliminating waste and increasing efficiencies goes hand-in-
hand with these goals.

59
In the last chapter, we covered some of the major concepts associated with Lean. In this
. These,
along with the concepts introduced in Chapter 1, are some of the building blocks used in
improvement projects and statistical process control.

The driving goal of Six Sigma is to reduce defects. By reducing defects, teams can increase
productivity, decrease overall costs, increase customer satisfaction, and create maximum
profit. One idea inherent in the Six Sigma methodology is that variance is the root of many
defects.

For example, if an oven heats to exactly 350 degrees in five minutes and stays at that
temperature until it is turned off, it is less likely to burn cookies. If a cook measures each
ingredient exactly, he or she is more likely to turn out cookies that consistently taste good.
Add variation in the process, and consistency is lost. When consistency is lost, defects are

might burn. If the cook puts in a cup of sugar instead of a cup and a half, the cookies might
not be sweet enough. Variation makes for inconsistent quality.

the cook set the oven to 400 degrees all the time and only used half a cup of sugar for each
batch? The process has no variation, and neither do the results. The cookies will always be
bland and burnt.

Six Sigma process improvement teams usually take a two-step approach to improvements.
First, they have to determine if the process is functional. In the cookie example, does the
recipe work at all? Is there even a recipe? Once the team determines there is a workable
recipe, they make improvements to remove the variation that causes outputs to deviate
from the result intended by the recipe.

60
 SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

The statistical measure used by teams to understand variation in a process is known as

standard deviation. Standard deviation is represented in math by the lower case Greek
letter Sigma the you saw in Chapter 1.

Standard deviation measures the distance between data points and the mean of all data. A
large standard deviation means an overall wide spread of points; a smaller standard deviation
means a closely clustered set of points.

The image above provides a graphical representation of deviation. Imagine the vertical axis is
a measure of time and the horizontal axis is a measure of temperature. The center line in each
image represents the mean temperature. You can see that the temperature over time varies
much more in the figure on the right.

Standard deviation is a statistical concept. The formula for standard deviation when dealing
with data of the entire population is:

61
 BASIC SIX SIGMA CONCEPTS

You would use this formula if you have all the data elements of a population and not just a
random sampling of data elements. For example, if you wanted to find out what the
deviation was in the size of pizzas made, you could ask staff to measure each pizza before
serving it. You would have the data for the entire population of pizzas for the day, so you
could use the equation above. However, if you wanted to calculate standard deviation when
you have sample data, you would use the equation from the next section.

Formula Key:

Standard deviation

= mean
tells you to add up the results of all the calculations done for the items listed in the
parentheses

N = the number of data elements for which you calculated standard deviation

X = a place holder for each data element

break it down and run through some exercises on calculating standard deviation manually,
but in practice, you will usually use a statistical software tool to make this calculation
automatically.

deviation of scores on the latest test. The scores from her class of 15 students are:

62
 SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

67, 68, 73, 74, 81, 85, 88, 88, 90, 90, 90, 93, 94, 98, 99

1. Calculate the mean.

To begin the standard deviation calculation, you need to know the mean for the population.
The mean is represented mathematically by the Greek letter mu, or . Mean is calculated
by adding all of the numbers and dividing that sum by the number of items in a data set. In
this case, there are 15 items.

67 + 68 + 73 + 74 + 81 + 85 + 88 + 88 + 90 + 90 + 90 + 93 + 94 + 98 + 99 = 1278

1193/15 = 85.2

2. Subtract the mean and square it.

The formula calls for you to take each number in the data set, subtract the mean from it,
and square the result. The first number is 67, so:

67 85.2 = -18.2

-18.2 * -18.2 = 331.24

If you apply that concept to all 15 numbers, you end up with a list of results:

331.24
295.84
148.84
125.44
17.64
0.04
7.84
7.84
23.04
23.04
23.04
60.84
77.44
163.84
190.44

3. Find the mean of the results.

63
 BASIC SIX SIGMA CONCEPTS

The rest of the formula under the square root sign simply tells you to add up all the
numbers you just calculated and divide by N, where N is the number of items in your data
set. Or, to put it another way, you need to find the mean of the new numbers you just
calculated.

The sum of the numbers above is 1496.4.

1496.4 / 15 = 99.76

This new number, 99.76, is called the variance.

4. Find the square root of the variance.

The standard deviation is the square root of the variance. In this case, the square root of
99.76, which is 9.987.

The standard deviation for the test scores is 9.987.

While statistics based on total population data are always more accurate than those based
y work from sample data more often. It just becomes too
expensive or even impossible to get population data for many elements. Sometimes, the

For example, if you wanted to understand temperature fluctuations in a warehouse, you

might record the temperature at a certain location every ten minutes. After several days,
you have sufficient sample data to analyze.

Other examples of sample data include:

A random sample of reasons for denied medical claims

Measurements for river height taken three times per day for a month

The formula for standard deviation based on sample data is:

64
 SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

Formula Key:

Standard deviation of a sample

x-bar = the mean of the sample

tells you to add up the results of all the calculations done for the items listed in the
parentheses

N = the number of data elements for which you calculated standard deviation

X = a place holder for each data element

Since mu is the mean of population -bar, which

is the average of the data points in your sample. Sigma has been replaced with s, but the
only mathematical difference is that you divide by N-1 instead of N to get the variance as a
way to make up for some of the accuracy lost in using a sampling.

the teacher had were a random sampling from all of her classes. The only difference in the
math would come in the second to last step, where we divide by 14 instead of 15, so:

1496.4 / 14 = 106.885

The square root of 106.885 is 10.338, which would be the standard deviation for the
sample.

65
 BASIC SIX SIGMA CONCEPTS

See for yourself:

Lab techs are measuring the response of bacteria to an ingredient in a potential

treatment. They want to know how long it takes bacteria to show a response. Sample
data for response times in minutes is:

2, 3.5, 2.3, 2, 2.5, 3.1, 2.2, 3.2, 4

Calculate the standard deviation.

once the statistical concepts behind the numbers are understood, statistical analysis
software, such as Excel and Minitab, can be used to accurately crunch numbers. Standard
deviation is automatically calculated in most statistical analysis software programs by
clicking a button after you enter your data sets. The standard deviation is also calculated
automatically by such software programs when you initiate other calculations that require
-depth in the chapters on
using Excel add-ons and Minitab for statistical analysis.

In the meantime, you can quickly calculate standard deviation in Excel using the standard
deviation function. To do so:

1. Enter your data set in a column

66
 SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

2. In a new cell, enter =STDEV()

3. Select the cells with data you want to calculate standard deviation for.

4. Hit enter

Note: The formula in Excel calculates a sample standard deviation using the N-1 math,
which means you can use this formula for samples and not for populations.

Standard deviation gives you an idea of how much variation actually exists in a process
while taking outliers somewhat into account. In the example of the grades from above, the
sample standard deviation indicates that most of the grades are going to fall within 10.33
points on either side of the average.

That tells the teacher that students have a fairly wide performance on her test. If the results
were an average score of 90 with a standard deviation of 3, he or she might assume that
students in class were learning and retaining the knowledge as expected. If the average
score was 64 with a standard deviation of 2, then he or she might assume students in class
were not retaining the knowledge as expected or there was some issue with the test
structure. Both of these situations indicate a small variance in the way students are
performing, which points to the success or problem being tied to the class, the teaching, or
the test.
67
 BASIC SIX SIGMA CONCEPTS

On the other hand, if the average score was 60 with a standard deviation of 30, then some
students were performing very well while others were performing poorly. This might
indicate to the teacher that some students are falling behind. If he or she took samples from
several classes, he or she might investigate and realize that the lowest scores were mostly
from one class, which could indicate that he or she forgot to adequately cover a certain
concept in that class.

Standard deviation alone serves as a pointer for where to investigate within the process for
problems or solutions. Another reason to calculate it is because it is involved in many of the
other statistical processes we cover in later chapters. Standard deviation becomes an
important concept in both analysis and statistical process control and often serves as the
starting point for further Statistical Six Sigma analysis.

The Pareto principle, also called the 80/20 rule, says that 20 percent of the causes lead to
80 percent of the effects. This there is also called the law of the vital few: the vital few
inputs drive the majority of the outputs.

The Pareto principle was first suggested by a management consultant named Joseph Juran.
Juran named the principle for Vilfredo Pareto, an economist in Italy who wrote that 20

common in various circles. Business professionals commonly state that 80 percent of sales
come from 20 percent of customers, and volunteer organizations usually operate with 20
percent of the people doing 80 percent of the work.

The principle is critical to Six Sigma not because causes and effects line up nicely via an
80/20 breakdown, but because it almost universally applies that a few inputs create more
impact than all of the other inputs. Individuals seeking to reduce defects can almost always
identify three to four inputs that, if improved, will create dramatic impact on the outcome.
While resources, costs, and difficulty of improvements also play a role in solution selections,
understanding which inputs or root causes are high on a Pareto chart let project teams
determine where improvements will make the biggest impact to the bottom line.

The Pareto principle is best displayed using a Pareto chart, which is a graphical
representation of data elements usually inputs or causes in a ranked bar chart. Unlike a
regular bar chart, the bars are arranged in order of height, with the highest on the left and
the lowest on the right. Statistical software used to create such charts adds formatting and
other elements automatically, but you can also create a basic Pareto chart in Excel.

68
 SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

To illustrate the
medical field specifically in the process for submitting medical claims. Payers often deny
claims, and they do so for a variety of reasons. When claims are denied, provider offices
have to rework, resubmit, or appeal the denials. Some denial reasons are not appealable,

im
denials. The office gathers data about the denials and creates a Pareto chart so the team
can see where the bulk of the denials are coming from. The data is listed below, followed by
a basic Pareto chart created in Excel.

Reasons for Denying Medical Claims

Reason Count
Duplicate claim 18012
Timely Filing 13245
No beneficiary found 10215
Claim lacks information 4548
Service not covered 2154
Medical necessity 1423
Date of service issue 526

69
 BASIC SIX SIGMA CONCEPTS

From the Pareto chart, you can see that the top three denial reasons account for 80 percent
of the denied claims. An experienced billing team could tell you three things just from
looking at this data:

The office has muda of rework. They are sending a large percentage of claims more
than one time.
The office has an efficiency problem. Almost a fourth of their claims are not making
it to the payer pior to timely filing deadlines.
The office has an insurance verification problem, because a fifth of their claims are
being sent with in

Addressing duplicate claims is important because it reduces rework and could enhance the

the timely filing problem first because timely filing denials are final, which means the office
is losing the revenue associated with all those claims. Filing claims on time is not difficult in
many cases, given the fact that most payers allow months or even a year for claims to be
-
hanging fruit in this manner.

70
 SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

like the one above

in Excel. Use the claims denial data or data of your own to practice making a Pareto chart.

1. Create a column for the data labels. Pareto charts work well when you have
quantifiable causes for a defect or other effect. In the example, the data labels are
the reason for the denial. No matter what type of data you are using, enter it in
order from largest to smallest for Pareto chart purposes.
2. Create a column for count. Enter the total for each cause in that column.
3. Create a column for cumulative count. This column provides a running total. You
can calculate the numbers manually or using Excel. In the data table below, you
would set C3 = B3. In C4, you would enter the formula =C3+B4. You can drag that
formula down and Excel will change the references for each cell so you get =C4+B5,

4. Create a column for percent. In the data table below, the formula for D3 is
=B3/$C$9. Cell C9 has the total of all denials, so we want to divide each individual
denial total by C9. The dollar signs in the formula let you copy it into each lower
cell. The first reference will change, moving to the next line, but the dollar signs tell
Excel to keep the C9 reference for each calculation. The final result is a table that
looks like this:

Reason Count Cumulative Percent

Duplicate claim 18012 18012 35.9%
Timely Filing 13245 31257 26.4%
No beneficiary found 10215 41472 20.4%
Claim lacks information 4548 46020 9.1%
Service not covered 2154 48174 4.3%
Medical necessity 1423 49597 2.8%
Date of service issue 526 50123 1.0%

5. To create the Pareto chart, highlight the information in both the Reason and
Percent column and select Insert Chart Bar chart.

71
 BASIC SIX SIGMA CONCEPTS

6. The bar chart will be created automatically. Select Add Chart Element Trendline,
and add either an exponential or linear trendline.

72
 SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

7. Select Add Chart Element Data Labels, and select the format of data label you
prefer for your chart.

73
 BASIC SIX SIGMA CONCEPTS

Pareto charts are helpful analytical tools when you need to analyze frequencies or causes of
problems. They also help narrow an approach for a problem that has many causes or is too
broad to address in a single improvement project. Like the claims denials example above,
you can find a single cause to work on that can yield large results across the entire process.

Pareto charts are also helpful when communicating information about causes to others,
especially those outside of the Six Sigma process. Although Pareto charts are a powerful
analytical tool, they also represent complex data in a visual format that is familiar to most
anyone. Business professionals know how to read a bar chart, and putting the chart in order
only makes it easier for individuals to see the true causes behind issues. For this reason,
many Six Sigma experts regularly include Pareto charts when presenting to business leaders
74
 SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

and others, especially if the data might be considered surprising or need visual
reinforcement.

Voice of the Customer, or VOC, is a foundational concept in many quality programs. The
goal of quality is to make a better, more consistent product. One of the ways you know

way to reach this goal is to seek feedback from the customer, making VOC data critical to
collect before, during, and after improvement projects.

Successful VOC programs are proactive and constant in their desire for feedback, and
technology makes it possible to seek customer feedback in numerous ways. Some methods
for capturing feedback include:

Surveys via telephone, mail, email, or online

Focus groups in person or online
Interviews
Beta or user testing
Feedback forms
Customer complaints
Social media or site interaction
Reviews
Forums

The VOC can be sought as a means to clarify needs and desires, clarify specific problems
with a process, or as a regular part of improvement, customer service, and marketing
agendas.

Asking the right questions, in the right way, helps you create powerful VOC campaigns that

campaigns in this section: general customer feedback and specific customer feedback.

General customer feedback is often obtained through feedback forms, customer complaint
records, and passive information gathering via websites or social media. Through such
methods, organizations are usually testing general waters to get a temperature reading: are
75
 BASIC SIX SIGMA CONCEPTS

customers happy overall, dissatisfied overall, and is there any direction as to the cause of
customer feelings?

Pick up a feedback form in any fast food restaurant or access the online survey usually
linked to on a rec
VOC campaigns.

Kroger, a grocery store chain in the United States, includes a link to a survey on most of its
receipts. The survey first asks for the date, time, and an entry code from the receipt. This
helps the company know where and when a person shopped so they can attribute feedback
to the right location and staff.

Next, the Kroger survey asks in which areas of the store a person shopped. The rest of the
survey asks specific questions about each area of the store a person visited, including:

What was the overall satisfaction with the store?

o Employee friendliness
o Prices
o Service
o Cleanliness
o Items being available
o Weekly specials
o Ease of movement
o Quality of brands
o Check out times
Whether the shopper is likely to recommend the store to another person in the
next 30 days.

These questions are designed to gauge general customer feelings on critical quality
elements for the store. Understanding your critical to quality factors, or CTQs, is important

General VOC feedback is often used as a smoke alarm. A smoke alarm is designed to alert
individuals in a business, home, or other building that the possibility of a fire exists. Smoke
alarms are set at a sensitive level, so they go off when smoke is present and people within
the building can take action. Often, the alarm and early action saves lives and can even
reduce damage associated with a potential fire.

76
 SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

VOC data can work the same way. If numbers change suddenly in a certain area, an

exist; early investigation and action can help prevent problems from becoming bigger or
more costly. For example, if a certain Kroger store always scored high in cleanliness, and the
numbers dropped consistently across a month, then store management might need to
revisit maintenance and cleaning training.

Results from general VOC feedback are also used in some organizations as an indicator of
quality for certain employees. Sales and services staff are often rewarded financially and in
other ways for high customer satisfaction scores. This also increases employee drive and
satisfaction.

Sometimes, organizations want feedback that is specific to a problem, product, or idea. The
same tools used in general feedback campaigns can be used in specific campaigns, but you
can also tailor the VOC tool to the need. If you want feedback about a new app, you could
use a beta test. If you want to test a product, idea, or marketing campaign, an in-person or
online focus group might be best.

For specific feedback, you have to ask specific questions. This is especially true if you are
seeking additional information or clarification of general feedback. For example, if Kroger
did see a problem with ratings on cleanliness, it might want more information about how
and where customers note uncleanliness. Without additional feedback, managers might
have staff mopping the floors more when customers really felt the store was dirty because
of a lack of lighting or because shelves were stocked in a sloppy manner.

Getting the right type of feedback and keeping costs and timelines within budget
requires selecting the right VOC tool for your project. The table below rates each tool on
relative cost and provides some brief pros and cons.

77
 BASIC SIX SIGMA CONCEPTS

Tool Cost Benefits Disadvantages

Feedback form Low -Gathers a lot of -An individual must

data from many decide to leave
sources feedback, skewing
results to people who
-Can be geared
feel strongly one way
toward numeric
or the other
data for easier
analysis

Survey via High -Can randomly -Requires a lot of labor

phone select, which hours
allows you to draw
-Customers may be
conclusions for the
annoyed by unwanted
entire population
phone calls

Survey via mail Medium -Can randomly -The customer must

select, which send it back for it to
allows you to draw count. Because many
conclusions for the
entire population. you have to send more
surveys to get a
-Lower cost
statistical sampling.
alternative to
phone or in-
person surveys.

Social media Low -Ongoing ability to -Requires an

seek feedback. established social
media following.
-Ability to ask
questions on the
fly.
-Relies on followers and
-Possibly the least fans, which means you
expensive option are asking for feedback
for VOC. from people who

78
 SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

already favor your

brand in some way.

Focus groups in High -Lets moderators -Limits data pool to

person seek more in- local customers or
depth answers or those willing to travel.
feedback
immediately -
assumptions about the
general population.

-Customers may be less

inclined to be honest
when face-to-face with
surveyors

Focus groups Low -Lets moderators -

online seek more in- assumptions about
depth answers or general population.
feedback
immediately

-
travel and you can
access customers
across the globe

User or beta High -Provides feedback -Takes time and

testing about a specific requires experienced
produce, service, users or testers.
or process.

When designing your own VOC tool, keep in mind how you intend to use the information
gained. If you want to input data into statistical analysis software to test hypothesis or

79
 BASIC SIX SIGMA CONCEPTS

create visual charts, then you need to ask questions that yield actual data points that can be
analyzed using statistics. A popular way to do this is with a Likert Scale.

Using a Likert Scale, you would frame all questions so they are answered via a 5-point
ranking. The ranking can be any number of things, but most commonly is some variation of:

Strongly agree
Agree
Neutral
Disagree
Strongly disagree

The answers are coded with numbers when data is entered into statistical software. For
example, an answer of strongly agree might be coded as 10. Agree would be 7, neutral 5,
disagree 3, and strongly disagree 1. By using numerical data, you can easily create charts
and graphs and run more in-depth statistical analysis, which is covered in future chapters.

We introduced some ideas about Six Sigma metrics in Chapter 1 when we talked about
sigma level and defects per million opportunities, or DPMO. Metrics are extremely utilized
when applying Six Sigma to processes and improvements, requiring that anyone working in
a Six Sigma environment be familiar with them.

Many Six Sigma metrics come with an equation, just like standard deviation. For example,
the equation for DPMO is:

(number of defects in a sample/opportunities for a defect in the sample) * 1,000,000

For example, if a mail-order retailer examines quality of the order process, it might sample
forms entered by customer service representatives. If each form has 10 fields, then there
are 10 opportunities for an error on each form. If the retailer reviews 90 forms, then there
are 10 * 90, or 900, total opportunities for errors.

During the review, the retailer finds 2 errors, or defects. To calculate DPMO, the math
would be as follows:

(2/900) * 1,000,000 = 2,222 defects per million opportunities.

80
 SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

DPU is a measure of how many defects there are in relation to the number of units tested.
DPU is concerned with total defects, and one unit could have more than one defect. The
formula for DPU is:

Number of defects found / number of units in the sample

For example, if a publisher printed 1,000 books and pulled out 50 books for quality checks,
it might be looking for the following defects:

Incorrect printing
Incorrect alignment
Missing pages
A loose spine
Torn cover

Out of 50 books, the publisher discovers:

3 books are missing pages

1 book is missing pages and has a torn cover
2 books have loose spines
1 book has incorrect printing and incorrect alignment

There are 9 total errors, as two books had two defects each.

The DPU is calculated by dividing defects by number of units sampled. In this case, 9/50 =
0.18.

DPU provides an average level of quality it tells you how many defects on average each
unit can be expected to have. In this case, that is 0.18 defects on average.

First time yield is the ratio of units produced to units attempted to produce. For example, if
you put 12 cookies in the oven, but only 10
cookies.

The formula for FTY is:

81
 BASIC SIX SIGMA CONCEPTS

Number of good units produced / number of units entering the process

In the cookie example, the FTY is 10/12, or .833.

Most products or services are created via multiple processes; you multiply the FTY for each
process to calculate an overall FTY. For example, consider the following process chain:

100 units enter process A and 95 units exit.

95 units enter process B and 85 good units are achieved.
85 units enter process C and 80 good units exit.

The FTY would be calculated as follows:

95/100 = 0.95 = FTY of process A

85/95 = 0.89 = FTY of process B
80/85 = 0.94 = FTY of process C
0.95 * 0.89 * 0.94 = 0.79

The overall FTY of the process is 0.79.

The rolled throughput yield, or RTY, provides a probability that a unit will be generated by a
process with no defects. One of the main differences between RTY and basic yield or first
time yield is that RTY considers whether rework was needed to generate the number of

rework that is inherent in a process, which means they often measure a process and deem
it successful even if muda is present.

RTY is calculated in a similar manner to FTY, but it takes rework into account. If process A
from the FTY example only achieved a yield of 95 because someone reworked five items to
make them good, then RTY calculations add five instances of rework into the ratio. The
formula is:

(Number of units entering - (scrap + rework))/number of units entering process

In the case of process A: (100 - (5+5))/100 = 90/100 = 0.9

Consider the following process chain:

100 units enter process A. Five are scrapped, 5 are reworked, and a total of 95 are
produced.

82
 SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

95 units enter process B. Ten are scrapped, 5 are reworked, and a total of 85 are
produced.
85 units enter process C. Five are scrapped, 15 are reworked, and a total of 80 are
produced.

The RTY is calculated as follows:

100 (5 + 5) = 90, 90/100 = 0.9 RTY for A

95 (10 + 5) = 80, 80/95 = 0.84 RTY for B
85 (5 + 15) = 65, 65/85 = 0.76 RTY for C
0.9 * 0.84 * 0.76 = 0.574

The overall RTY for the process is 0.574, which is a much lower rate than when you look

RTY indicates that there is waste in the process in the form of rework.

See for yourself

A government agency handles applications for assistance for local families. The process for
each application includes:

-
- A separate staff member reviews the information and uses an income scale to
determine if the family is eligible for any assistance
- The second staff member sends the family a letter stating their options for assistance

All of the applications and customer feedback for March were reviewed, and the team
found the following information:

- 643 families sought assistance in March

- 3 families were not able to complete the application process because the
representative took too long to see them
- 50 applications could not be passed to the second rep because of incomplete
information
- 45 applications did not have complete information at first but that information was
later received
- The second staff member was able to process all completed applications she received
- Of all letters that went out to families, 10 included incorrect information

Calculate the FTY and the RTY for this process.

83
Understanding how to approach the problem beginning to identify the problem and
defining it with a statement is critical to creating a foundation for successful Six Sigma
project,

general, digging deeper via a series of why questions, and creating a general problem
statement as a launching point for a project.

Because Six Sigma approaches things with a statistical mindset, it considers all problems as
a function. Using mathematical symbols, this looks like:

The y=f(x) statement can be used in two ways. First, it is a general map for stating a
problem. Y (the problem) occurs because some X (input or cause) is occurring. In reality, Y is
usually occurring because of some group of causes or inputs, which means there are going
to be more than one X inputs.

The idea can also be applied to specific processes and outcomes within the problem. As you
get more and more granular, the y=f(x) concept becomes increasingly mathematical; in
many cases, you can graph the relationship between the output (y) and the input (x).

84
 SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

that might occur for a large HVAC service provider. The manager of a service team has
discovered that service calls are taking much longer than expected; in fact, his five team
members take 1.75 times longer on average than other service reps in the company to
handle all types of calls.

To find out what might be causing the situation, the manager researches the problem by
talking to the reps, talking to the customers, and going out on random calls with all five
representatives. He makes the following observations:

One representative is a native to the area the team services, which means he or
she knows many of the customers personally. This results in friendly chatter
that lengthens the time on the job.
One representative is providing homeowners with very in-depth explanations
and education about HVAC issues, sometimes over and beyond what the
homeowner would ever need to know regarding his or her HVAC unit.
One representative is new to the job and takes longer to complete each task
because he or she is unsure of the work, has to double-check the work, or calls
another rep to ask questions about the work.
The remaining two reps perform work in times that are on par with company
averages.

The manager distills this data down to two overall causes for the problem:

Too much talking (reps one and two)

Inadequate training

The problem can now be stated as a function:

The extra time is a function of too much talking and inappropriate training.

The manager also now has two root causes to address. The example is simple, but it
illustrates the basic concept in defining a y=f(x) relationship for a problem and its causes.

relationship is always present.

Some other examples of y=f(x) relationships include:

Low customer satisfaction with hamburger taste is a function of an uncalibrated

grill.

85
 Low employee morale is a function of a poor time-off approval system.
Customer wait times are a function of technology distractions for employees.

Data analysis is one of the best ways to validate a y=f(x) assumption, but teams who are
familiar with processes can often arrive at some basic relationships through a process
known as the 5 Whys. This is a brainstorming tool that asks increasingly granular why
questions about a problem or process, seeking to understand the root cause or actual
problem. The 5 Whys can be used to define a problem or to begin seeking causes.

For example, consider the hamburger example above. Teams addressing a problem of
customer satisfaction might begin doing so because feedback forms have shown a lower-
than-normal satisfaction with food quality over the past week.

The team first asks: Why are customers dissatisfied with the food?

Looking at feedback tied to orders, the team notes that the customers who are rating the
food poorly are mostly customers who ordered hamburgers of some type. The answer to
the first question is that the customers are dissatisfied with the food because they are
dissatisfied with the hamburgers.

Why are customers dissatisfied with hamburgers?

The team looks at written feedback on forms or speaks with customers directly and
discovers that many customers feel that their hamburgers were undercooked. The new
answer is that customers are dissatisfied with hamburgers because the meat is
undercooked.

Why is the meat undercooked?

An investigation into the kitchen reveals that the grill is not properly calibrated and is
providing inconsistent results. At this point, you have the y=f(x) relationship, but the team
could keep asking questions.

Why is the grill not properly calibrated?

Further investigation shows that the morning shift, responsible for calibrating the grill, has a
new grill cook. During training, education on performing this function was omitted. The grill
is not properly calibrated because the employee responsible was not properly trained. Now
the team has a specific cause and a solution: train the grill cook.

86
 SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

In a Six Sigma environment, the team might move on with one more question: Why was the
grill cook not properly trained? This might lead to the development of a consistent training

In the hamburger example, it only took four why questions to get to the root of the
problem, and a fifth question started pointing to controls or long-
always this easy; the tool is called the 5 Whys because it often leads to answers within five
questions. However, teams could ask a dozen questions if they begin at a very high level and
work down through a complex process.

One benefit of 5 Whys is that it only costs your team a small amount of time to use a team
familiar with a process can conduct a complete 5 Whys session in less than an hour if a
moderator keeps things on task. Because of its simplicity, the 5 Whys tool can be used for
almost any problem. Use it to address a problem team members bring up, to address a
problem a supervisor noticed, or to address the vague feeling that there is a problem when
no one has been able to define what is actually wrong. At the very least, a 5 Whys session
facilitates communication and thought.

In a Six Sigma project environment, 5 Whys is usually deployed when processes involve
human interactions or people-powered inputs, though it can be an effective start to
brainstorming on any process.

Since a 5 Whys session is usually based on input from subject-matter-experts, gather people
who are close to the process. On a white board or web conference screen, display a basic
problem statement as you understand it. This problem statement is not going to be detailed
e next session a 5 Whys session is often one of the
tools you use to get to that detailed statement.

Examples of statements you might see in a 5 Whys station include:

Customers are not happy with the selection of produce

Customers are receiving orders late
The printing process is resulting in too many defects
Lead times on the bottling process are excessive
Employees are not happy with vacation schedules

87
These are all fairly general statements that simply say something about defects or
dissatisfaction. Begin by asking the highest-

down.

The team works together to provide a high level answer to the question. Employees are not
happy wit
the answer down under the question, then create the highest-
can about the new answer.

answer is that supervisors take so long to approve vacation requests that other employees

everyone.

answer might be that the time-off system is too cumbersome, so supervisors put off
approvals until they have a lot of time to manage them.

hen
moving from screen to screen and each approval requires a vast number of clicks and
entries.

Now, the team has a root cause: the system itself is inefficient, which leads to problems
down the line. If the team can correct the programing issues and encourage supervisors to
approve vacation requests faster, employee morale can be improved.

A Six Sigma improvement project usually starts with a formal project statement. This is
different from the basic statements used to launch a 5 Whys session. A strong problem
statement is similar to a 30-second elevator pitch, which executives and sales people across
the globe use to hook clients or business investors on an idea. The problem statement, like
that pitch, provides enough information that a busy executive can understand what the
issue is and why there is a need for an improvement effort.

Project statements should include:

Where and when the problem was recorded or was occurring

A measurement of magnitude for the problem, preferably with some tie to cost

88
 SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

A brief description of the problem that could be understood by professionals

not closely aligned with the process (avoid too many niche words and acronyms
if you will be presenting information to non-niche professionals)
A brief notation about the metric used to measure or describe the problem

In the first quarter, the California distribution center sent 108,000 packages. Of those
packages, 15,000 were returned, resulting in a 13.8 percent return rate. The rate of return is
above the accepted 7 percent rate and cost the company an additional $372,000 for the
quarter. Over the course of the year, the current process could result in additional costs of
over $1.4 million.

This problem statement covers all the basic information:

When? During the first quarter of this year.

Where? The California distribution center
What? Returns
How many? 15,000, or 6.8 percent above expectations
What is the magnitude? The cost could be $1.4 million a year

strong problem statement because it answers all the basic questions and it provides a
significant reason for leadership to invest interest: $1.4 million a year is a big loss.

The Canton, Ohio bakery is producing undercooked bread. Customer dissatisfaction with the
bread is resulting in returns and bad word of mouth. The bread is supposed to be baked at
350 degrees for 40 minutes.

problem occurred, how it was measured, and what the true magnitude is. The problem
statement also begins going into possible root causes when it includes how the bread

This statement might be better framed as:

In November and December 2014, customer satisfaction complaints were traced back to
bread baked in the Canton, Ohio facility. The facility produced 300,000 loaves during that

89
time period and received 50,000 complaints of bread being undercooked. Bread returns and
loss of sales related to quality are estimated to be $125,000 per month.

expect to get all the information into a couple of sentences. To avoid leaving out
information, it helps to use a list and to consider yourself a problem-statement journalist.

When journalists write a report, they are looking to answer some specific questions: What
happened? Who did it happen to? When did it happen? Why does the audience care?

The same is true when you are writing a problem statement. Follow the problem statement
checklist:

o Where did the problem occur?

o When did the problem occur?
o What process did the problem involve?
o How is the problem measured?
o How much is the problem costing (in money, time, customer satisfaction, or
another critical metric)?

Use the checklist to construct the problem statement, and then ask yourself: Could
someone else answer all the questions in the checklist from your problem statement alone?
Before you present your statement to a boss or other decision-maker, test it out with a
coworker or someone who is not as familiar with the issue as you are.

Here are two problem statements. See if you can answer all of the questions in the checklist
just using the information provided.

The call center in Jacksonville, Florida, handled 36,000 calls in February 2015. Of those calls,
8,000 had an average speed of answer (ASA) over the contract-required 15 seconds. Those
8,000 service-level-agreement violations resulted in costs of $200,000.

The call center in Ohio has a service-level-agreement issue that is costing approximately
$9,000 per day.

90
 SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

Problem statement 1 answers all of the questions on the checklist:

o Where did the problem occur? Jacksonville, Florida

o When did the problem occur? February 2015
o What process did the problem involve? Answering phone calls
o How is the problem measured? Average speed of answer
o How much is the problem costing (in money, time, customer satisfaction, or
another critical metric)? $200,000 per month

Problem statement 2 does not answer all of the questions:

o Where did the problem occur? Ohio

o When did the problem occur? Unknown
o What process did the problem involve? Unknown
o How is the problem measured? Unknown
o How much is the problem costing (in money, time, customer satisfaction, or
another critical metric)? $9,000 per month

Problem statement 2 would benefit from adding a place, a specific reference to a process,
and a specific metric.

Another way to tell you have a strong problem statement is that you can create an overall
project objective statement or goal directly from the problem statement. Consider the two
examples above.

The team working with problem statement 1 might create an objective that states:

The goal is to reduce answer speed SLA violations in the Jacksonville call center by 50
percent within three months. The potential savings to the company is $100,000 per month.

The team working with problem statement 2 would not be able to create a goal statement
with this much detail. They would simply be able to say they hope to reduce the service-
level-agreement violations in the facility.

Specific problem and objective statements are critical to Six Sigma project success for
several reasons. First, being as specific as possible sets up appropriate expectations. In the
first example, leadership has a specific expectation of the project: the team is going to work
to reduce average speed of answer, and success is a reduction of 50 percent. No one is

91
going to expect the team to solve another problem, such as customer satisfaction with
phone operators. That is out of scope for this project.

In the second situation, the problem and goal statements are not specific enough. What SLA
violations is the team addressing? What, exactly, does success look like? Is the team
expected to reduce costs completely? Not being specific enough sets you up for failure.
Leadership might expect you to address service level agreements that have to do with how
reps route phone calls, but you are only intending to address service level agreements that
relate to the speed with which calls are answered. Leadership might think success is a 75
percent reduction in costs when you intend to work toward a 25 percent reduction.

Creating strong problem statements lays a stable foundation for the rest of your project,
gives the team a beacon when they get overwhelmed with information, and reduces the
chance of scope creep and misunderstanding.

Scope is the definition of what is included and what is not included in a process or
improvement project. You begin defining scope with your problem statement. The
information you include in the statement gives clues to what you will be working on, and
the goal statement provides appropriate limits on the work to be done.

Six Sigma projects are not everlasting initiatives, though the culture of improvement that
comes from Six Sigma is. This means your individual project needs a specific, challenging,
but attainable goal. Once that goal is met, the project is concluded and you begin looking
for a new problem to improve upon.

Scope creep occurs when teams look to make infinite perfections on a process, attempt to
reach unrealistic goals, or begin to reach for processes or problems that are out of the
original scope. For example, consider the problem statement from one of our examples in
this chapter:

In the first quarter, the California distribution center sent 108,000 packages. Of those
packages, 15,000 were returned, resulting in a 13.8 percent return rate. The rate of return is
above the accepted 7 percent rate and cost the company an additional $372,000 for the
quarter. Over the course of the year, the current process could result in additional costs of
over $1.4 million.

A related goal statement might be:

92
 SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

The goal is to reduce the return rate to the accepted 7 percent and save the company
$372,000 per quarter.

In scope for this project are processes related to shipping and returns only insomuch that
they impact the return process. At some point, the team might stumble upon a packing
process that is using too much material, thus costing the company an additional $50,000
per month. Unless the packing process is causing the returns which is not likely in this
situation this issue is not in scope for the team and they should not seek to fix it. The team
can, however, note the issue or report it so that a future project might be launched to
address the problem.

It takes discipline and organization to address only that which is in scope for a project.
Understanding the relationship between problems and inputs and knowing how to create a
strong problem statement are the first steps to controlling an improvement process.

93
In Unit 1, we introduced Six Sigma as a concept and covered a lot of principles that
are foundational to creating and maintaining improvement in a business. In Unit 2,
n a process, and
how Six Sigma projects can improve processes. The concepts you learn in Units 1
and 2 become the bricks used to build project work that we discuss throughout the
rest of the book.

A process is a collection of tasks, steps, or activities that are performed, usually in a specific
order, and result in an end product such as a tangible good or the provision of a service. In a
business, multiple processes work together to achieve organizational goals. Technically, the
business or organization itself can be seen as one enormous process. For example, a law
firm that handles criminal defense cases operates via a huge, complex process. Defendants
and their cases enter the process. The output of the process is the result of the case: a
bargain with prosecutors, a win or loss in court, or dismissal of charges prior to court.

Within the huge process that sees the defendant through to his or her outcome, there are
hundreds, possibly thousands, of smaller processes. There are processes within processes. A
lawyer and team of paralegals might move through the process of negotiation; a legal
secretary might go through the process of setting appointments. Holding depositions,
making copies, sending letters, and filing legal documents are all examples of processes. At
the most detailed level, even answering the phone or typing a letter can be considered
processes.

94
 SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

In Chapter 6, we briefly introduced the idea of scope and scope creep. To define and
maintain proper scope, we said that a Six Sigma team had to identify the processes that
were related to a process improvement or project. In the legal firm example, a project to
reduce the time it takes to set appointments would likely not include a process for filing a
legal brief. To know that, howe

components of processes and a format for mapping those components known as a SIPOC.

definition above is just that: Basic. Before we begin defining the components of a process,
ess

Whether physical, digital, or ideological, every process is a series of some number of steps.
You can put those steps on paper in the form of instructions often called a standard
operating procedure in a formal business training or policy document -- or a visual diagram
known as a process map. A process map uses standard shapes and connections to create a
map of a process that can be understood by most employees and any Six Sigma team
member.

Processes all take a certain amount of time, and processing time can change with a variety
of factors. Process maps and documents can only record information such as the average
time a process takes or measures of variation in the processing time. This information is
often noted in such documents because it provides valuable information to teams, but real-
time observation of a process almost always provides better information about processing
time.

A retail chain might create a process map for restocking a certain area. The process
documentation notes an average time of two hours to fully restock the shelves in the
defined area. In an effort to obtain more data about the process, a Six Sigma team observes
employees actually performing job functions in real time at various times of day for two
weeks. Some notes that come from those observations include:

Stocking in the evening takes only minutes.

Stocking during the day is hampered by the movements of customers.

95
 What is a Process?

Stocking work performed during peak shopping hours usually takes the longest.

With just this information, you can probably see an easy way to reduce stocking time in this
example: move stocking duties to non-peak times when possible. Simply understanding the
steps to stock the area is not enough to understand the process; you also have to gather
data about process times.

Almost any process in a business will be dependent upon one or more other processes.
Remember, the business itself is a series of linked processes all working toward the same
goal or goals. Sometimes, interdependencies are noted on processes maps. Other times,
interdependencies are resource-related.

For example, consider a very simple passenger train scenario. The train leaves station A with
passengers, carrying them to station B. Before the train can leave, the engineer must be on
board and prepared to operate the machine. Safety checks, clearance from the rail yard, the
closing of all the doors: these are all processes that must be completed before the train
leaves the station. The process of the train transporting passengers is dependent upon the
completion of other processes.

When working with processes during a Six Sigma improvement project, teams must be
aware of interdependencies. What does any process you are working on rely? What relies
on your process? The first is important because you might need help from processes or
people upstream from your process when making improvements. The second is important
because you have to know how your improvements will impact downstream processes and
people
company or organization as a whole if it hinders the performance of a downstream process.

Processes require resources. Like a motor vehicle requires fuel or electricity to run, a
process requires resources such as power, people, cash, digital bandwidth, computer
equipment, machinery, supplies, parts, and even skill. Since someone in an organization has
96
 SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

to approve and pay for resources, project teams must understand the resources involved,
the cost of those resources, and the owners of the processes and resources in question so
they can make appropriate requests about needing additional resources.

Processes are made up of components that include inputs, outputs, events, tasks
(activities), and decisions. Inputs enter the process when a specific event occurs; tasks and
decisions are performed upon or with the inputs. At the end of the process, an output is
generated. Most of the time, the idea of process components is introduced with a simple
factory-based illustration: raw goods of some type enter the factory, work is performed,
and finished goods leave the factory. For example, if a factory makes hard candy, things
such as sugar, water, plastic, and electrical power enter the factory. Equipment and
employees take the inputs and work with them. The end result is a wrapped piece of candy
ready for the store.

The figure below illustrates the idea of process components using a pizza shop example. An
event the ordering of a certain pizza begins the process. You can see all of the
in detail in the section
that follows.

97
 What is a Process?

Figure 1 Medium Cheese Pizza Process

98
 SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

Input refers to anything that enters a process or is required to enter a process to drive the
creation of an output. In the pizza example above, the inputs are all the ingredients needed
to make the pizza. You might also consider factors such as oven temperature, type of oven,
el to be inputs in the process.

Understanding all inputs to a process is important in Six Sigma because inputs are often
causal or related to causal factors regarding a process. Inputs or the results of those
inputs can cause errors or defects in the process. The cookies burn when the oven is too

information: these are all inputs into processes that are also
causing problems within the processes.

Other reasons for defining inputs when working with a process include:

Understanding the resources required for a process to run

Identifying extraneous i
Understanding costs for the process
Understanding how the process relates to processes that come before it

Remember, in a business, processes are linked together to accomplish a final goal or goals.
The inputs entering process B might be the outputs leaving process A.

The output of a process is the service or product that is used by the customer of the
process. In the pizza example, the output is the cheese pizza the customer is going to eat. In
the candy factory example, the output is the hard candy that will be sold by the retail store.

The process customer is not always the traditional end customer who purchases a product
or service. Customers can be internal or external. An example of a process serving an
internal customer might be seen in a business office that employees a receptionist to
answer phone calls. If the receptionist takes messages or transfers phone calls, he or she is
serving both the person on the phone and the person who is receiving the call or message.

In some cases, the customer of a process is not even a person. Many processes feed other
processes. In a pharmacy, the process of entering data about a prescription feeds the
process that bills an insurance company for the medication.

99
 What is a Process?

From a Six Sigma perspective, an output is almost always of more value to the ultimate
process than the input is. The process itself involves adding value of some kind to the
inputs. A bakery puts raw bread dough through the oven to add value to it: the result is an
edible and tasty product that a consumer is more likely to purchase or to pay more for.

Events are specific, predefined criteria or actions that cause a process to begin working. A
process that performs well responds to an event just like a light bulb responds to the action
of a switch being pulled. Six Sigma teams must determine what events trigger a process
because it helps them understand why a process is being performed and whether the

Consider an example about compliance audits in a financial sector company. Perhaps a

company created a specific audit process that initiates when red flags on accounts are
raised; the audit process is comprehensive and usually takes an average of 80 labor hours.
You can imagine that it is an expensive process to run. In investigating, a Six Sigma team
identifies the event associated with the process: anytime a discrepancy in an account is
noted by a clerk, the compliance process is triggered.

The team investigates and realizes that this is true even when the discrepancy was minor
a few dollars or less or the clerk was able to reconcile the discrepancy later in the day. The
team might suggest that the relationship between the process and the event is a problem.
The process is running at times when it might not be valuable for it to do so.

Tasks, or activities, are the heart of a process. Just as the heart pumps blood through your
body, the tasks within a process pump the inputs through, turning them into the outputs.
Tasks are the physical, automated, or computerized actions within a process. Examples of
tasks include:

A machine joining two metal parts with a weld

A person entering data into a software program
A computer processing data to create a report
An email being written
A piece of computerized work being routed within a workflow system
A chef chopping ingredients for a recipe

100
 SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

Decisions are closely related to tasks and can be tasks themselves. A chef preparing
ingredients for a soup dish has to chop those ingredients, but he also has to decide how
much of each ingredient he needs. His decision will likely be guided by the recipe and the
number of people he has to feed.

Decisions within a process are typically governed by a set of rules. Sometimes those rules
are formally documented; other times, decisions are made via informal rules along with
staff knowledge and experience. Processes that are governed by informal rules can have
problems of consistency; even when all staff are experienced, they could have individual
variations on performing a task. And, as we discussed in Unit 1, variation can lead to more
opportunities for defects and a reduction in quality.

Using the task examples above, here are examples of decisions in a process:

A person entering data into a software program makes a decision to select a certain
drop down category because of training or rules provided by the software
A computer processes a report; the result of that report is a number above a set
threshold, so the computer sends the report to a person
When writing an email, a person chooses to include certain specific information,
such as an order number or customer number, because it is protocol to do so when
sending this type of email

between all the components are equally complex. Inputs can be outputs from previous
processes; outputs can be inputs in the next process. A decision might result in an event
that starts a new process, but it can also be the factor that decides which task begins. As Six
Sigma teams work with processes observing them, diagramming them, and measuring
them the teams begin to understand the relationships of the components, and that helps
them make decisions about possible improvements and changes.

As teams work to improve processes, they need to understand who the process owners are.
Depending on the business organization, process owners can be the people with the power
to approve changes. In some organizations, the lowest-level owner might not have veto or

101
 What is a Process?

decision power about all changes, but he or she is held responsible for the performance of
the process.

A process owner can be:

A person in charge of a very specific process or function

A team supervisor or department manager
An executive-level individual who is probably responsible for a number of
processes in his or her division

The responsibilities of a process owner are often defined by the infrastructure of a specific
business, but commonly, a process owner will:

Monitor how the process performs, usually using one or more metrics or regularly
reported data elements.
Understand how the process fits into the overall business, why the output of the
process is critical to business goals, and what inputs feed the process.
Ensures the process is documented via standard operating procedures (SOPs) and
that process documentation is kept current and accurate.
Ensures operators within the process have the resources and training they need to
complete their jobs.

In a Six Sigma environment, process owners might also ensure a control plan is in place and
regularly review the process for possible improvement opportunities.

Finally, all processes generate some form of data

information is inherent in any process. A computer program that automatically routes work
in a workflow might generate data such as the number of items in work queues, how many
items were worked that day, the time items have been waiting in queues, how many items
were transferred, and where those items were transferred to. A process for filling bottles
with liquid might generate data such as how much liquid is placed in each bottle, how many
bottles per hour are filled, and perhaps variation between bottles.

process is in control and successful.

102
 SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

The SIPOC diagram is often an important part of the define stage of a Six Sigma project. But
you can use the SIPOC diagram anytime you want to learn more about a process or
understand how a process in a business environment is linked to other processes.

SIPOC stands for Suppliers, Inputs, Process, Outputs, and Customer. For the purposes of a
SIPOC, inputs and outputs follow the same guidelines described previously in this chapter.
Suppliers are the people, processes, and organizations that supply inputs to your process.
Customers are the people, processes, and organizations that make use of the outputs of
your process. The process itself is the series of steps that take the inputs and make them
outputs.

SME: Subject Matter Expert

The SIPOC diagram is one of the most often used tools
An SME is someone who is
for understanding process components and process
closely associated with or
relevance because it is so effective and simple. Teams
familiar with a process or work
can create SIPOC diagrams in a single brainstorming
function. Six Sigma teams invite
session, though effective diagramming usually
SMEs to participate in
requires the presence of a process owner and one or
discussion, process mapping
more SMEs who are familiar with the process on a
sessions, or problem and
daily level.
solution brainstorming, because
SIPOC diagrams are also infinitely scalable. Teams can SMEs have valuable insight that
diagram processes at a very minute level, but they can might not be provided by high-
also use SIPOC to diagram an ent level process owners or a review
walk through creating a SIPOC diagram and then of the data.
provide some examples of SIPOCs at various levels to
illustrate scalability.

You can create a SIPOC diagram as an individual exercise or within a team environment.
SIPOCs can be created using a computer and software tool such as Word or Excel, but you
can also draw them freehand on a whiteboard or piece of paper. Freehand diagramming is a
valuable brainstorming tool because teams can quickly edit the rough draft of the diagram
as they discuss a process. Keep this in mind :many of the diagrams presented in this book

103
 What is a Process?

look clean because they have already been typed and edited. As you diagram your own
processes, they can tend to look messy at first with edits, arrows, scratch outs, and inserts.
can limit the ideas and information
that flow during the process. You can always create a clean copy of the diagram for
presentation purposes when you are finished brainstorming.

A SIPOC diagram is based on swim lanes. Swim lanes let you show how cross-functional
activities and resources relate to your process. A SIPOC diagram gets five lanes: one each for

like the figure below.

Suppliers Inputs Process Outputs Customers

Before beginning a SIPOC session, set a definition for where your process or responsibility

ect.

Naming your process helps the team identify more readily with a specific aspect of the
business. For example, a team working to improve processes within a medical office might
process, the
team has put some scope limitations in place: the team will talk about things related to
gathering information from patients. The scope is further limited to the process by which
staff gathers information from new patients.

As you work through the SIPOC diagramming exercise, you can point back to the name and

104
 SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

You can complete SIPOC swim lanes in any order, but best practices usually have teams
enter data in the following order:

Process
Outputs
Customer
Inputs
Suppliers

returning to swim lanes repeatedly to move information around and add new information.

-level or detailed map of the actual process, so keep teams high-

level when completing the process swim lane. You can simply enter the name of the process
in that section, or you can list some of the high level steps required for the process. Listing
beginning to
visualize the process usually helps ideas flow about how the process is connected to other
processes and resources in the company.

To keep the session from turning into a detailed process mapping activity, ask the team to
describe the process in less than five to seven steps. Keep things simple by limiting process
steps to short verb-

Once you have a rudimentary process definition, begin with either inputs or outputs. Ask

go into the outputs swim lane.

answers in the customer swim lane. Remember that customers can be external or internal,
and another process can be the customer in cases of automation.

105
 What is a Process?

You can divide the idea of inputs into two types, if you like. First, you have the actual inputs
the goods and services that are transformed by the process to create the outputs. Second,
you have enablers
for the process to
function. Machinery is an enabler. In a process that cuts metal parts from a steel sheet, the

your SIPOC helps you define the process and provides additional information for later in the
project.

Once you have a list of inputs, ask the team,

or internal. A vendor might provide the raw sugar that goes into the candy in a factory; the
marketing department might provide the leads that the sales department uses to create
orders.

Suppliers can also be other processes, particularly in an automated environment, and you
can have a list of several suppliers for one input in a raw SIPOC diagram. For example,
support tickets come into the Information Technology (IT) department. The supplier of the
ticket could be both the end-user submitting the ticket and the automated process that
routes the ticket to the appropriate work queue. If you are documenting enablers, you
might record the end-user as the supplier and the automated process as the enabler.

Ensure that your understanding of the process at this high level is accurate by validating

about, though. Invite other SMEs or the process

owner to review the diagram briefly with the team and provide feedback.

One of the best ways to create an initial SIPOC diagram during a team session is on large
pieces of paper or a whiteboard. Create swim lanes by drawing them on the whiteboard or
hanging a piece of paper for each swim lane on the wall. Provide the team with sticky notes
and markers; write on sticky notes instead of writing directly on the board or paper. This
lets you move components around quickly as you work through the diagram.

106
 SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

Here are some sample SIPOC diagrams. The first diagram is at the highest level: the process
is the business itself. The second diagram features an automated process. The third diagram
illustrates a people-powered factory process and includes enablers.

This diagram shows the SIPOC for a mid- -level,

simplified SIPOC that shows how customers and vendors provide information and items; the
printing company then turns those inputs into products such as printed business cards. The
final product goes to individuals, businesses, and marketing professionals who placed the
order.

Suppliers Inputs Process Outputs Customers

Paper Orders/customer Receive Business Individuals

vendor specifications order cards

Business
Ink vendor Paper Layout Brochures owners
designs

Copy and Ink Banners Marketing

print Print and signs departments
machine designs
provider Designs
Mailers
Deliver
Customer printed
product Letterhead

107
 What is a Process?

-level SIPOC diagram.

However, if the team includes members from outside the division or company, such as
vendors or consultants assisting with an improvement, then starting with a high-level
diagram can help those outside of the business understand the overall goals of the
company.

The diagram below represents an automated process in a mail-order pharmacy. The process
in question puts labels on bottles that are to be filled with corresponding medications. The
scope of the process is only the labeling of the bottles.

Suppliers Inputs Process Outputs Customers

Bottle sorting Unlabeled Choose Labeled Bottle-filling

machine bottles bottle size bottle station

Label machine Data for Print label

labels

Prescription Affix label

software Labels

Ink and label Ink for

vendors printing

Because this is a process within a chain of automated processes, almost all of the
components are machines, processes, and things. Prior to labeling, a machine sorts bottles
by size. That machine feeds the labeling station as needed. After the labeling is done,
another station fills the bottles.

108
 SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

The SIPOC diagram below illustrates how enablers might be recorded for your process. The
process in question takes place in a factory that makes furniture; in this process, a person
attaches legs to a barstool on an assembly line. For the purposes of this illustration, leg
attachment is the last step in the completion of the product, which means the product
moves from the leg attachment station to packing and shipping.

Suppliers Inputs Process Outputs Customers

Upholstery Stool top Align legs Barstool Packing

station with legs station
(provides final attached
top of stool) Legs Attach legs
with screws

Warehouse Screws
(provides legs, Place
screws, and protective
protective Protective cover
cover) cover

Enablers:

Conveyor machine that

moves products

Drill for application of screws

Without the conveyor machine, the people in involved in this process would have to move

109
 What is a Process?

but it enables the process to move at a more efficient pace. A case could be made that the
screws can be installed manually
the process to move at a speed required for mass production.

With just this simple SIPOC diagram of a process, a Six Sigma team would already have
some idea about where variation could be hiding, what drives efficiencies in the process,
and how the process relates to the overall business.

Whether working in a team or on your own, choose a process you know about and practice
creating your own diagram. Pick a process associated with your business or a business
example you have experience with. Use the following templates to get started.

110
 SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

Suppliers Inputs Process Outputs Customers

Enablers:

111
 What is a Process?

Suppliers Inputs Process Outputs Customers

112
One of the most concise definitions of quality comes from the International Organization for

The same document defines requirements as expectations or needs that are implied,
obligatory, or stated, and the ISO notes that requirements can be generated by different
interest points. A Six Sigma team should be interested in requirements generated by all
interest points, but often focuses most on those generated by the customer. Various types
of requirements might include:

that a customer wants the product he or she ordered. Expectations of delivery

speed might be stated in the form of feedback in customer surveys.
Compliance or regulatory rules, which are obligatory. For example, banks must
protect credit card information they are obligated by rules from government and
-DSS).
Similarly, healthcare organizations must protect the confidentiality and security of
patient data; they are obligated to do so under the Health Insurance Portability and
Accountability Act (HIPAA).
Brand expectations, which come from in-house leadership. Brand expectations are
typically stated; while not obligatory in the sense of being backed by regulation,
companies for which high-quality, a specific voice, or other unique factor is a
component of branding might treat brand expectations as obligatory. Coca-Cola,
for example, has a brand that is recognizable around the globe. While components

regulations and might not be required by customers, Coca-Cola itself holds these
components as important and puts resources and effort into them because it
values its brand.

7
https://www.iso.org/obp/ui/#iso:std:iso:9000:ed-3:v1:en:term:3.9.11
113
 QUALITY

process improvement as well as costs associated with quality in general.

Critical to quality characteristics, or CTQs, are the factors or parameters that are the major
drivers of quality within an organization or process. Usually, CTQs are key characteristics
that can be measured; where the performance of said metric provides information about
whether or not the customer is going to be satisfied.

CTQs are closely related to CTCs, or critical to customer characteristics, but they are not the
same thing. Something can be critical to quality even critical to how a customer ultimately
feels about a service or product without being critical to the customer directly. CTQs are
internal concerns, but they drive CTCs.

s to understand the difference and the

relationship between these two factors.

When a customer purchases a pair of pants, he or she is usually concerned with how the
pants fit and look. Are they comfortable, is the size correct, and does the clothing match the

can take customer feedback on various types of pants and learn that a certain fabric with a
certain cut is most comfortable for the target audience. The manufacturer can also
determine appropriate measurements for each size. During the manufacturing process,
these critical-to-quality factors are applied: only fabric that meets the specifications
identified is used. The fabric is then cut to specific measurements and sewn together in a
specific manner measurements and sewing methods are critical to quality.

each fabric piece or the way the seams were sewn. They want to put on a pair of pants and
experience a comfortable fit.

performed as expected. The feedback suggests that the chocolate is too sweet the taste
and the sweetness of the chocolate is a critical to customer characteristic.

114
 SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

The company might tweak its formula, reducing the amount of sugar that goes into the
chocolate. The recipe and the amount of sugar -- is a critical to quality factor in this case.
But what if the customer feedback indicated that health-
want to buy a chocolate bar with so much sugar in it? Then the amount of sugar in the
recipe becomes both a CTQ and a CTC. It is critical to the quality of the taste of the bar, but
customers might also look at the nutritional information on the bar and make purchasing
decisions based on the amount of sugar in the chocolate.

If a business wants to launch a mobile app for its customers, then an obvious customer-

about the process the business needs to go through to launch the app on the platform in
question, but the business must meet the criteria for Apple, Android, Windows, or other
mobile operating systems. Those requirements become some of the CTQs for the mobile
app development, even though certain requirements from the platforms might not appear
to be at all related to statements from customers about desires or needs.

In a process improvement environment, CTQs are critical to narrowing work scope and
understanding how to enact change. Consider the 80/20 rule discussed in Chapter 5. Often,
CTQs are the factors, characteristics, or outputs that drive 80 percent of customer
satisfaction. By improving these few critical factors, teams can substantially impact
customer satisfaction and the performance of the overall process. Identifying CTQs lets
teams create the most improvement possible with the time, money, and people resources
available.

Outside of a project environment, understanding CTQs lets organizations stay on top of

quality. By managing a few critical metrics, teams can ensure excellent output in a
continuous fashion and identify potential areas for improvement before they become
customer-facing problems.

In Chapter 5, we introduced the concept of the Voice of the Customer, or VoC. Six Sigma
teams usually start with some type of VoC data when they are defining a problem and
working on goals for a project. Either the team conducts surveys to hear from a statistically

115
 QUALITY

relevant group of customers during the first few phases of a project, or the team receives
feedback from internal customers about a process. Sometimes, the VoC information a team
begins with is something as simple as a champion or executive-level individual making a
statement about expectations for an internal process or project. When VoC data is limited
in such a fashion, teams might have to work harder to validate assumptions with data
before moving on to CTQ analysis.

To gain a better understanding of how to measure the quality of a process, teams must
convert VoC statements to CTQs. One of the best ways to do this is through a diagramming
process known as a CTQ tree.

A CTQ tree begins with specific and critical customer needs, breaks that need down into
drivers, and uses the drivers to create requirements. Specific requirements are easier to
convert to measurable quality components. While each CTQ tree is unique, they begin with
a common form. The common structure of a CTQ tree is shown below.

requirements. Some customer needs will have more drivers than others; some drivers will
have more requirements. You might also have multiple CTQ trees t to create
one for every need you identify that is critical to a customer.

116
 SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

Begin the CTQ tree process by creating a list of needs that are critical to the customer. A
bank working on processes dealing with online checking access might identify accessibility,
user-friendly interfaces, and security of information as the major critical-to-customer needs,
for example. Define needs in broad terms to help catch all drivers and requirements later in
the diagramming process.

The best way to define needs is to directly ask customers for feedback, but time and

of data collected via recent surveys or feedback forms, which is the next-best thing. In the
absence of customer feedback, brainstorm critical needs with a group of employees who
has knowledge of and experience with the customer. Subject matter experts from sales,
customer service, and complaint departments can often provide viable information when
the customer is the end-user. You can also begin a CTQ tree with the outputs of your SIPOC
diagram; depending on how you structured the outputs on a SIPOC, you might need to
define critical quality factors for the output as a starting point for your CTQ tree diagram.

Once you have a list of critical needs, work with one need at a time to create a tree similar
to the diagram above. Identify quality drivers that must be present or met for the customer
need to be fulfilled. For example, customers of an HVAC service company might require
excellent service. Drivers for that need might include friendly service technicians, helpful
and knowledgeable employees, and a timely response to service calls.

Drivers are the tra

necessarily have to be able to measure drivers, but you want them to be a bit more detailed
than the broad customer needs you already identified and you want to be flowing in the
direction of measureable factors when possible.

Requirements are the most detailed breakdown regarding critical to quality characteristics.
These are the things that you can measure that lead you to understand whether drivers are
performing appropriately so customer needs are met.

117
 QUALITY

You can see in the above CTQ tree that drivers have been converted to requirements and
each of those requirements can be measured with numbers. In some cases, driver-to-
requirement conversions equate to a hard number. The chocolate bar example at the
beginning of this chapter features such a scenario: if the company deems that the recipe
with a quarter cup of sugar is the correct recipe, then the process metric for quality is
exactly a quarter cup of sugar.

In other examples, companies might provide leeway for exceptions or the understanding
that a process is not going to hit an exact number every single time. For example, an HVAC
team ca
made and if the service technicians can arrive earlier than 24 hours, most people would
want them to. That means the company has to create a definition for the requirement: how
many hours is it before a customer considers a technician untimely? In the case of our

Because teams will use the requirements from the CTQ trees to develop process
measurements and metrics for success, it is extremely important that each requirement is
vetted before teams incorporate it into the project or process. Requirements should be
compared to VoC data, to existing measurements, and to experience and knowledge from

118
 SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

equirements

needs work.

Six Sigma teams must always be aware that quality comes at a cost. When talking about
quality costs, many organizations consider what is known as the cost of poor quality, or
CoPQ. The cost of poor quality is defined as the costs or expenses associated with defects
created by a process. Quality actually has a broader cost avoiding poor quality comes at
an expense as well. The cost of quality, or CoQ, covers the expenses associated with
maintaining good quality throughout an organization or process. Sometimes, this is referred
to as the cost of g
measurements, how they relate to Six Sigma in general, and how they relate to Six Sigma
teams and projects.

In some ways, the cost of poor quality is easier to measure than the cost associated with
overall quality. CoPQ is usually broken into two major categories: costs associated with
external failures and costs associated with internal failures. External and internal failures
are often referred to as the costs of nonconformity they are the expenses that occur when
outputs do not conform to critical to quality requirements.

External failures usually occur after products or services have been delivered, which means
they are directly associated with customer dissatisfaction. External failures might include
revenue losses associated with a reduction in sales because of the quality of products,
services, systems, or information. Other types of external losses include expenses
associated with repairs, returns, or rework associated with a customer complaint; expenses
associated with warranties; or loss of revenue or sales because of customer ill will or bad
word-of-mouth.

requirements set by the company, and the product or service is provided to the customer in
an unsatisfactory fashion. Internal failures are usually handled by scrapping the work,
119
 QUALITY

redoing the work, or repairing the work. Obviously, such rework results in added material
and labor costs, but it also results in losses associated with delays, shortages of parts or
inventory, and lack of flexibility or the ability to adapt. For example, if a process has such
poor quality that 50 percent of the items produced by it require rework, then the process
might be producing 40 percent less on a daily basis than it could be. That means the process
can serve fewer customers, generate less output, and contribute less overall to the

Understanding the cost of poor quality is critical to Six Sigma organizations because it lets
leaders understand how financial needs are related to the need for quality improvements.
The higher the cost of poor quality, the more likely an organization will work toward
improvement.

At a project or process level, the cost of poor quality might help determine budgets for
improvement. If poor quality within the process is costing an organization $5,000 a month,
a project that costs $20,000 but saves $3,000 a month in quality would pay for itself in just
seven months. On the other end, a project that costs $100,000 when the costs of poor
quality are only $1,000 a month is less likely to make sense.

The equation for CoPQ is:

CoPQ = External Failure Costs + Internal Failure Costs

While the equation seems simple, identifying all of the costs associated with poor quality
can be difficult. Most experts use the metaphor of an iceberg to explain the hidden costs of
poor quality. On the surface, you see the very small tip of the iceberg the obvious costs of
poor quality. These might be things such as scrap, reprocessing, warranty claims, customer
returns, and extra shipping.

Beneath the surface, however, an iceberg is always much bigger. The same is usually true of
the cost of poor quality, and hidden costs might include:

Loss of customer loyalty

Loss of morale
Loss of employees if morale remains low for extended periods
Conflicts associated with scheduling or rescheduling
Higher risks of compliance issues, including fines
Higher administrative costs
120
 SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

Unpredictable revenue, sales, or production

Calculating the cost of poor quality is extremely difficult on an enterprise-wide level and still
moderately difficult on a process level. A method for listing all possible costs and

develop a streamlined method that is used throughout the enterprise when calculating
CoPQ. At the very least, Six Sigma experts in the organization might consider defining a
specific way of listing costs of poor quality company-wide so that various process teams are
using similar measures when they report to leadership.

The cost of quality, or CoQ, includes the cost of poor quality and the cost of good quality. In
addition to internal and external failure costs, CoQ includes prevention and appraisal costs.
Prevention and appraisal costs are often referred to as the costs of conformity they are
the expenses related to ensuring outputs conform to critical to quality requirements.

The costs of prevention are the expenses that are related to any activity meant to stop an
error or defect from occurring. Error-proofing, which is covered in detail in later chapters on
controlling processes, results in prevention costs. For example, if a company produces
baked goods, at some point in the process people or machines must measure ingredients to
add to dough batches. One way to error-proof such a process is to provide specialized
machinery that will only allow a specific amount of each ingredient to be introduced to a
batch. Such a machine would likely be quite expensive; it would also have to be managed by
a qualified operator and maintained by appropriate repair and cleaning staff. All of that
activity would generate costs which might be considered preventative in nature.

Other types of prevention costs include expenses related to quality planning, reviews, or
education and training focused on quality. Quality review and evaluation processes also
create prevention costs, whether those reviews are related to suppliers, products,
processes, or people. Customer surveys, the creation of technical manuals, work to create
and manage requirements and specifications, and the management of job descriptions can
all lead to prevention costs. Even housekeeping costs might be considered preventative
costs, especially if a clean work environment is required to reduce flaws or errors in product
manufacturing.

121
 QUALITY

Appraisal costs are those associated with any activity meant to ensure high levels of quality
across a process or organization. If a manufacturing plant hires a quality control specialist,

return the work for correction or report the level of quality as a metric, then the salary of
that person and any expenses related to his or her employment are appraisal costs. In some
cases, those expenses might also be considered prevention costs, but they would not be
counted twice when calculating CoQ.

Other types of appraisal costs might include expenses related to quality audits on products,
services, or processes, the cost of calibration and measurement equipment or software, and
the costs of field tests. Prototype inspections, consulting expenses, financial reporting and
auditing, security checks, safety checks, supplier certifications, employee surveys, and
customer feedback are all further examples of areas where appraisal costs might exist.

The equation for CoQ is:

CoQ = CoPQ + Prevention Costs + Appraisal Costs

The same challenges inherent in calculating CoPQ also exist when calculating CoQ. The
same iceberg analogy is relevant, and prevention and appraisal activities often have hidden
costs such as unnecessary overtime, paperwork, or system expenses.

Traditional wisdom might say that if the cost of poor quality goes down, the cost of good
quality is likely to go up. You have to spend money on quality to have good quality, in other
words. While historically that might be true for many organizations, it is not the case in a Six
Sigma company. Because Six Sigma works to create quality that is inherent in the process
meaning things are done right the first time and defects are reduced the costs of quality
often go down while quality itself goes up.

In Chapter 1, we showed that a process with a higher sigma level (and thus, higher quality)
has fewer defects. Defects decrease in an exponential manner as sigma level rises. Because
there are fewer defects, the costs of poor quality are exponentially reduced as well. But
time and again, Six Sigma has also reduced the cost of overall quality. As the sigma level of
processes is increased via the application of Six Sigma tools and methodology, the cost of
both prevention and appraisal goes down as well.
122
 SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

One way of relating the cost of quality and perhaps the most common way of doing so
among corporations is as a percent of sales. The cost of quality as a percent of sales
typically aligns so closely with sigma values that you can predict the cost of quality based on
in relation to sigma
values are shown in the table below. As you can see, as companies improve their sigma
levels, they experience a substantial savings in the cost of quality.

Sigma level Cost of Quality as a Percent of Sales

2 Above 40%

3 25 to 40 %

4 15 to 25%

5 5 to 15%

6 Less than 1%

Six Sigma is one of the best methodologies for managing the cost of quality because it
works to build quality into every process. When approaching an organization or process that
has a high cost of quality, teams and leadership can apply a triage-based method to reduce
those costs. While no organization can remove quality costs 100 percent, the goal should be
zero costs of failure, either internal or external, and minimal preventative and appraisal
costs.

expensive to detect costs associated with nonconformance, and improvements that seek to
correct causes for a few critical failures can have a big impact on overall quality and total
cost. Instead of adding layers of quality programs over processes to ensure quality which
only adds prevention or appraisal activities and increases the cost of quality Six Sigma
improvement projects build preventative measures into processes themselves. In other
words, efficient Six Sigma processes are self-regulating. They have built in checks and
balances that work to constantly reduce defects and rework.

The benefit of building failure stop-points into a process include:

123
 QUALITY

Earlier detection when errors do occur, which keeps hidden costs down. When
mer, you avoid many
issues and costs associated with low morale, reduced customer loyalty, or
product returns.
Employees are able to support and manage higher quality. When quality is
something employees have ownership of, they are more likely to work hard to
create the best possible output. In contrast, appraisal-style quality programs
can spark feelings of paranoia at being closely watched or create a relationship
in employee minds between the idea of quality and the idea of reprisal or
correction. Poor performance on a continuous basis does need to be addressed,
but employees should not default to a negative mindset when they hear the
word quality.
In-process quality assurance is actually more effective than post-process or
over-process prevention and appraisal methods. Statistical process control and
Six Sigma improvements can push a process to six sigma level. To ensure the
same level of performance 3.4 defects per million opportunities quality
assurance employees would need to review millions of parts and ensure only a

organization.

After teams use the Six Sigma and Lean process management methods discussed
throughout this book to reduce failure costs, teams can turn to prevention and appraisal
costs. Often, in a process that is functioning at a high sigma level, prevention and appraisal
activities are a form of muda. They can be expunged from the process without impacting
the end product, quality, customer satisfaction, or employee morale. In some cases,
removing prevention or appraisal from processes actually creates a positive impact on
quality, production, and customer and employee satisfaction.

The first step to removing quality-related muda is identifying it. Process maps, spaghetti
diagrams, and value stream maps are valuable tools for uncovering
need to be included in a process. All of these tools are covered in depth in Unit 9.

owners or a Six Sigma team, asking questions about value and necessity can also help
identify muda of prevention and appraisal. If you think an activity related to a process might
be an unnecessary form of prevention or appraisal, ask:

124
 SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

Does the activity itself add any value to the output?

Does the activity substantially reduce the time it takes for the process to
produce an output?
Does the activity substantially increase the cost of the process?
If the activity is designed to prevent defects within the process, can the activity
be made more efficient?
If the activity is designed to prevent defects, can the activity be made less
expensive?
If an activity is designed to capture quality data about the process for reporting
purposes, are those reports necessary?
If quality reports are necessary either because of obligatory requirements such
as compliance or because the reports provide value in another process can
the reports be automated to reduce associated expense?

The answers to these questions help teams identify areas where muda can be removed or
where quality-related processes can be improved.

When Six Sigma teams are expunging quality-related costs and unnecessary activities from

programs with some other form of control. By the end of this book, you will have learned
about a number of tools, from statistical process control to poka yokes, which you can use
to create quality within a process.

or even an entire department devoted

to Six Sigma
leadership must buy into the belief that quality is better when controls are incorporated
within the process, and they have to be willing to communicate this fact via training and
example. Some companies choose instead to use Six Sigma to improve processes while
maintaining the expense of traditional quality, compliance, and audit departments. For
some industries, such as healthcare or finance, audits and other appraisal and prevention

125
bottom line and/or customer satisfaction. Since Six Sigma works best when it is
implemented as a company-wide culture, project selection should work as an enterprise-
wide function. This chapter covers a number of tools and methods for brainstorming and
selecting projects that are most likely to bring significant improvement to processes and
serve overall business goals.

A critical part of Six Sigma success for organizations is knowing when teams reach maximum
project load. Even when organizations hire employees dedicated to process improvement,
they can only sustain a certain number of improvement projects without substantially
reducing the positive outcomes of those projects. While project work, including data
gathering and analysis, might be handled by employees committed 100 percent to
improvement projects, teams usually have to engage with and pull resources from regular
staff members. An organization that juggles too many projects puts daily output at risk. In
seeking to improve processes, a company that selects too many projects at one time could
actually negatively impact quality.

No formula exists for how many Six Sigma projects a company should run at a given time,
but a few well-designed projects are more likely to make greater impact than many poorly
designed, overlapping, or unfunded projects. Organizations should only launch projects they
can:

Fund. Six Sigma projects take monetary resources, which means organizations
must prioritize based on financial criteria.
Support with people resources. Six Sigma projects require work from

that draw heavily on IT resources or attempt multiple, simultaneous projects

that need input from the Director of Compliance on a regular basis. Relying too
heavily on resources for multiple projects can burn out employees, decrease

126
 SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

morale, impact quality, and impede work that is necessary to keep the business
running from day to day.
Manage. Project teams require leadership; Six Sigma teams are usually run by
Black Belts, sometimes along with certified Project Managers. Since Black Belts
are supported by Green Belts who handle much of the data collection and
analysis work, a single Black Belt can usually manage more than one project at
a time if needed. This is especially true for experienced Black Belts who are not
responsible for any type of daily operation. Even so, organizations with limited

strain on those resources.

When companies are working to apply Six Sigma culture to the entire enterprise, executive
leaders and other decision makers should work directly with Six Sigma experts to identify
improvement opportunities and launch projects. Doing so lets leadership align project
selection with organizational goals, ensure projects are organized in a way that matches

five-step procedure for identifying viable Six Sigma improvement projects.

1. Data-Based Review of Current State of the Organization

Organizations can begin with a high-level look at internal and external sources of
information about performance. Internal information might include complaints or issues
raised by employees, existing performance metrics or reports, financial reports, and quality
reports. External sources include all of the Voice of the Customer tools we covered in
Chapter 5. In reviewing internal and external information, organizations should ask:

What types of things are customers or employees complaining about?

Where is the organization falling short of benchmarks or competitor
performance?
What needs do customers have that the organization is not meeting?
What needs might customers have in the near future that the organization is
not yet able to meet?
What processes are outputting the most defects?
What processes are known for the most rework?
What are the slowest or most expensive processes in the organization?

127
 SELECTING THE RIGHT PROJECTS

What are some obstacles keeping the organization from reaching its goals?

2. Brainstorm and Describe Potential Projects

Answers to the questions in step one become a brainstorming list for potential projects.
What types of things are customers complaining about? Perhaps surveys and feedback
forms show customers complaining about long shipment times, poor quality of products, or
rude customer service. With just a single question, an organization has a list of possible
projects:

A project to reduce the time it takes for customers to receive orders

A project to increase the quality of products
A project to create better customer service

Admittedly, the scope is enormous with these examples, so organizations would need to
look for a bit more detail. Why do customers think the quality of products is low? In Chapter
6, we covered the 5 Whys brainstorming method, and that method is relevant here. During

questions to get a more granular look at project possibilities.

For example, if a feedback form for a carpet installation company indicates that customers

customers are unhappy because carpet edges are coming up shortly after the carpet is
installed. Why is this happening? The short answer is that something is wrong in the

list of possible projects.

Creat
brainstorming session. As issues are raised, more information might be required to list
possible projects, but if you gather the right group of people for a few braining storming

Remember, the point of this exercise is to call out possible areas for improvement, not
validate assumptions or come up with solutions.

Once teams have a large list of possible projects, they should begin creating short
descriptions that will become the basis of step three. The descriptions also let teams quickly
identify things that are not actually problems or would not apply within an improvement
project environment. Descriptions should include answers to three questions:
128
 SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

How is the issue painful to the customer, the employees, or the organization? In
short, how does the issue impede someone from getting what they want or
need?
What is the goal that would be accomplished with an improvement?
Why should an organization address this issue now?

something similar to, though slightly less formal than, the problem statements discussed in
Chapter 6. Using the carpet installation example, for example, the description might be:

Customers are not satisfied with carpet installation because edges are coming up within a
few weeks of installation. The poor edges are creating safety and aesthetic issues and
increasing expense and rework for teams who have to return to sites to address defects.
The goal is to reduce the number of times carpet edges come up by 80 percent. The
organization should address the issue because it is costing $20,000 per month in errors.

Some basic idea of what the team wants to do is provided, and leadership has a very real
measurement of why the improvement is important. Even better, the measurement --
$20,000 in additional costs each month can be compared to other project opportunities.

3. Apply Some Basic Criteria to Shorten the List

Once a list of possible projects is created, teams can apply some very basic criteria to
remove projects that are inappropriate, would not work with Six Sigma methodology, are
not property scoped, or have little likely return on investment. This step usually begins
during the second part of step two, when teams are creating short descriptions of possible
projects.

First, teams can remove items from the list where there is no real pain point. If a significant

improve. For example, if a single employee complaint about the efficiency of a piece of
software made it onto the initial list, a company wouldn
turned out no one else was having the issue.

Second, teams can remove issues that have very obvious problems and/or solutions.
Consider the carpet installation problem: if the issue of edges cropped up in the last month
and someone on the team reviewing potential problems recently received an email about
defects in edging materials from a carpet vendor, the solution might be obvious. Perhaps
the vendor sent notification that the materials in a certain batch of carpet were faulty and
provided instructions for a solution. In this case, action is required on the part of the
129
 SELECTING THE RIGHT PROJECTS

about

need to spend time going through the DMAIC phases.

4. Create Unique Business Criteria

teams should create and

apply business criteria to further filter the list. Business criteria usually come in the form of
expenses, monetary gains, impact on customer satisfaction, and urgency. Some questions
teams might ask include:

How will the improvement impact revenue-facing measurements such as profit,

orders, or income?
What savings will the improvement create?
How is the problem trending? Is it becoming a bigger and more urgent issue
quickly, or can the organization operate with minimal impact without making an
immediate or near-future change?
How much will the improvement cost?
How many employees/employee hours will be required for the improvement?
What resources are required for the improvement?

5. Use Business Criteria to Prioritize Project Lists

Using the business criteria, teams should prioritize projects and select projects from the top
of the prioritized list for immediate work. One of the best ways to prioritize projects is to
create a selection matrix with defined criteria and a numerical ranking system.

For example, using the example questions in step four, we might create the following list of
criteria:

Potential savings
Potential cost
Potential increase to revenue
Ability to access resources needed

A matrix can be created using the criteria and a list of projects. Teams can then rate each
project against each criteria using a numeric scale. In the example below, we applied a scale
from 1 to 10, with 1 being the most negative and 10 being the most positive.

130
 SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

-world
numbers. For example, when rating savings, a higher number just means a more positive
expectation. In this case, the positive expectation would be a high amount of savings. When
rating costs, however, the higher number (and more positive expectation) would relate to a
project with a lower overall cost.

Revenue Access to
Savings Costs
increase resources Total
Project
1 8 9 10
1 28
Project
5 5 4 6
2 20
Project
10 2 2 2
3 16
In the table above, you can see that Project 1 has low expected savings, but the team also
estimates it will have a low overall cost, drive a high increase in revenue, and has easy
access to resources. Project 3, on the other hand, has a high expected savings, but negative
ratings in all other categories.

To rank projects, add up the scores for all categories and order the projects from highest to
lowest by total score.

Teams can choose to create their own criteria for a project selection matrix, or they can use
a 15-point viability model as defined below. One benefit of the project viability model is
that it provides some weighting, letting teams make some criteria more important than
others. It also removes some of the objective nature of the selection matrix defined in the
previous section.

This model is based on 15 criteria, which are defined in the table below.

Criteria Definition

1. Sponsorship The project is likely to be sponsored at a high

level. (For more information on project
sponsorship, see the team building information
in Chapter 10). Sponsorship increases the chance
that teams will have access to the funds and
resources required for a successful potential
131
 SELECTING THE RIGHT PROJECTS

project.

2. Corporate alignment The goals of the project are aligned with the
goals of the business. Working on potential

can reduce business effectiveness.

3. Data Data is available or can be accessed so the team

can design project metrics. Without access to

If data is excessively time-consuming or

expensive to collect, then the potential project is
usually not the best choice.

4. Definition of defect There is a specific, well-defined defect or

problem. Without a well-defined defect,
potential projects run the risk of scope creep.

5. Stability The potential process is stable and there are no

expectations that the process is going to be
overhauled, redesigned, or changed in the near
future. There is usually no reason to spend time
and money improving a process that will
drastically change soon anyway.

6. Customer The planned goal of the potential project would

create a substantial and positive impact on
customer satisfaction or perception of quality.

7. Benefits The potential project has a strong cost-benefit

ratio.

8. Timeline The timeline for a potential project is relatively

short. Timelines for most Six Sigma improvement
projects are around 6 months, though some do
run longer. Longer timelines decrease the chance
that an improvement fits within the DMAIC
methodology.

132
 SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

9. Solution The potential project purpose is to find a solution

that is not already known or defined. As we
previously stated, if a solution is obvious, you

10. Implementation is A solution identified and verified by the potential

likely project is likely to be implemented. If, for any
reason, change is very unlikely within a process,
then going through Six Sigma improvement work
is a waste of resources.

11. Required investment The potential project requires a large investment

of cash. Generally, the greater the cash or capital
investment required, the less likely a project will
be selected or a solution will be implemented
due to cost-benefit analysis.

12. Available Six Sigma The Black and Green Belts required for the
Resources project are available.

13. Inputs can be For a Six Sigma process improvement project to

controlled be successful, at least some of the inputs must
be within control of the team or organization.

quality of a part that is provided wholly by a

vendor.

14. Redesign The process can be improved as is and

need a complete redesign.

15. Process quality is

improved/maintained quality of service or products along the value
chain.

133
 SELECTING THE RIGHT PROJECTS

Based on the above criteria, teams create a matrix.

Most
Most ly
Weig No ly No Possibl Yes Yes
ht (1) (2) y (3) (4) (5)
Is there a sponsor or champion?
Do project goals align with corporate goals?
Is data available or accessible?
Are defects well defined?
Is the process stable?
Are there customer benefits to the project?
Are there company benefits to the project?
Can the project be completed within 6
months?
Is the solution unknown?
Is it likely a discovered solution will be
implemented?
Would a new solution cost little to no cash?
Are Six Sigma team members available for
the project?
Can inputs in the process be controlled?
Can the process be improved without a full
redesign?
Will the improvements maintain or improve
quality across the value chain?

134
 SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

Teams then apply a numerical weight to each criterion. Weight each criterion on a scale of 1
to 5, with 1 being least important and 5 being most important. For example, our team with
the carpet installation issue might create weights as follows:

Weight
Is there a sponsor or champion? 3
Do project goals align with corporate goals? 4
Is data available or accessible? 3
Are defects well defined? 3
Is the process stable? 1
Are there customer benefits to the project? 5
Are there company benefits to the project? 5
Can the project be completed within 6 months? 3
Is the solution unknown? 4
Is it likely a discovered solution will be implemented? 3
Would a new solution cost little to no cash? 5
Are Six Sigma team members available for the project? 3
Can inputs in the process be controlled? 5
Can the process be improved without a full redesign? 2
Will the improvements maintain or improve quality across the
value chain? 5

Next, teams should answer each question by marking a 1 in the relevant box on the grid;
the answers correspond with no, mostly no, possibly, mostly yes, and yes. The complete
grid for our carpet installation problem is featured below.

135
 SELECTING THE RIGHT PROJECTS

Most
Most ly
Weig No ly No Possibl Yes Yes
ht (1) (2) y (3) (4) (5)
Is there a sponsor or champion? 3 1
Do project goals align with corporate goals? 4 1
Is data available or accessible? 3 1
Are defects well defined? 3 1
Is the process stable? 1 1
Are there customer benefits to the project? 5 1
Are there company benefits to the project? 5 1
Can the project be completed within 6
months? 3 1
Is the solution unknown? 4 1
Is it likely a discovered solution will be
implemented? 3 1
Would a new solution cost little to no cash? 5 1
Are Six Sigma team members available for
the project? 3 1
Can inputs in the process be controlled? 5 1
Can the process be improved without a full
redesign?
2 1
Will the improvements maintain or improve
quality across the value chain?

5 1

Once a matrix is completed for each project, teams must calculate and compare the score
for potential projects. These calculations are completed via the following steps.

1. Divide each weight by 3; a weight of 3 equals 1, but a weight of 5 equals 5/3, or 1.7

2. Convert each of the 1s listed on your grid to a weighted value by multiplying it by the
converted weight from step one. For example, the weight for the first question on the grid

136
 SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

above is 3. We divided 3/3 to get 1. We would multiple 1 * 1 for the first row. The next row
is weighted 4; 4/3 is 1.3. The numbers have all been converted in the grid below.

Mos
Most tly
Weig No ly No Yes Yes
ht (1) (2) Possibly (3) (4) (5)
Is there a sponsor or champion? 3 1
Do project goals align with corporate
goals? 4 1.3
Is data available or accessible? 3 1
Are defects well defined? 3 1
Is the process stable? 1 0.3
Are there customer benefits to the
project? 5 1.7
Are there company benefits to the
project? 5 1.7
Can the project be completed within 6
months? 3 1
Is the solution unknown? 4 1.3
Is it likely a discovered solution will be
implemented? 3 1
Would a new solution cost little to no
cash? 5 1.7
Are Six Sigma team members available
for the project? 3 1.3
Can inputs in the process be controlled? 5 1.7
Can the process be improved without a
full redesign?

2 0.4
Will the improvements maintain or
improve quality across the value chain?

5 1.7

3. Sum the numbers in each of the five columns.

137
 SELECTING THE RIGHT PROJECTS

Mos
Most tly
Weig No ly No Yes Yes
ht (1) (2) Possibly (3) (4) (5)
Is there a sponsor or champion? 3 1
Do project goals align with corporate
goals? 4 1.3
Is data available or accessible? 3 1
Are defects well defined? 3 1
Is the process stable? 1 0.3
Are there customer benefits to the
project? 5 1.7
Are there company benefits to the
project? 5 1.7
Can the project be completed within 6
months? 3 1
Is the solution unknown? 4 1.3
Is it likely a discovered solution will be
implemented? 3 1
Would a new solution cost little to no
cash? 5 1.7
Are Six Sigma team members available
for the project? 3 1.3
Can inputs in the process be controlled? 5 1.7
Can the process be improved without a
full redesign? 2 0.4
Will the improvements maintain or
improve quality across the value chain? 5 1.7
1.3 4.4 4.7 3.3 4.4

4. Multiply each of the summed weighted scores by the number at the top of the column.
answers is 1.3. Multiplying that by 1
equals 1.3. The other columns are calculated as:

2 = 8.8
4.7 * 3 = 14.1
3.3 * 4 = 13.2

138
 SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

4.4 * 5 = 22

5. Add up the answers from the previous step. In this case, the total is 59.4.

6. Divide the sum from step five by the sum of the weighted totals from step three. In this
case, 59.4 / 18.1 = 3.28

7. The answer from step 6 is the score for your project.

Once you score each potential project, you can determine if it is a viable project within a
DMAIC methodology with the following key:

Score DMAIC Viability

< 2.0 Not viable for DMAIC

2.0 to 3.0 Possibility viable, but organizations

should validate further

Above 3.0 A viable DMAIC project

It should be noted that the 15-point matrix described above can only be used to determine
if a project is viable within a DMAIC structure. A process might still need to be improved

structure might let Six Sigma teams approach the improvement. The differences between
DMAIC and DMADV methodologies, and how to determine which method is best for a
project, are covered more in-depth in Chapter 11.

The goal of a Six Sigma team is not to define appropriate projects at an enterprise level. A
department or team responsible for only a few processes might be seeking to make an
improvement. In an organization where Six Sigma is important to business culture,
departmental leaders are likely familiar with some Six Sigma tools and might even be Green

139
 SELECTING THE RIGHT PROJECTS

Belts or Black Belts themselves. While these leaders have daily responsibilities that are not
Six Sigma related, they can bring Six Sigma thought processes to their department.

Departmental leaders might want to identify potential opportunities to present to

leadership. They might also want to identify areas where they and their teams can work
toward improvement themselves. In some organizations, department leaders can run
smaller versions of projects with the guidance of on-staff Six Sigma experts especially
when such projects would require little in the way of capital or resources.

Departmental staff can use all of the tools in this chapter to identify possible projects.
Often, though, they are close enough to the situation that they can identify possibilities for
improvement without going through brainstorming stages. If data is already present,
departmental staff might use Pareto charts to identify some areas where improvement
would create results; they can then use the selections matrix to validate those assumptions
and prioritize efforts.

140
 SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

Consider a problem or need for improvement in your own company or one you faced in a
past work experience. Practice completing the project viability matrix using the template
below.

Weigh No Most Possibl Most Yes

t (1) ly No y (3) ly (5)
(2) Yes
(4)
Is there a sponsor or champion?
Do project goals align with corporate
goals?
Is data available or accessible?
Are defects well defined?
Is the process stable?
Are there customer benefits to the project?
Are there company benefits to the project?
Can the project be completed within 6
months?
Is the solution unknown?
Is it likely a discovered solution will be
implemented?
Would a new solution cost little to no
cash?
Are Six Sigma team members available for
the project?
Can inputs in the process be controlled?
Can the process be improved without a full
redesign?
Will the improvements maintain or
improve quality across the value chain?
TOTAL
S:
Score:

141
Six Sigma is typically managed on two levels within an organization. First, the culture of Six
Sigma must be managed at an enterprise-wide level, usually by a group or council of senior
managers, such as executives, with the guidance of a Master Black Belt or Black Belt.
Ultimately, this group sets the tone for Six Sigma within an organization, provides final
approval on projects, and holds others accountable for metrics, performance, and success.
While many of these individuals might also work as sponsors or champions on projects, as a
-to-day project details.

Some roles of a high-level Six Sigma leadership group include:

Creating a rationale for the use of Six Sigma in the organization and supporting
process improvement as a cultural goal.
Setting clear objectives for Six Sigma initiatives to ensure that project goals
align with business goals.
Holding Six Sigma teams and the organization accountable for improvements
and performance.
Demanding and reviewing measurements of results
Communicating wins and losses to the team in an honest manner.
Rewarding teams and individuals for Six Sigma successes.
Advocating for resources and funding for necessary improvement projects.

Six Sigma must also be managed at the team level, which is the primary focus of this

team, and talk about managing a team with timelines and schedules, milestones, budgets,
and a defined measure of success.

-made team ready to begin work on every project that comes

up. Six Sigma teams must be uniquely tailored to the goals and processes at hand. The same

142
 SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

Six Sigma experts Black Belt leaders, data analysts, or project managers might work
across multiple projects, but individual subject matter experts and team members only
bring high value to the team if they are familiar with the process or have some related
education, knowledge, or skill to offer. Not all team members will serve consistently
throughout the entire life of a project, either. This is often why companies send existing
employees for Six Sigma training rather than hire Six Sigma experts.

Executive leadership groups working with Six Sigma leaders and experts usually put teams
together. Any process improvement team should have, at minimum:

A Six Sigma leader

A process owner
An expert on the process
Someone to manage budgeting and accounting

Some of those roles might be handled by the same person; the process owner might also be
the process expert. Depending on expectations of needs, the team might also need to
include technical resources, such as a programmer or IT leader, as well as individuals from
human resources, compliance, legal, or other ancillary departments.

When putting teams together, organizations should remember that three basic team
member types exist with relation to a Six Sigma project. First, there are the regular team
members. These individuals participate in all activities of the team and attend all or almost
all of the team s meetings. Regular team members include project leaders, process owners
and experts, and identified subject matter experts who the team or executives feel would
be critical components of their group.

Second, ad hoc team members provide expertise on an as-needed basis. Usually, these are

take these employees from their job functions for every single team event, as that would
negatively impact the state of current production. Instead, these employees are included in
team meetings as needed when additional information or assistance is required.

Finally, resource team members are only included when the project team leader feels they
are needed in a meeting or team event to provide expert information, counsel, or help in
accessing resources. Resource team members are usually members of ancillary
departments such as accounting, human resources, or compliance. Resource team
members might also be managers or leaders in departments that are related to the process
143
 BASIC SIX SIGMA TEAM MANAGEMENT

being improved. For example, if a team is seeking to improve a customer service

department, they might need help with inputs from the marketing department; someone
from the marketing department could be added as a resource team member.

Most Six Sigma process improvement teams are relatively small: five regular team members
is considered a good number on average. Adding too many regular team members can
create communication problems, make it difficult to manage brainstorming sessions, and

that team members are serving on multiple projects. While ad hoc or resource team
members can serve several projects and handle their own work on a daily basis, regular
team members should not be asked to serve on more than one team and handle daily
workloads. In fact, organizational leaders might want to consider reducing work
requirements for team members who are serving as full-time members on a project.

Other tips for selecting team members include:

Choosing employees who are knowledgeable about the customer, product, or

process related to the project.
Choosing employees who have shown a willingness and ability to work toward
improvement in a team environment.
Selecting employees who have access to and an understanding of the data
required to learn about and measure the process or problem.
Picking employees who can provide at least five hours of work per week to the
team.
Matching the skills of employees to the projects at hand; if a project is likely to
include all technical improvements, you would be less likely to add a team
member who is skilled in marketing.
Removing political obstacles through team selection; if a specific person in an
organization is likely to be an obstacle to a team, sometimes putting that
person on the team can increase the chance that they will buy into the process.

The team member roles described in this section are based on Six Sigma process
improvement best practices, but best practices also say that teams and team leaders should
not be overly rigid. Experienced Six Sigma leaders and experts understand how to work

144
 SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

within best practices while also creating unique team structures that are tailored to the
project or process at hand.

in previous chapters. In most Six Sigma

environments, these are the senior-level leaders who oversee projects at the highest level.
Even the Black Belt must report to the project sponsor or champion. The senior leader is
usually responsible for the final result of a project, which means he or she usually wants
regular reports about progress; sometimes, the sponsor or champion is the liaison between
the team and the leadership council. As the senior leader, the champion or sponsor is also
responsible for assisting the team with obtaining funds and resources to ensure project
success. Some additional duties within this role include:

Coaching the team, particularly at the project charter stage. The sponsor often
provides input into what is in scope on a project and who might be included on
a team.
Locating resources for the team, including support from other departments,
money, equipment, time, and labor hours.
Handling matters of politics within a corporate
have to.
Working with other managers within the organization to help the team
succeed in improving a process and transitioning improvements to a daily work
environment.

The business or process owner is usually someone who is directly responsible for the
process in a leadership capacity. Usually, the process owner is the person who is going to
is ready to be
rolled out to all team members or used on a daily basis. Because of this, the process owner
is usually included in the team because he or she must understand how and why any
change is made. The process owner must also be familiar with methods of control that are
created by the Six Sigma team as he or she will become responsible for maintaining and
monitoring those controls once the process is transitioned from a team environment to day-
to-day production.

A process owner usually also acts as a process expert on a Six Sigma team. The process
owner has insight into the existing process, understands the needs of the customers and

145
 BASIC SIX SIGMA TEAM MANAGEMENT

employees related to the process, and might already have access to data regarding the
always the only process expert on a team, however; in
-to-day interaction
with the process to be an expert.

When leading or managing a Six Sigma team, Black Belts and others do have to be wary of
process owners who are resistant to change or who believe they have all the answers.
Someone who is set in his or her ways might not want to involve other team members or
ders who
are also process owners might be afraid that a team member will outshine them or threaten
their position, which could lead them to block team members from participating on a team.
These are some of the political and human resource problems Six Sigma leaders run into,
and Black Belts and project leaders must work tactfully with champions, sponsors, and
process managers to resolve such issues.

Six Sigma projects are usually led by certified Black Belts, although some organizations do
allow Green Belts to act as leaders on small initiatives with occasional feedback and
guidance from Black Belts. In most organizations, the Black Belt holds primary responsibility
for the regular work performed by a team and usually only works with one team or project
at a time.

Best case scenarios let organizations align Black Belts with projects in areas they are already
familiar with. For example, a bank might have several Black Belts on staff. Each Black Belt
might specialize in working with certain processes or departments; one might usually work
with compliance and audit processes, another with accounting, a third with customer-facing
processes, and a fourth with online processes. Since Black Belt resources might be limited,

improvements
managers or other individuals are certified as Black Belts and can lead processes in addition
to their regular responsibilities, although this can put an undue burden on employees and

Black Belt project leaders often work to:

Help create a rationale for a project.

Provide input for the selection of project team members.
Lead teams throughout all the phases of DMAIC, which are covered in depth in
Unit 3.
146
 SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

Educate and support team members as they learn about and use Six Sigma
tools.
Provide oversight through time management, decision making, and planning.
Maintain schedules and timelines, sometimes in conjunction with a certified
Project Manager.
Provide expertise in the form of statistical analysis or guidance with analysis.
Assist with project transition.
Report to sponsor or champion on a regular basis.
Provide documentation at the end of the project.

In some organizations, Master Black Belts play an overall role in leading multiple Six Sigma
projects. Master Black Belts act as coaches to multiple teams; Black Belts leading Six Sigma
teams can work with Master Black Belts to solve especially difficult problems or seek help
for complex statistical analysis. Master Black Belts provide continuing education to both
Black and Green Belts, helping team members to constantly improve their grasp of Six
Sigma methodologies.

Some organizations use traditional project management techniques alongside Six Sigma
improvement methodologies. In these organizations, a project manager is usually assigned
to a Six Sigma project. While structures vary by organization, the project manager does not
usually lead the team. Instead, the PM offers leader support to the Black Belt by keeping up
with documentation and timelines, helping keep meetings on track, and ensuring items are
followed up on after meetings. At first, you might think that adding a PM to a team would
cause problems for a Black Belt, but when the two roles work together, the Black Belt
benefits. With a PM worrying about timelines or whether the meeting is getting too far off
track, a Six Sigma exert is free to concentrate on the brainstorming session or statistical
analysis at hand.

Not all Six Sigma teams use timekeepers, but they can help keep meetings on track, reduce
the chance of scope creep, and increase overall productivity. The timekeeper can be any
person on the team who is not regularly engaged in leading meetings, brainstorming

such a rigid fashion that the benefits of fluid discussion and brainstorming are lost, but he or
she should gently steer teams toward following agenda schedules or provide the project
leader with an indication that time is up for the topic at hand.
147
 BASIC SIX SIGMA TEAM MANAGEMENT

To function properly, a timekeeper needs an agenda to follow. It is usually the

responsibility of the Black Belt or project manager to provide a detailed agenda for each
meeting. The agenda should include clear indications regarding how long each item is
expected to take, though teams should always be aware that agendas might be changed
during the meeting at the discretion of the project manager or project leader.

Team leaders should pick a timekeeper who is organized and level-headed. In the heat of
and
the timekeeper is a member of the team. In addition to regular duties as a team member,
the timekeeper is expected to:

Keep an eye on the agenda and the time

Let team members know when the time for a certain agenda is almost up;
teams might want to set up a five-minute warning rule so they have a few
minutes to wrap up a discussion
Signal that the time is up for a certain discussion or item

While project leaders can choose to ignore agendas, they should also back up the

heckled by other team members for noting the time.

A lot of discussion occurs in the midst of Six Sigma brainstorming and team sessions, and
someone needs to record that information. Notes are important because they help team
members review what was discussed, create lists of follow-ups and actions from a
discussion, and record charts, graphs, and diagrams that were created during brainstorming
processes. While everyone can take notes, the team leader should appoint one person as
the official scribe for the team. Sometimes, that person is a certified project manager
working in conjunction with a Six Sigma team leader. Other times, it is a member of the
team who is seen as detailed and organized.

The Black Belt or other project leader should never be the scribe; it is too difficult to take
notes while leading a discussion or exercise. The Black Belt might make some notes during
the discussion, but he or she is likely to miss important details while working directly with
other team members.

The scribe should create notes or minutes of the meeting in typed format and disseminate
those notes to all team members as soon as possible following a meeting. Team members
can review the notes and add any missing information if desired; often, organizations create
148
 SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

portals or shared file systems so teams can keep notes and all other documents in an easy-
to-access location.

One challenge in recording the discussions of a Six Sigma project meeting is in recording the
diagrams and brainstorming that occurred. This is especially true if teams use whiteboards,
paper, or sticky notes to create diagrams; the scribe is not always equipped with the skills or
the software to recreate a computerized version of such documents. One tip for recording
such information that is used by many modern Six Sigma teams is to take a picture of the
diagrams with a smartphone or digital camera. The images can then be uploaded into the
Green Belt can convert the raw
diagrams to a computerized version for the purpose of presenting information to leadership
or other departments if desired.

In the beginning of this chapter, we covered the three major types of team members:
regular, ad hoc, and resource. Selecting members for each of these roles is up to the project
leader, the sponsor or champion, and the overall organizational leadership team. In addition
to the project leader, process owner, and process expert, Six Sigma teams are usually
comprised of one to three other regular team members. In addition to acting as timekeeper
or scribe as directed by the team leader, team members also:

Participate in brainstorming sessions, discussions, and other team activities.

Collect data and perform analysis under the direction of the Black Belt. Often, team
members performing these functions are Green Belts.
Perform work between meetings as required by the project leader.
Report the results of and progress on individual assignments to the team.

Review work performed by other team members and the team as a whole, offering
suggestions and feedback.

Scheduling and maintaining that schedule is an integral part of the Six Sigma project
process. Organizational leaders need to understand how long a project will take, when
results can be expected, and when team resources will be freed up for other endeavors.
Without this informa

149
 BASIC SIX SIGMA TEAM MANAGEMENT

two methods for creating a project timeline or schedule and touch on the importance of
milestones.

Six Sigma projects

concept of the DMAIC method. DMAIC breaks a project up into five phases: Define,
Measure, Analyze, Improve, and Control. Experienced Six Sigma experts with some data and
information about a project and process can usually provide a very basic and raw estimate
of time by assigning a certain number
most of the phases are likely to overlap.

To create a raw timeline for a project, a Black Belt or other Six Sigma leader usually starts
with an overall time requirement. He or she either estimates the total time required for an
improvement or works with a deadline imposed by the leadership group. For example, the
leadership group might say that an improvement needs to be completed within four
months.

Using a four month timeline and what information is already available about the process,
problem, and resources, the Black Belt might create an estimated timeline for the DMAIC
process that looks something like the figure below.

The estimated timeline is for 16 weeks; the expert believes the Define phase will take 3
weeks and the Measure phase will take 5. The Measure phase overlaps with both the Define
and Analyze phases, which is normal with Six Sigma projects.

The benefit of this approach is that you can generate a timeline quickly. The disadvantages
are that someone without experience of Six Sigma and a fair amount of knowledge of the
process being improved can easily misjudge the time required for each phase and
leadership might consider this a hard timeline, which can create unrealistic expectations.

150
 SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

When presenting such a timeline, make sure everyone knows that it is a rough estimate and
the time for each phase can change as you go through the process.

The critical path method is a more detailed way of defining timelines for various elements
of a project, but it does require more information and input from a project team. This
line until the project is
underway; a critical path diagram could be one of the activities the team undertakes as part
of the Define phase.

A critical path diagram can be created for the entire project or for each phase of a project.

a project to reduce bad debt (uncollected invoices) in a medical billing environment as an

example.

1. Identify the critical needs or activities to complete the project or phase of a project.

To complete the define phase of our project to improve bad debt in a medical billing setting,
the team needs to choose a team, charter the project, define the problem, and create a
baseline metric.

2. Put critical activities in order.

The order with which the team should accomplish the tasks defined in step one is:

Choose a team
Charter the project and define the problem (these tasks can be done
simultaneously)
Create a baseline metric

3. Assign a time to each task.

A Six Sigma expert estimates it will take one week to choose a team, one week to create a
charter, one day to create a problem statement, and two weeks to create a baseline metric.

4. Create a diagram of the tasks, stacking simultaneous or parallel process and including
time figures.

151
 BASIC SIX SIGMA TEAM MANAGEMENT

The diagram is created from left to right. The items on the left must be done before the
items to the right can be completed. When items can be done at the same time, they are
stacked.

5. Draw a critical path through the diagram.

When steps are stacked, the critical path goes through the step with the longest time
estimate. For example, the team might create a problem statement while they are working
on a project charter; the project charter takes from Monday through Friday to complete.
The problem statement is complete on Tuesday. However, the team is not done with all of
the steps in that series until they are done with the project charter.

6. Add up the longest times from each section.

152
 SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

In this case, the team adds 1 week, 1 week, and 2 weeks to get to 4 weeks total for the
Define phase.

timelines for extremely complex projects or processes.

Once a timeline is established, set up milestone meetings and dates to help keep the team
on track and notify the sponsor or champion of progress. In a DMAIC project, milestones are
usually set at the end of each phase (Define, Measure, Analyze, Improve, and Control).
However, teams can set custom milestones, and sponsors might require specific milestones
if they are approving large resourcing or funding requests for a project.

You can also set up milestones within a team environment to manage goals and tasks; these
milestones can be kept within the team. For example, a team working for a chain of
sandwich shops is hoping to improve the process by which sandwiches are put together.
They have set up the following milestones:

Define: January 21

Measure: February 12

Analyze: February 22

Improve: March 15

Control: April 10

The milestone dates are when the team or the Black Belt will meet with the sponsor to
present the findings or results of each phase of the project. Each date gives the team
something to work toward. However, the team has determined that certain tasks must be
accomplished during the Measure phase. First, they have to create some definitions of
terms so everyone is on the same page when discussing measurements. Second, the team
has to gather data about the temperature at which ingredients are stored and cooked.
Finally, the team wants to actively observe sandwich shop employees in order to measure
the time it takes to make various sandwiches.

The team might set up internal milestones for the Measure phase, stating that definitions
will be created by January 25, temperature data collected by February 5, and time data
collected by February 10.

153
 BASIC SIX SIGMA TEAM MANAGEMENT

By breaking each phase, and each larger task, into smaller parts, it is easier for the team to
stay on track and complete work. Smaller tasks seem more manageable, so they are more
likely to be accomplished.

Teams, and especially team leaders, must always be concerned with project budgets. While
success is rated by end customers in terms of performance, quality, and satisfaction, Six
Sigma teams also answer to corporate leadership. For leaders, success is also measured in
terms of time and budget. A strong timeline and good milestones help you meet time
requirements, and an understanding of financial drivers, strong communication, and
financial oversight help you keep a project within budget.

One of the challenges when dealing with budgets in a Six Sigma project is that all team
members are not always completely aware of financial drivers. In some cases, financial

employee pay made public to various team members, for example. Some information and
analysis might need to be performed solely by a project-leading Black Belt in such cases,
especially if data is critical or sensitive.

Outside of concerns with sensitive data, process improvement projects work best when all
team members are made aware of as many of the drivers and data as possible. When teams
know how much funding a champion is willing to seek on their behalf, they can make
realistic decisions about how to improve a process. Sometimes, the solution that is most

$80,000 capital investment in machinery, for example.

Budget concerns vary by organization. In some organizations, leaders are most concerned
with specific expenditures by a team, including expenses on new equipment, hiring new
personnel, or purchasing new products or software. Some organizations take a more
granular approach to project budgets, considering the expense of hours spent by the team
on the project as well as the expenses associated with training and implementing a solution
outside of the team environment.

Six Sigma team leaders must ensure they understand how leaders and organizations
manage project budgets. Working for the first time with an organization or sponsor means
having honest and thorough conversations about how budgets are calculated, how much

154
 SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

sponsors are willing to work for

expected role in maintaining budgets is.

Finally, Six Sigma teams must create a well-defined measure of success. To best manage a
Six Sigma project and team, leaders have to ensure all team members, leaders, and

-defined,
teams risk concluding a project without satisfying the customer, sponsor, or all members of
ike, the team could

failure.

In the end, successful Six Sigma team management hinges on many of the same concepts as
successful leadership in other endeavors does. Choosing the right people, being clear about
expectations, approaching work in an organized manner, and being honest and open about
progress helps every member of the team succeed and provide value.

155
One of the things that sets Six Sigma apart from some other quality improvement and
management methodologies is a structured approach to every project. Projects that are
meant to improve an existing process follow a road-map for success known as the DMAIC
process; DMAIC is broken into five phases: Define, Measure, Analyze, Improve, and Control.
The main activities of a DMAIC project include identifying the critical inputs or causes (the
Xs) that are creating the problem (the Y), verifying those causes, brainstorming and
selecting solutions, implementing solutions, and creating a control plan to ensure the
improved state is maintained.

The DMAIC methodology is designed to be fairly inclusive the vast majority of teams who
are seeking to improve a project will be able to fit their activities in to the DMAIC steps
because those steps are designed to allow some flexibility. Sometimes, though, teams

improvement for the organization. Instead, a process might need to be completely replaced
or redesigned to meet goals for customer satisfaction or organizational improvement. In
such cases, teams can employ the DMADV method.

DMADV stands for Define, Measure, Analyze, Design, and Verify. The principles governing
the method are very similar to those governing DMAIC, but the last two phases are geared
toward rolling out and testing a completely new process. Six Sigma teams might approach
improvements through DMADV if:

The business wants to launch a new service or product.

Business leaders decide to replace a process because of upgrade needs or to align
business processes, machinery, or employees with future goals.
A Six Sigma team discovers that improving a process is not likely to provide the
success desired from a project.

Most teams go into the project knowing whether they are employing DMAIC or DMADV
approaches, but some DMAIC projects can become DMADV projects usually during the

156
 SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

Define, Measure, or Analyze stages when the team realizes the need for a complete
process replacement. Switching to the new methodology during the middle of the project
might require some shuffling of resources and could impact project schedules, which means
keeping champions and sponsors informed of team progress and decisions is imperative.

formally use the DMADV approach for any

project, in part because they find it easier to stick with the nomenclature of a single
methodology. These organizations might still complete process redesign projects by altering
some of the activities handled in the Improve and Control stages of DMAIC. In short, the
teams do use the DMADV approach, but they use the verbiage associated with DMAIC to
streamline Six Sigma education across all levels of
the organization.
What is Change Management?

Change management refers to a

The major differences between DMAIC and DMADV closely-managed process of
are the goals the team sets and the outcome of the making changes in an
completed project. In some ways, a DMADV project organization. Often, companies
might feel like it has a more tangible outcome, but in use change management policies
reality, both methods are seeking to deliver better and rules to govern how changes
quality, better efficiency, more production, more are made to software,
profits, higher customer satisfaction, or some infrastructure, or processes that
have compliance or audit
the phases of both methods in this chapter. But first, elements.
e quick definitions of each phase as
During change management,
they relate to DMAIC versus DMADV.
teams must document all activity
in keeping with corporate
policies and report changes and
During a DMAIC project, the Define phase is results to an oversight
concerned with identifying the problem, defining committee. Sometimes, Six
requirements for the project, and setting goals for Sigma projects involve changes
success. Requirements and goal setting might relate that are also governed by these
to a variety of factors and are dependent somewhat policies, which means Six Sigma
on guidance from the leadership team and expected leaders must be prepared to
budgets, and Six Sigma leaders can use various tools report to change management
within the phase to create flexibility that allows for a committees.
variety of project types.

157
 INTRODUCTION TO DMAIC AND DMADV

In a DMADV project, the Define stage is slightly more rigid. Teams also have to identify a
problem and begin defining requirements, but requirements must be made within a
change-management environment. Sometimes, organizations have a change management
program in place, which means Six Sigma teams must incorporate all requirements of that
program into the DMADV phases. The team also works to define customer requirements to
create a measuring stick to which the process development can be compared.

The DMAIC Measure phase is when teams use data to validate their assumptions about the
process and the problem. Validation of assumptions also merges into the analyze phase.
The bulk of the measure phase is occupied with actually gathering data and formatting it in
a way that can be analyzed. Measuring can be one of the most difficult tasks in a Six Sigma

create queries for digital data, sift through enormous amounts of data to find relevant
information, or capture data by hand in some manual process.

After validating assumptions from the Define stage with actual data, the team might revisit
problem statements, goals, and other process-related definitions. If the team leaves Define
wi
also work during Measure phases to measure key inputs and steps in the process in
preparation for Define.

Teams working through a DMADV approach might do some of the same things during the
Measure phase, but activities are typically more targeted. Teams will likely collect data and
measurements that help them define performance requirements for the new process.

During the Analyze phase of a DMAIC project, teams develop hypotheses about causal
relationships between inputs and outputs and between Xs and Ys, they narrow causation
down to the vital few (using methods such as the Pareto principle), and they use statistical
analysis and data to validate the hypothese
Analyze phase tends to flow into the Improve phase in a DMAIC project; hypothesis testing
to validate assumptions and possible solutions might begin in Analyze and continue into the
Improve phase.

158
 SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

A team using DMADV might also identify cause and effect relationships, but they are usually
more concerned with identifying best practices and benchmarks by which to measure and
design the new process. Teams might also begin process design work by identifying value-
and non value-added activities, locating areas where bottlenecks or errors are likely, and
refining requirements to better meet the needs and goals of the project.

Six Sigma teams start developing the ideas that began in the Analyze phase during the
Improve phase of a project. They use statistics and real-world observation to test
hypotheses and solutions. Hypothesis testing actually begins in the analyze phase, but is
continued during the improve phase as teams select solutions and begin to implement
them. Teams also work to standardize solutions in preparation for rolling improved
processes to daily production and non-team employees. Teams also start measuring results
and lay the foundation for controls that will be built in the last phase.

The fourth phase is where DMADV projects begin to diverge substantially from DMAIC
projects. The team actually works to design a new process, which does involve some of the
solutions testing mentioned above, but also involves mapping, workflow principles, and
actively building new infrastructures. That might mean putting new equipment in place,
hiring and training new employees, or developing new software tools. Teams also start to
implement the new systems and processes during the fourth phase.

For DMAIC and DMADV teams, the control or verify phase is where loose ends are tied and
the project is transitioned to a daily work environment. Controls and standards are
established so that improvements can be maintained, but the responsibility for those
improvements is transitioned to the process owner. During the transition, the Six Sigma
team might work with the process owner and his or her department to troubleshoot any
problems with the improvement.

Consider the following improvement projects. Which methodology do you think a Six Sigma
team might use to approach each project?

1. A business wants to create a smartphone app to help customers make and manage
appointments.

159
 INTRODUCTION TO DMAIC AND DMADV

2.
to get appointments, appointment communications are confusing, or patients show
up
3.
small size boxes.

The team handling an improvement for the business in example 1 would choose a DMADV
approach
need that already exists and is improving an overall process the setting of appointments
the app itself is a new process and a new product. The app will need to be designed,
integrated into existing systems, and the final product tested before full implementation.

possible that the team might determine during the process that one solution might be to
develop a new appointment-setting software or replace existing software with something
from a different vendor. In some cases, that might warrant a switch to DMADV, but, as
previously stated, not all organizations would do so. Some organizations would continue
with the DMAIC process and modify the activities in each phase to fit the needs of the
project at hand.

Example 3 is a classic example of what brings many teams to the DMAIC method. The
ation knows that goals and expectations
are not being met. A leadership team might work with subject matter experts and one or
more Six Sigma experts to discover more about the processes involved before settling on
one or more improvement projects.

During the Define phase of a Six Sigma process improvement project, teams create what is
known as a project charter and a basic plan for work. A charter is a synopsis of the project.
It provides some common information and a summary of what the team hopes to
accomplish. The charter usually features a list of team members, names of those
responsible for outcomes, a problem statement, a goal, and some basic definitions of scope
and metrics for success. Some charters also include a rough timeline estimate for the
project.

Also during the Define phase, teams create or list measurable customer requirements and
create high-level documents about the process (including process maps). Often, teams will
start with a SIPOC diagram to help them begin to understand a process. Teams should also

160
 SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

identify stakeholders during the Define phase. Stakeholders are individuals, both within and
without an organization, who have some level of influence on the success of an
improvement project. By understanding who stakeholders are, teams can remain in contact
with various persons throughout the project, communicating with those stakeholders as
needed to ensure future viability of any improvement that is created. One way to identify
stakeholders is through a Stakeholder Analysis.

One of the biggest challenges Six Sigma teams face when in the Define phase of a project is
generating positive, targeted momentum that sets the foundation for the rest of the
project. As a Six Sigma team leader, you can increase chances of success by keeping the
team as focused as possible during the Define stage. Begin by explaining the Six Sigma
process and the purpose of the project for any ancillary team members who may not be
familiar with Six Sigma and DMAIC. Next, work as a team to create ground rules for how the
project will run including how meetings are organized and managed, how information will
be communicated, and what each team member might be responsible for during the
project.

Create a charter and project plan so the team has something to focus on. If possible, have
the Champion of the project spend time with the team. Hearing directly from an executive
leader about expectations and the support of leadership for the project helps motivate a

Define is also a good time to explain the roles of scribe and time keeper and talk about the
purpose of brainstorming. Some Six Sigma leaders like to let team members take turns
facilitating various exercises for the group, as this integrates each person more tightly
within the process and helps team members at all levels learn more about Six Sigma.

Once a team has a good grasp of what the process does and how it works, what the
problem is, and what the goal for the project is, the team moves from Define to Measure.
Usually, the transition between phases is marked by a tollgate review wherein the team
presents its Define work to a champion or a Six Sigma leadership board. The champion or
board provides feedback and makes the decision about whether the team is ready to move
on to Measure.

161
 INTRODUCTION TO DMAIC AND DMADV

During the Measure phase, the team is concerned with creating a baseline metric for the
process and refining problem statements and other outputs of the Define stage. Creating a
baseline metric lets teams understand how a process should be measured and how the
process is really performing before improvements begin. It also provides a comparison point

DMAIC method.

One of the biggest challenges, especially for teams and team members who are new to the
Six Sigma method, can be deciding what to measure. Many times, inexperienced teams end

process. Because the Measure phase starts with some educated guesswork and trial-and-
error, teams and Six Sigma leaders have to keep a close eye on progress and redirect work
when measurements are not creating the answers or production required.

A successful Measure phase requires strong observation skills, an understanding of the

reasons behind measure, knowledge of data types such as discrete and continuous, tools for
measurement assessment, and a strong background in statistical analysis. Some of the tools
often deployed in the Measure phase, such as the CTQ tree and sigma level calculation,
were covered in previous chapters.

The Measure phase is often the most challenging phase for a Six Sigma team leader,
especially when working with teams that are inexperienced in the methodology. When
teams start to really dig deep into a process and begin to measure things, they often get a
true idea about how challenging the problem really is. They might also have a difficult time
understanding how and when to meas
collected before can be time consuming and tedious. Because of all these challenges, teams
might enter what is called a storming stage team members question the viability of the
project, rail against the Champion or the team leader, complain how much time the project
is taking from other duties, or stop showing up to meetings altogether.

Six Sigma leaders can reduce the impact of storming on a team by demonstrating a calm
approach to each aspect of the project and redirecting the strong emotion of storming to
more productive work. If you can identify an easy task or problem, letting the team work on
that and accomplish something immediately can reduce the excitement of storming; Six
Sigma leaders should also ensure work is fairly distributed and that each team member
knows exactly what his or her responsibilities are.

162
 SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

Once measurements are collected or are in the process of being collected Six Sigma
teams usually move on to the Analyze phase. Again, a tollgate review is often conducted
between phases, but the lines between Measure and Analyze are often blurrier than the
lines between Define and Measure. In some cases, a team has to measure, analyze, and
then measure some more

Analyze phases are when teams perform detective work on the process. Using the clues
gathered during the Define and Measure phases, along with information provided by the
sponsor, process owner, and subject matter experts, teams attempt to identify root causes
for a problem; they also use statistical analysis and other tools to verify causes before
turning to the work of identifying possible solutions. During the Analyze phase, teams use a
variety of tools some of which were introduced in earlier chapters. Tools common in the
Analyze phase include Pareto charts, run charts, histograms, cause-and-effect diagrams,
scatter diagrams, process maps, and value analysis.

As teams work through the Analyze phase, they also start preparing for the Improve phase.
During Analyze, teams might begin working on possible solutions and selecting solutions,
developing improvement plans, and preparing some basic documentation about
improvement work. Whether a team begins this work during Analyze often depends on the
individual project and the manner in which the Six Sigma team leader would like to proceed.
The Six Sigma team leader should

the work done for the Analyze phase.

Teams in the Analyze phase might continue to s

during Define or Measure phases, they might begin to do so in Analyze. Six Sigma leaders
can use the same tips for controlling storming in the Measure phase in the Analyze phase.

Another common challenge for Six Sigma team leaders is introducing and explaining
statistical concepts during the Analyze phase. When other team members or even the
champion of the process are not familiar with statistical analysis, presenting advanced
analysis in terms of statistical verb
understand how you came to the conclusions you are presenting, which makes it less likely
they will get behind the solution or improvement in a positive way.

163
 INTRODUCTION TO DMAIC AND DMADV

Six Sigma experts should be aware of the knowledge limitations of various team members
and work to both present information in a way that is understood by everyone and continue
to add to team member knowledge by explaining concepts when possible.

During the Improve phase of a project, a Six Sigma team selects a final solution and begins
to put it in place. Sometimes, teams will select more than one solution, especially if a few
smaller solutions are highly related and work together for an overall solution. It can be hard
to determine which solution actually improves a proces
practice to implement one change at a time and verify that change before moving on to
something else.

Teams might also come up with many possible solutions, all of which would provide some
improvement for the process. They should use a solutions selection matrix or other Six

again that the best solution is not always the solution that provides the most improvement.
Solutions that are so expensive or disruptive that they cause disadvantages that outweigh
any benefits should never be selected by project teams.

During Improve, Six Sigma teams must continue to keep the project definitions in mind. The
solution must address a root cause verified in the Analyze phase; the root cause must be
directly related to the problem stated during the Define phase. After selecting solutions,
teams must test them using statistical tools and real-world sampling to ensure effectiveness
before deploying solutions to a live work environment.

Possibly the most common problem that plagues Six Sigma teams during the Improve phase
is project fatigue. By the time teams come to Improve, they have been working on a project
for weeks or even months; for many team members, the project work is on top of regular
work. Fatigue or frustration might push team members to select and implement solutions
just to have the project completed. Six Sigma leaders have to work to keep teams motivated
on quality and improvement.

The best way for a Six Sigma team leader to create strength as the team nears project
completion is to build a good foundation for Six Sigma in the earlier phases. Teams that
understand the DMAIC process and have at least basic understanding of Six Sigma and
statistical analysis by the Improve phase are more likely to stick with planning, analysis, and
the DMAIC method.
164
 SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

Six Sigma team leaders should also continue to foster a team approach to all aspects of the
project. One challenge for some leaders is the temptation to take measurements and

easier to handle decision-making and analysis on your own, especially when you are dealing

team members, though, and can result in a Six Sigma leader without direct process
knowledge making the wrong decision. Keeping the team involved and making exercises
and meetings fun and productive helps you make it through the Improve phase.

Control is the final phase for Six Sigma teams employing the DMAIC process. During the
Control phase, teams usually handle four tasks: creating the foundation for process
discipline, finalizing documents regarding the improvement, establishing ongoing metrics to
evaluate the process, and building a process management plan that lets the team transition
the improvement to the process owner.

Tools used by a team during the Control phase include documentation checklists, control
charts, response plans, process maps, and process dashboards.

The Control Phase is often easy for a team because the work of the team has already
reached a crescendo. In a well-run DMAIC process, the Control phase is a time of wrapping
up loose ends and arriving at the end of a project. At the same time, teams might find it
challenging to let go of a process they have put so much time into. By the time teams reach
the Control phase, they might have been working with a process for months. If a Six Sigma
leader has done his or her job, the team has taken ownership of the process and feels
personally tied to the quality and output, making it hard to turn the work over to other
teams or employees.

Six Sigma leaders can help team members transition a project by preparing them in advance
for this phase. You might also find ways to incorporate team members into meetings or
presentations where project results are being shared. Six Sigma leaders should always host
a meeting to wrap up the project. The meeting should be somewhat celebratory in nature
if budget, time, and policy allows, Six Sigma leaders might consider having lunch or snacks
at the meeting
members to identify something they learned that can be applied to their own work. This
helps team members see that Six Sigma is an ongoing culture within an organization, and
165
 INTRODUCTION TO DMAIC AND DMADV

the end of
continuous improvement.

Recognition is extremely important when ending a Six Sigma project. Team members might
have put in extra hours to provide excellent work on a project while maintaining their own

duties, so they are going above and beyond what might normally be expected of them. Six
Sigma leaders should make it a point to recognize the work of team members in front of a
project sponsor or champion, and, when possible, in front of the department for which the
improvement is being made.

Design is the fourth phase of DMADV; it replaces the Improve phase of DMAIC. DMADV is
one approach for what is called Design for Six Sigma, or DFSS. Another approach is called
DMADOV, which stands for Define, Measure, Analyze, Design, Optimize, and Validate.
Teams using the DMADV approach usually combine the activities from Design and Optimize,

The Design phase of DMADV is when teams create a new process or develop a new product.
A Six Sigma team would have previously done all the work to lay the foundation for
development during the Define, Measure, and Analyze stages, which means most of the
Design phase is taken up with the actual work involved in creating the process or project.

Using the plans, instructions, or maps created in earlier phases, the team either creates a
product themselves or works with vendors, manufacturers, or other employees to create
the product. For example, if the DMADV project involved the creation of a new app for
customers or employees, the team might work with staff in the programming and technical
departments. They might also work with a vendor who will be supplying the app or software
in question; in such a case, a representative from the vendor should have been part of the
team throughout all phases.

During Design, a team will also test the product, process, or service. Testing can be done in
testing environments, in limited production environments, or via Beta testing. Usually, the
team rolls out the new process or product to a limited number of internal or external
customers; those customers provide feedback and the team uses the feedback to
troubleshoot the new process or product as needed. Seeking feedback and troubleshooting
the new process to create the best possible solution is where the Optimize in DMADOV
comes in.

166
 SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

In a process redesign project, all phases are essential, but Design is often seen as the most
critical. Teams that falter in the design phase can waste the work that was put into other
oject fatigue just as work requirements pick
up for everyone involved. Six Sigma team leaders can help improve the chances of a
successful Design phase by following the tips for managing Improve phases. Teams should
also be realistic about target dates for design work. Promising a complete solution in a short
time period pleases leadership at first, but if teams are rushed, they tend to deliver low-
quality processes. If you promise a too-good-to-be true timeline, you also run the risk of
running far behind schedule, which can impede the work of other projects and process
improvements.

The Verify phase of a DMADV or DMADOV project is very similar to the Control phase of a
DMAIC project. The new process, product, or service is transitioned out of project mode and
handed off to a process owner or employees who work daily with the process or product in
question. Control plans, including control charts, might be put in place by the team to track
ongoing results, and almost all of the tools used in a DMAIC Control phase are relevant to
Verify.

One of the differences between Verify and Control is that DMADV teams might take time to
complete further CTQ analysis at the end of a project so they can identify new critical-to-
quality factors. This is done because the process or product is different than it was when the
team first started working. While the team should have made educated guesses about CTQs
for the new product and used those CTQs in planning and designing they could not
predict 100 percent how the customer might react to the new product or process. A new
process might have a capability the old one did not; having that capability, the customer
might decide it is the most important factor in quality about the process or product.

At the end of the Verify phase, a team delivers a final product or process that meets the
needs first identified in the Define stage. The process or product should be free of known
problems and defects wherever possible, and teams should have provided a way to manage
and control the process through statistical control charts, Lean templates, and policies.

167
 INTRODUCTION TO DMAIC AND DMADV

One of the major differences between DMAIC and DMADV is the possible timeline. We
previously stated that a problem fits the DMAIC model if it can be solved in less than six
months. While some DMADV projects might only take a few months, many process or
product designs can take years. Because of this, the concluding challenges for a DMADV
team are similar to those in a DMAIC environment, but they might be heightened by the
length of time a team has spent on a project.

Team members who have spent a year or more working to develop a new process or
product might feel like the end of the project threatens their job. This is especially true
when team members have not been handling regular work duties in addition to product
duties. Six Sigma leaders and champions can reduce these worries by communicating next
steps and expectations clearly with staff.

Team members who have been working on regular job duties alongside project work for
years might find it hard to return to regular duties without something else to work on. One
of the benefits of Six Sigma is that team members learn to expect more of themselves, their

employees returning to daily work and help them apply what they learned in a positive
fashion within their respective departments.

Finally, Six Sigma team leaders should ensure that a DMADV project closes on a positive
note by validating all team members and ensuring process owners have all the tools they
need to accept the new process without disrupting work.

You should now have a basic understanding of how a Six Sigma team approaches a problem
or process improvement. Whether improving an existing process or creating a new process
or product, teams work through phased approaches. The phases of DMAIC and DMADV
provide control and organization for a project, help keep everyone on task, and let teams
break up what can seem like enormous tasks into chunks that are tolerable. As the old
adage says: How do you eat an elephant? One bite at a time. Similarly, the phased approach
of Six Sigma breaks up the elephant so teams can work on it one bite at a time.

168
Six Sigma teams enter the project process with various levels of information. Sometimes, a
problem is fairly well defined before the team begins work, particularly in organizations that
use a Six Sigma leadership council to choose projects and create teams. Other times, teams
begin work with little information except that a problem of some type exists because
the outcomes of a process are not as expected. Teams might not know where errors are
occurring or even begin a project with a complete understanding of the inputs and outputs
associated with the process.

Whatever knowledge teams begin with, the define phase is when teams move from very
basic information about a process or problem to the knowledge and organization necessary
to enter measure and subsequent other phases with a successful foundation. In the define
phase, teams set rules, create a charter that will govern efforts moving forward, identify
stakeholders and customers, define a process through process mapping, and prepare for a
define tollgate before entering the measure phase.

A project charter, or team charter, is a short document that includes information about the
team and what they plan to accomplish. The purpose of the charter is to set expectations
that can be agreed upon by the team as well as the sponsor or executive leaders, keep the
team focused on the goal, ensure the project remains aligned with the goals of the business,
and documents the fact that control of a process is being moved from a business executive
or manager to a Six Sigma project team.

Minimally, team charters should include:

169
 DEFINE

A complete and concise problem statement that follows the guidelines set out
in Chapter 6.
A list of critical to quality metrics, or those measurements that will ultimately
determine project or process success. Critical to quality was introduced in
chapter 8.
The names and roles of each person on the team. Selecting team members and
appropriate team member roles are covered in Chapter 10.
A list of both internal and external process customers. Use a SIPOC, discussed in
Chapter 7, to begin defining internal or external customers.
The name of a sponsor and/or champion.
A duration for the project.

Teams might also include information such as a list of non-customer stakeholders, an

estimated schedule for each phase of the project, scope definitions for the process or
project, and financial drivers for the project.

The information for

session; the charter is an outcome of the entire define phase, not a quick notation at the
beginning. By taking time to properly consider all elements of a team charter, Six Sigma
teams create a stronger foundation for the rest of their work.

Businesses that are implementing Six Sigma organization-wide might consider creating or
using a specific template for team charters. Templates streamline define phases and make it
easy for leadership teams and other employees to understand critical process components
at a glance. While final team documentation is likely to be extensive, and even in the define
phase, teams themselves might work with lengthy requirements documents, charters
themselves should be as concise as possible. Some organizations distill charters to a single
page while others use multipage documents. A sample one-page charter template is
attached at the end of this chapter.

are worth mentioning again.

170
 SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

The business case might also be referred to as the financial drivers behind a project. Related
closely to the problem statement, the business case is a short statement that provides a
reason the project should be undertaken. The problem statement tells someone where,

we said dollar amounts or another financial metric were important to include in the
problem statement. If you include a business case in your charter, you would build on that
basic financial statement to explain why, specifically, the loss of money, efficiency, or
quality is important to consumers, employees, or the organization. You might also make an
argument for why the problem must be solved now; in essence, why is this project being
run now in place of another project?

We introduced the concept of scope briefly in Chapter 6. For the purposes of the team
charter, the scope should include a hard beginning and end of the process or problem being
considered. You might also include a short list of items or activities that are in scope and out
of scope for your project. A SIPOC diagram helps teams identify the parameters for a
project, and you can also use the In and Out of the Box method described later in this
chapter to understand the intended scope of a project.

The scope should be clear. Listing the scope for a project or process

different points the beginning of the order stage or the end of fulfillment. A better scope
ces an order and ending when the

replacement processes out of scope for a project so that they are only dealing with original
orders. Successful projects have a well-defined scope that is approved and backed by a
project sponsor or champion.

Listing major stakeholders on the charter helps the team remember who and what they are
likely to impact in addition to end customers. Having the list visible during meetings reduces
the chance that the team will initiate changes that might have a negative or unwanted
effect on other process owners or processes, and it helps direct the team to resources
outside of the team that can provide help, access, or information to areas related to the
project.

171
 DEFINE

Team members and roles were covered in Chapter 10, and the team charter simply needs to
list the names of all team members along with their role and expected time commitment.
Adding time commitments to the charter helps sponsors and executive leadership
understand the human resource requirements for the project; often a Six Sigma team
leader has to seek approval for staff members from other areas to devote a specified
amount of time to the project.

Time commitments can be listed in hours per week but are often listed as a percent of the
matter expert who is expected to attend all
of the team meetings to provide input, but is not expected to complete data collection,
analysis, or improvement work, might be listed as providing 10 percent of his or her time to
the team. A list of team members in a charter might look something like the list below. You

the project.

Mike Smith, Black Belt, 100%

Chase Michaels, Green Belt, 100%
Lisa Javes, Green Belt, 100%
Rosalie Myers, Process Owner, 25%
Brent Reed, subject matter expert, 10%
Brenda Tran, subject matter expert, 10%

In Chapter 10, we covered creating a draft schedule for a Six Sigma project. The diagram
included in Chapter 10 that broke down the timeline for a project is called a Gantt chart.
Adapted by Henry Gantt in the early 20th century, a Gantt chart is a bar chart that displays
the phases of a project according to time. One of the benefits of using a Gantt chart to
display a rough project schedule is that it can easily be included in a one-page project
charter; anyone reviewing the charter can quickly visualize the time element required for
the project.

Teams should ensure a date is provided for the end of each of the DMAIC phases and that
all team members agree that the dates are plausible given what the group wants to do. In
some cases, milestones might be set by the project sponsor or champion, but the team
should agree that milestone dates are possible. If dates seem implausible, teams can
present a counter schedule with logical arguments regarding why the original schedule

172
 SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

In addition to milestones at the end of each project phase, Six Sigma teams might also want
to set milestones for work within each phase specifically for the more laborious measure,
analyze, and improve phases. While the team should document all milestones it agrees on,
detailed mileston -page charter document.

Everyone needs to know how the team is going to measure success. If a sponsor is
measuring success on customer satisfaction scores and the team is measuring success on
internal quality scores, ideas about the outcome of the project are likely to differ. Usually,
measures of success can be pulled from the critical to quality metrics discussed in Chapter
8. If teams can convert a CTQ to a measurement, they can understand what major metrics
determine success of a project. While teams might begin to gather measurements or look at
existing measurements while in the define phase, finalization of metrics can extend into the
measure phase.

Financial information is already likely included on the charter in both the business case and
the problem statement. Teams might include expected financial benefits in the business
case section of a charter, but it must be included somewhere. For some sponsors and
executive leaders, the financial benefit is the most important piece of information included
in a charter. An estimated savings or increase in revenue also provides a measuring stick by
which leaders can consider requests for resources for a project.

A Six Sigma expert should never

almost always better to under-promise and over-perform. If you tell leaders a project will
save $500,000 in the first year

project saves only $80,000. As with any aspect of a Six Sigma project, be as accurate as
possible, but be conservative with estimates when accuracy is in question.

Before a Six Sigma team presents a charter for approval, it should take time to review the
document as a group to ensure the charter lays a foundation for success. Some questions a
team might ask itself about a charter include:

Is everything especially the goals, financial expectations, and timeline

challenging but realistic?

173
 DEFINE

Can everyone on the team devote the committed amount of time to the
project?
Is the project backed by a sponsor or champion with enough influence to drive
critical assistance and resources?
Does the team expect to be supported by auxiliary departments such as
information technology, human resources, compliance, accounting, or legal as
necessary for project success?
Does the team expect to have the necessary freedom to implement a solution
it designs after the solution is approved by the sponsor, champion, or executive
steering committee?
Does the team have a leader who is well-versed in Six Sigma tools and project
management?

If the answer to any of the questions above is no, then the team could be setting itself up
for failure. Before moving forward, the team should address these concerns and, if possible,
make changes that convert no answers to yes answers.

Before moving forward with any work even defining a team charter
Six Sigma team to establish some basic rules and requirements for the team. We touched
briefly on this in Chapter 10 when discussing management of a team. The ground rules for a

have to be part of an official charter document. The reason for documenting the rules and
having all team members approve them is because a single team member cannot later claim
to be ignorant of the rules.

democratic process. Some of the more common sense or critical rules can be provided by
the Black Belt or team leader. For example, ground rules should cover topics such as who
should attend each meeting and the fact that team members should hold certain
information confidential. A Black Belt might simply state that team members should
observe confidentiality and attendance rules, be on time to meetings, and respect each
other. The team itself will likely vote on the frequency of meetings and when meetings
should be scheduled. Seeking team member input ensures that all team members can
actually co

schedule through the entire life of some projects.

174
 SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

Black Belts might also provide some tips and suggestions for how team members should
participate during meetings, particularly during group brainstorming sessions. For more
information on running a brainstorming session, see Chapter 35. Black Belts should also
dictate the rules for creating an agenda and running a meeting according to the agenda,
though they might delegate some of these functions such as time keeper and secretary.

ncluding the SIPOC diagram

and the 5 Whys. Process maps and value stream mapping are two advanced Six Sigma tools
that are often used in the define stage. Some Six Sigma teams begin using run charts to start
defining a baseline in the define phase; run charts are covered in Chapter 13 on the
measure phase.

ll cover three additional tools that are common to the define phase: the
Stakeholder Analysis, the In and Out of the Box Method, and the Is/Is Not Matrix.

A stakeholder analysis is a quick way to identify how various people within an organization
relate to a project and how the team should keep them informed. Begin the analysis with a
grid drawn over an x and y axis. The vertical axis represents the amount of power a person
has in the organization. The horizontal axis represents the amount of interest a person has

whiteboard or large flipchart. Draw the basic diagram, as seen in the figure below, in large
format.

175
 DEFINE

Provide the team with sticky notes. Ask them to write down possible stakeholders for a
project or process. Stakeholders are anyone who has an interest in the project, who might
benefit from the outcome of the project, or who might be impacted by the work done via
the project. Take a few minutes to discuss the names that were brainstormed and discard
any the team feels are not actually stakeholders.

Once the team decides who the stakeholders are, begin placing the names on the chart
according to power and interest. People with both low power and little interest will be
placed in the lower left corner. People with higher power and low interest are placed in the
upper left; people with less power but a lot of interest are placed in the lower right corner,
and people with high levels of power and a lot of interest are placed in the upper right
corner. A name might be placed high and to the right of the lower-left corner if the

the person quite crosses the line. Likewise, a name placed at the lower bounds of any
quadrant might show lower amounts of power than those placed nearer the top of the box;
the stakeholder analysis lets teams prioritize stakeholders in this manner, even within the
four quadrants.

176
 SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

Each of the quadrants of the diagram correlate loosely to the type of stakeholder, providing
some guidance on how the team might interact with each person or department listed.

Top Left: Keep Satisfied

Stakeholders that fall into the top left quadrant of the diagram have enough power that
-to-day
outcomes. The team should ensure these individuals are satisfied in whatever way they do
interact with the project. Teams might also consult with these individuals at various times
during the project. For example, a team working to solve problems of shipping in a
warehouse might need to consult with procurement managers at some point because the
team identifies a need for a different type of packing tape. Procurement leaders have
power over obtaining the resource, but they might not have a great deal of interest in the
project overall.

Sometimes, teams might identify a stakeholder that falls into this section and realize that it
would be valuable to the team for that particular stakeholder to be more invested in the
project. Six Sigma leaders can work with stakeholders to try to move someone from low
interest to high interest categories this is a political tactic that some teams use to bolster
support for a project.

Bottom Left: Minimal Effort

Stakeholders that fall into the bottom left quadrant have the least important connection to
a project. Teams will mostly communicate general information about a project via
newsletters or email to these stakeholders. While these stakeholders take minimal effort
from teams, some situations might exist where teams want to move stakeholders from this
box to the lower-right box.

Bottom Right: Keep Informed

Individuals in the lower-right box have a strong tie or interest in the project, but do not have
access to power to support projects from a resource standpoint. These stakeholders might
include employees in departments related to the process being improved or subject matter
experts that will be consulted about individual aspects of a project.

While stakeholders in the lower-

can act in support of a project, often in the form of a goodwill ambassador.

Top Right: Key Player

177
 DEFINE

Individuals in the top-right quadrant are either key players regarding a process or executive
leaders with the ability to assign resources to a project. These are the individuals teams will
report to at various tollgates; often, the executives responsible for the ultimate success of a
process or project appear in the top-right quadrant.

The In and Out of the Box method is a quick and easy method that helps teams define
project or process scope. Begin by drawing a large box on a whiteboard or flip chart. Provide
markers and sticky notes for the team. Ask the team to write down elements of the process
to be worked on, including resources, activities, and people. Each item should be written on
a sticky note. Work together to create as complete a list as possible, placing the sticky notes
on a wall or table as you go. Make sure everyone on the team understands that there are
iscussion. By
brainstorming items with little-to-no discussion, teams can capture more ideas, leading to a
more accurate picture of what is in and out of scope.

Once a comprehensive list is made, begin working as a team to assign each item to a place
inside, on the line, or outside of the box. Items outside of the box are those that will be
considered out of scope for the project. Teams might place items outside of the box
k on
the items, or they have specifically been told not to include the item in the project.

Items that are within the box are considered in scope for the project. These should be
elements of the process the team can reasonably be expected to influence. If
sure yet where an item might fall, they should place it on the line. Items that go on the line
might be ones that the team hopes to be able to impact but for which the team leader
needs to seek permission or assistance from someone outside the team.

Once all items are placed on, in, or outside of the box, review the placements as a team and
make any changes. Document the exercise by photographing the diagram or recreating it on
a computer. The team might reflect back on the diagram when attempting to control scope
creep or considering who they should approach for help with a project.

The Is/Is Not Matrix is another quick brainstorming tool teams can use to define scope. It
can also be used to help define some of the information necessary to a problem statement.
The matrix works by considering specific things about the process or project and coming up
with both is and is not answers.
178
 SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

For example, if a Six Sigma team is tasked with determining why the furnaces in a certain
factory are not heating to proper temperatures, they might create an Is/Is Not Matrix like
the one below.

Is Is Not

Where South plant North or East plan

What Steam furnaces Wood furnaces

When January 2015 Prior to January 2015

The matrix clearly shows that the scope of the project only includes work at the south plant
on the steam furnaces. The problem was noted in January 2015, which provides the team
with a starting point for gathering data. This is a very simple matrix; teams can ask as many
questions as they like to narrow down scope or better understand processes and projects
through the Is/Is Not structure.

A successful define phase ends with all of the following deliverables:

o A comprehensive project statement

o A team charter
o An understanding of the process and a project diagram or map
o An understanding of the Voice of the Customer

A definition of what success will look like that has been agreed on by the team members
and any sponsors or executive leaders

179
 DEFINE

Team Charter Template

Project Name:
Team Members Sponsor:
Name Role Time
Commit

CTQs:

Financial
drivers:

Internal
Customers:

Non-Customer In Scope
Stakeholders
External
Customers:

Out of Scope

180
 SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

Problem Statement:

Objective/Goal

Project Schedule:

181
Moving from the define phase to the measure phase of a project, Six Sigma teams continue
to delve into the process, now coming to understand processes more fully through data.
The measure phase is often the most laborious phase for the team, especially when data is

continue building on the concepts of measure introduced in later chapters on statistical

analysis.

One of the first steps of the measure phase is determining the capability of a process. This
step can be completed before a team formally leaves the define phase if the data needed to
perform sigma level calculations is available. Calculating sigma levels for a process was
covered in Chapter 1. In addition to sigma levels, teams might also calculate various metrics
for a process, including defects per million opportunities, FTY, or RTY, which were all
covered in Chapter 5.

The Failure Modes and Effect Analysis is a tool that can be applied by a Six Sigma team in
any phase from define to analyze. Often, teams begin working with FMEAs in measure
because it helps them identify risk priorities for various inputs and errors within a process.
Used properly, the FMEA uses systemic data and team input to set the stage for root cause
analysis in the next DMAIC phase. Remember, while tollgates do occur and teams move

phases. Teams might begin working on measure phase tasks before leaving the define
ertain that teams will begin some analysis while still collecting data.

Ultimately, an FMEA tool should be used when teams need more detailed information
about inputs and possible associated fail-points than the tools discussed in the define
chapter allow. The FMEA offers some of the information that is offered by SIPOC, but it also
provides evaluations of the inputs. Teams typically create FMEAs in a spreadsheet program,
as some calculations are required during the process.

To create an FMEA, create a spreadsheet with the following column headers:

182
 SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

1. Process step
2. Potential failure
3. Potential failure effect
4. SEV
5. Potential cause of failure
6. OCC
7. Current monitor/control
8. DET
9. RPN
10. Recommended changes/actions
11. Who and When?
12. Action completed
13. SEV
14. DET
15. RPN

Columns 1 through 9 of the FMEA might be completed during the measure phase while
columns 10 through 15 are more appropriate for the improve phase.

Begin by identifying all possible process steps, activities, or inputs in column one. In column
two, indicate what might go wrong for each process step. Note that you can list process
steps more than once if there are multiple opportunities for error within each step. If the
e for a
majority of the steps.

In column three, enter a short description of the impact of the failure on the customer.
Incorrect measurement can result in increased variance in a product, for example. In the
SEV column, rate the severity of the possible failure you described in the previous columns.
Rate the severity from 1 to 10, with 1 being no effect, 5 being minor disruption to
production, and 10 being severe enough to endanger a process or person.

In column five, enter the potential reasons the specific failure might occur, and in the OCC
column, enter a numeric rating for how often the failure might be expected, with 1 being a
very unlikely failure and 10 being an almost inevitable failure.

In column seven, create a short description of the current controls that are in place to
monitor the process or prevent the failures the team has described. In the DET column, rate
the ability of the process or staff to detect failure if it does occur. Rate detection between 1

183
 MEASURE

and 10, with 1 being a process that includes automated detection that rarely fails and 10
being no detection at all.

Finally, calculate the risk priority number by multiplying the severity, occurrence, and
detection ratings, as in the example below.

A Six Sigma team working on a project to improve the speed with which refunds are
processed to customers is creating an FMEA. One row of the FMEA includes the following
information:

Process step: Refund request is entered in system.

Potential failure: Incorrect amount is entered.
Potential failure effect: The customer receives more or less refund than
anticipated.
SEV: 8
Potential cause of failure: Data-entry employee transposes numbers or makes a
similar typing mistake.
OCC: 10
Current monitor/control: A supervisor randomly reviews a sample of refund
requests to ensure accuracy.
DET: 7
RPN: (SEV * OCC * DET) = (8*10*7) = 560

The team completes a second row as follows:

Process step: Refund check is printed.

Potential failure: The printed check has defects that make it difficult to cash.
Potential
new one.
SEV: 9
Potential cause of failure: Printer is misaligned or out of ink.
OCC: 1
Current monitor/control: The person who signs the checks reviews the checks
as they sign them.
DET: 2
RPN: (SEV * OCC * DET) = (9*1*2) = 18

184
 SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

The potential failure in the first example has a much higher risk priority number, which
means, as the team moves forward, they are more likely to work on solving that potential
failure. During analyze and improve phases, the team would recommend changes,
implement the recommended actions, and rescore the process to determine if the RPN of
the changed process is lower. If it is higher or the same, then the change was not a good
one and the team might need to try again.

Creating a baseline metric for a process begins in the define phase, but teams cannot leave
the measure phase without a strong understanding of current process performance. That
understanding begins with figures such as sigma level, but teams should also define a
process-specific metric where possible and gather historical data regarding that metric so
they have something to compare future data against to prove that improvements were
made.

Ideally, the team would have access to historical metrics for the process. In some cases, the

and cover it in depth in the units on sampling.

Before creating and displaying a baseline metric via graphical representation, you have to
understand the type of data you are dealing with. Data is either discrete or continuous, and
teams collect data either as a population sample or a process sample. How teams collect
data and the type of data collected determine how the data can be viewed graphically and
analyzed.

Discrete data is categorical in nature; it is also referred to as qualitative data or attribute

data. Discrete data falls into three categories: ordinal, nominal, and binary, or attribute,
data; some data collected can be expressed in one or more of the discrete categories. For
example, student test scores can be conveyed in an ordinal fashion via the grades A, B, C, D,
and F or in a binary fashion via the Pass/Fail distinction.

Discrete data can be displayed via Pareto charts, pie charts, and bar charts. In some
instances, the data can be converted to run and control charts using variation within the

185
 MEASURE

understanding why a team might want to convert discrete data to be used in a control
chart.

Within discrete data, binary or attribute data is usually the easiest data to collect. Attribute
data records one of the other answers. Does the person choose paper or plastic? Is the
room hot or cold? Is the glass empty or full? Is the light on or off? Depending on the
scenario, attribute data can be very accurate. The light is either on or off; the switch
position tells you that. Attribute data in this case can be automated with the right
technology, which means it would be highly accurate. Whether the cup is empty or full is
another story, because there are so many variations between completely empty and
completely full. If the data is being collected by people, personal biases might enter the
equation. Teams can remove some of those biases and better ensure accurate
measurements, which will be covered in the sections on measurement systems.

Continuous data is quantitative data and is measured in units. For example, the time of day
is measured in hours. Temperature is measured in degrees, and almost anything can be
converted to continuous data by making it a percentage.

Continuous data is visualized in graphs such as histograms and box plots. Box plots are
discussed in chapter 14, and histograms are covered in depth in the chapters on statistics.
Continuous data can also be viewed in the form of run and control charts.

Sometimes, a process or activity can be measured in both discrete or continuous data.

Depending on the purpose of the measurements, teams might need to pick between the
two data types. For example, if a Six Sigma team has identified room temperature as an
input into the quality of product, they will want to monitor the temperature of the room.
They can do so by recording the temperature in degrees every ten minutes; that data would
be continuous. Alternatively, the team might create a tick sheet, having someone make a
mark every hour to note whether the temperature was in the 40s, 50s, 60s, 70s, or 80s with
regard to the Fahrenheit scale. That data would be discrete.

In this particular example, most teams would choose to record the continuous data. Exact
temperature measurements every 10 minutes provides a lot more information than
whether the temperature of the room was in the 70s at the turn of the hour. The
continuous data could be converted to provide teams with the discrete data easily; the
discrete data in this case and in most cases could not be converted to continuous data.

186
 SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

What is true in the example is true for most scenarios. When possible, teams should convert
measurements to continuous data. Continuous data:

Provides more information than discrete data does.

Is typically more time-consuming to collect than discrete data unless teams have
access to automated or computerized data collection.
Is more precise than discrete data.
Lets teams remove variation and errors inherent in estimation and rounding.

Data can be classified at four basic levels: nominal, ordinal, interval, and ratio. Attribute, or
binary, data is actually a limited form of nominal data.

Nominal is considered to be the lowest data classification level and simply involves applying
number labels to a qualitative description so statistical analysis programs and tests can be
applied to the data. The n
about whether the data is better or worse than other data in the listing in nominal data,

An example of nominal data might be applied in a list of birth states for a classroom. In a
class of 30, the number of students born in various states breaks down as follows:

Texas: 6
Louisiana: 4
Arkansas: 10
Mississippi: 1
Oklahoma: 9

In nominal data, each state would be provided a numeric label:

1. Texas
2. Louisiana
3. Arkansas
4. Mississippi
5. Oklahoma

187
 MEASURE

For nominal data, central tendencies are calculated not with means or medians, but with
mode. For example, a list of the nominal data in our example would be as follows:

1, 1, 1, 1, 1, 1, 2, 2, 2, 2, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 4, 5, 5, 5, 5, 5, 5, 5, 5, 5

The mode is the number that appears most in the set; in this case, 3.

Statistically, analysis is limited with regard to nominal data, but some tests can be
performed with statistical analysis software.

Ordinal data is considered to be a higher form of data than nominal, though it still uses
numbers and categories to identify data elements. With ordinal data, though, the numbers
themselves actually provide some meaning. The numbers used in the FMEA scales at the
beginning of this chapter were ordinal data. The numbers are qualitative in nature, but they
are also ranked. Central tendencies with ordinal data are measured by either the mode or
the median, and common uses for this type of data include ranking various things against
each other or rating a specific thing, such as a movie or pain level.

Ordinal data can be arranged in an order that makes sense: on a 1 to 10 scale, Suzy rated
the movies as 2, 5, 6, and 9. If one is the worst and 10 is the best, then we can assume Suzy
liked the last movie best.

While ordinal data comes with a logical order, the intervals bet
mean anything. If Suzy rated movie A as a 10 and movie B as a 9, the conclusion is that she
liked movie A better. Exactly how much more she liked the movie is not discernible using
ordinal data.

Interval data is an even higher form of data classification. Interval data provides numeric
values that can be arranged in a logical order with meaning; the big difference between
ordinal and interval data is that the difference between each interval value provides
meaning. If Frank is keeping track of the temperature in his house and he sees that at 8:00
a.m. it was 76 degrees and at 9:00 a.m. it was 80 degrees, he not only knows that 9:00 a.m.
was hotter; he knows that it was 4 degrees hotter at 9:00 a.m. than at 8:00 a.m. Interval
data is continuous, or quantitative, and offers more flexibility when it comes to statistical
analysis.
188
 SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

Ratio data is considered to be the highest of the data classifications. Ratio data has an
absolute zero point, can be both discrete or continuous in nature, and provides the highest
capabilities for statistical analysis in many cases. Some examples of information that can be
recorded using ratio data include force, defects per million opportunities, voltage, height,
units per hour, and volume.

Measurement systems analysis applies scientific principles to help teams analyze how much
variation a system of measurement brings to a process. The purpose of the MSA is to
identify errors of accuracy within data collection tools. Teams can then redress

the possibility of errors into account when performing analysis on data.

During measurement systems analysis, teams should review multiple components of

possible measurement error. Six Sigma teams analyze:

Whether bias occurs in the accuracy of measurements

Whether the measurement has the proper resolution
What measurement scale linearity exists
Whether measurement activities are stable over time
Whether measurements are repeatable and reproducible

Depending on the measurements the team is dealing with, the MSA can be time consuming
and is often why the measure phase of a DMAIC project is one of the longest.

In this stage of MSA, teams define the difference between the most accurate measurement
possible and the data being collected by the current measurement system. The goal of a
measurement system is accuracy: coming as close as possible to a defined target, if not the
exact measurement. For example, in a computer manufacturing plant, one employee might
solder a chip to the motherboard. For the rest of the chips and wires to be added to the
motherboard, the chip must be placed within a 2 mm area. In this case, a measurement tool
might be implemented with a required accuracy of plus or minus 0.5 mm to ensure the chip
is placed within the area targeted.

Teams can ensure accuracy of data by verifying that the gauge used to collect data is
performing accurately. If a digital scale is being used to weigh ingredients, teams should
189
 MEASURE

calibrate the scale using calibrated weights. If templates are used on a factory floor to make
measurements more efficient, teams should ensure those templates are accurate by
comparing them against known measurement tools such as verified rulers and scales. Note
that, for the purpose of the MSA, accuracy reflects the performance of the measurement
tool, not the operator. Whether the employee uses the measurement tool correctly or
records the amount correctly is considered a concern of precision, which is covered later in
this chapter under R&R Gages.

Once a Six Sigma team is confident that a measurement tool is properly calibrated, they can
instruct employees or others who are responsible for recording data. Data should be
accepted as it is collected for most efficient access and because early review can turn up
specific problems with data collection. When possible, teams should not round data but
collect it as it is recorded.

If data is being collected manually, employees should have a data collection template that
prompts them to collect data at appropriate times and record information about the data
collection event, including the person collecting the data, the machine or process involved,
conditions of the environment especially those that are different from normal conditions
or might have a direct impact on measurements and the measurement tool being used if
multiple tools are an option. These details help Six Sigma teams rule out outliers, which are
discussed in the next chapter.

Before measurements are passed to the analyze phase, Six Sigma experts should review
data to ensure there are no misplaced or missing decimal points, that duplicate entries
-
ddressing obvious data problems
before beginning analyze phases reduces the chance that teams will come to false
conclusions about root causes or viable solutions for a process.

Measuring at the correct resolution ensures that a measurement system can detect change
in the data or process appropriately. For example, if a Six Sigma team is working to improve
a process that cuts pipe for bathroom fixture installations, it might be concerned with the
length of the pipe. In reviewing the measurement system for the cut pipe, the team finds
that the process includes measuring the pipe to the nearest centimeter. If, however, pipes
that are off by several millimeters cause issues in the installation, then the nearest
centimeter measurement is not a small enough resolution.

190
 SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

A good rule of thumb to follow for resolution is called the 10-bucket rule. Break your
measurement resolution into a tenth of what is required. If the pipe must measure within a
range of 5 mm to perform, the measuring tool should measure to the ½ mm. In another
example, a food service department might be tasked with maintaining the correct
temperature in a freezer. To monitor the temperature, an employee records the
temperature once per hour. If temperatures fluctuate quickly in the freezer, a change in
temperature that would impact quality of food or ingredients might come and go between
recordings. In this case, the proper resolution might be gained by recording measurements
every 10 minutes or every six minutes for 10 readings per hour. Even better, in a freezer
with a digital thermostat connected to a network, teams might be able to access readings
recorded every minute.

Resolution is usually one of the easiest things to correct within a measuring system, but it
-effective or plausible to measure at the most detailed resolutions. Teams
should consider resource requirements when developing a measurement system. If,
however, the most detailed resolution is possible, measurements obtained will provide
more information about the process and a larger sample size from which to work.

Linearity describes how a measurement system performs across a range. A standard metric
ruler in the hands of most people is fairly accurate at measuring centimeters, but is less
accurate at measuring millimeters or kilometers. A scale with a range between 0 and 10
kilograms might measure less accurately at either end of the range.

Taking measurements at various ranges with an existing measurement system and

comparing those measurements to data gathered with tools known to be accurate across all
ranges can help teams find errors of linearity. In some cases, teams can develop
mathematical equations to account for the discrepancies. For example, if the scale is
accurate at 5 kilograms, but is off by an extra quarter of a kilogram for each kilogram
thereafter, a measurement of 8.5 kilograms would actually be:

8.5 ((8.5-5)*.25) = 7.625

If mathematical adjustments are not possible, then teams should not use measurement
systems to measure ranges where linearity errors regularly occur.

191
 MEASURE

Stability describes the consistency of measurements over time. If operators are measuring
in the same way and using the same tools
problems described above then measurements should reflect stability on a control chart.
Control charts are introduced in Chapter 16 and covered in depth in later chapters on
statistical process control.

If the variation of measurements, as reflected on a control chart, do not indicate stability,

then teams might want to first rule out a problem with the measurement system before
determining that the process is out of control.

Gage R&R tools are used to ensure repeatability and reproducibility with regard to
measurement systems. In most cases, Gage R&R tools apply to measurement systems that
involve human operators and appraisers. Six Sigma teams apply Gage R&R tests to find
weaknesses within such measurement systems.

In Gage R&R testing, repeatability means that a single employee, using the same
measurement system and appraising the same things, can repeat his or her measurements.
Reproducibility means that multiple employees using the same measurement system and
appraising the same things come up with measurements that match or are very close to
matching.

Most Gage R&R tests fall into two types: attribute and variable. The premise for testing
each type of measurement is the same, though the criteria and statistical analysis following
the test differ slightly.

An attribute Gage R&R is used when Six Sigma teams are analyzing measurement systems
for go/no go data. For example, if operators review an item in the product line and decide
simply to pass or fail it, this would be an attribute measurement. In the example of the
freezer measurements above, an employee might simply be tasked with recording whether
the temperature was in an appropriate range: a yes/no measurement. As previously stated,
attribute measurements provide the least information about a process, so in the case of the

was within appropriate range can be determined systemically from the temperature data.

192
 SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

When attribute data is used, an attribute Gage R&R is used to test the measurement system
following the steps below.

1. Select at least two appraisers.

2. Provide a number of samples. Label the samples in a way that you know which

3. Record the actual attribute measurement for each sample according to the best
possible (most accurate) measurement you have.
4. Have each appraiser record the attribute measurement for each sample
provided (go/no go; yes/no; hot/cold; pass/fail; etc.).
5. Repeat the process with the same samples and appraiser, randomizing the
order in which you present the samples. Randomizing sample order the second
time appraisers are presented with them reduces the chance that appraisers
remember what measurement they recorded the first time and record the
same measurement by default.
6. Enter all data into a spreadsheet or Gage R&R file similar to the one below that
shows a test of a pass/fail measurement.

Sample Actual Appraiser 1 Appraiser 2 Agreement

Label Attribute
Trial 1 Trial 2 Trial 1 Trial 2 Yes/No?

1 P P P F P No

2 P P P P P Yes

3 P P P P P Yes

4 F P F F F No

5 P P P P P Yes

6 F F F F P No

7 F P P F F No

8 P P P F F No

9 F F F F F Yes

10 P P P P P Yes

193
 MEASURE

From the Gage R&R above, you can see that the measurement system is reproducible only
50 percent of the time, making it a poor measurement system. It is repeatable 90 percent of
the time for Appraiser 1 and 80 percent of the time for Appraiser 2, and the appraisers are
accurate 80 percent and 70 percent of the time respectively. Given these results, a Six
Sigma team might determine that there is some problem of clarity with instructions for how
to determine whether a sample is a pass or a fail. The chart above only provides data for a
set of 10 samples; more accurate attribute Gage R&R testing usually requires at least 20
data points.

Not all data is attribute data, which is why teams can also perform variable Gage R&R tests.
While the raw data from a variable Gage R&R test can provide a Six Sigma team with a
picture of whether a measurement system is obviously failing or not, statistical analysis is
usually required to make a true determination about the performance of a measurement
system. This is because, with variable measurements, some differences between
measurements and operators is likely, particularly when measuring to very small or large
figures or capturing data in a moving measurement.

Set up a variable Gage R&R test in much the same way you set up an attribute test, using
two to three appraisers and at least five to ten outputs to be measured. Have each
appraiser measure each sample two or three times, randomizing the order in which samples
are presented to avoid appraisers remembering the measurements initially entered. Record
all data on a variable Gage R&R template, such as the example below.

194
 SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

Sample Actual Appraiser Appraiser Appraiser 3 Variation

Measurement 1 2

The statistical analysis performed in Excel SPC or Minitab by a Black Belt or Green Belt
typically returns four figures:

% Study Variation
% Tolerance
% Contribution
Number of distinct categories

Teams should look to ensure all four elements of a variable Gage R&R test calculation are in
t comes with a scale for safe, or
green, zones along with caution zones and failure zones. If one of the elements falls into a
caution zone and all others into the green, then a team will likely conclude that the
measurement system is sufficient. In some cases, all or a majority of caution zone scores
might be deemed acceptable, particularly if making the measurement system any more
accurate would be costly or cause application issues for other processes. Measurement
systems that score in the failure zone for any element should probably be repaired or
replaced.

Common criteria used to judge each element of the variable Gage R&R calculations are as
follows:

195
 MEASURE

Element Pass Caution Fail

% Study Variation 0 to 10 10 to 30 30 and above

% Tolerance 0 to 10 10 to 30 30 and above

% Contribution 0 to 1 1 to 9 10 and above

# of Distinct 10 or more 6 to 10 1 to 5
Categories

Note: Another tool that is effective in identifying variation in a measurement system is

called the ANOVA, or Analysis of Variance. ANOVA is also useful for analyzing variation of
any type, and will be covered in Unit 5 on intermediate statistics.

Once teams are sure the best possible measurement tools are in place, they can begin
collecting data to be used in the analyze phase of the DMAIC project. The most accurate
conclusions come when a team can analyze data for the entire population, but that is rarely
possible due to time and cost constraints. If you can gain access to automated data or data
warehouses, you might be able to collect population data or extremely large sample sizes
that better approximate population data. Otherwise, Six Sigma teams must randomly
sample the population that is available and use those samples to draw conclusions about
the population as a whole.

To ensure samples can be used to draw statistical conclusions, they must be handled
correctly and be the appropriate size.
sampling strategies that Six Sigma teams might use and why.

Simple random sampling works when there is an equal chance that any item within the
population will be chosen. For example, if you put 20 marbles of the exact size, weight, and
texture in a bag and blindly select one, each marble in the bag has a 1 in 20 chance at being
selected. If the marbles are different sizes or weights, those differing attributes can impact
the chance that each marble will be selected. Heavier marbles might sort to the bottom of
the bag; bigger marbles might be more likely to be picked up.

196
 SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

Random sampling for statistical analysis requires that the sample will represent similar

population is decided by a number of factors.

Stratified sampling occurs when the population as a whole is divided or can be divided into
subgroups with differing attributes. For example, if a shipping company wants to test the
accuracy of its estimated shipping times against actual shipping times, it might assume that
the results will vary according to the distance a package has to travel. By randomly selecting

up with samples for packages delivered within a 200-mile radius.

To prevent bias in the data, the shipping company might divide the population into four
subgroups:

Deliveries within 200-mile radius

Deliveries within 201 to 400-mile radius
Deliveries within 401 to 600-mile radius
Deliveries over 600 miles

By sampling randomly from the stratified subgroups, the team ensures a sample size with
less bias.

Sequential sampling involves selecting every X item for inclusion in the sampling. Sequential
sampling can be used when teams are collecting data at intervals such as time. The team
might collect data every 10 minutes. Sequential sampling can also be used to sample
physical items; every 5th item on a product line might be reviewed. Given the right
parameters and enough time, sequential sampling can provide valid statistical results.
Teams must be cognizant, however, that the sequence of the sampling could, in rare cases,
skew results. It is possible, for example, that something occurs during every 5 th iteration of a
process that causes a difference to occur.

Non-random sampling should not be used when dealing with statistical analysis because it is
more likely to introduce user or sampling error. While all sampling comes with some form of

197
 MEASURE

error, random sampling errors can be calculated and accounted for in analysis. The same
cannot be said of non-random samples.

Non-random sampling includes convenience or judgment sampling. Convenience sampling

occurs w

type of analysis only truly tells the team how the process performed at that exact moment
in time.

Judgment sampling occurs when an expert or knowledgeable person is tasked with

,

most cases, the team members select what they believe is better quality work, skewing any
results from the sample.

One of the major deliverables coming out of define and measure phases is the baseline
metric. How is the process performing now, and what measurement will the team use to
compare current performance to post-improvement performance?

Baseline metrics are numbers, but most teams find that presenting the metric graphically
resonates best with business resources and executives. Visual representations also provide
teams with a quick way to determine if progress is occurring.

The type of visual representation you use depends on whether your major metric is discrete
or continuous. Discrete data can be displayed on Pareto charts (see Chapter 5) and
continuous data can be displayed via run charts. You can also use variation or other
calculations to convert discrete data to continuous data for display in run charts and control
charts (see Chapter 16 for information about control charts).

A run chart can be used to monitor the performance of any variable or process over time.
With a single, intuitive chart, Six Sigma teams can display trends, shifts, and cycles within a
process; they can also monitor a process for concerns, though run charts are not as
effective at this as the very similar control chart is.

A basic run chart is simply a line plot of the data over time, which means anyone can create
the chart. Most Six Sigma run charts also feature a line representing the median of all data
points for visual reference. Depending on the type of information being charted, you may
198
 SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

need to convert data to a ratio for a more accurate run chart. For example, if you are
plotting the temperature of a surface over time, there is no need to convert data. If you are
plotting the number of patients readmitted to a hospital within 30 days of being discharged,
then it helps to convert the data to a percentage of the number of patients discharged
within the same time period. In a 30-day period where 10,000 patients were discharged,
you can expect a higher number of returns than a period during which only 5,000 patients
are discharged.

The figure below illustrates a run chart of temperature over time. You can see how
temperature changes through time and begin to see some possible trends. A Six Sigma team
would be able to zoom out, viewing the run chart over more time to validate trend
assumptions. You can also see that the median temperature for the process is 33.

The run chart below indicates the number of returns per hundred sales for each month of
the year. You can see that returns as a rate of sales increases steadily during the first part of
the year before holding steady from May through November. The orange line indicates the
median returns per hundreds sales, which is just under 7.

199
 MEASURE

Statistical analysis software, including Minitab and Excel SPC, creates all elements of a run
chart automatically from entered data, but anyone can use basic Excel functions to create a
run chart if needed.

First, create a data table.

Creating a data table for a single attribute, such as temperature, just requires entering the
time labels in one column and the attribute measurements in another. For the example,

1. Enter the data labels (month, week, hour, etc.) in the first column of Excel.
2. In the next column, enter the corresponding measurements for the attribute you
are interested in: in this case, the total number of returns per month.
3. In the third column, enter the total number of items you are comparing the
attribute to: in this case, the total number of sales per month.

200
 SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

4. In the fourth column, calculate the percentage the first column of data is of the
second. In this case, the percentage of returns per sales for each month. The
calculation is achieved in this case by the formula =B2/C2 for January, =B3/C3 for
February, and so forth.

201
 MEASURE

5. Decide whether you want to create a run chart showing percentages, or if you
would like to create the chart illustrating rate per 100, per 1,000, etc.
6. If you want to illustrate a rate per (x), multiply the percentage calculation in the
fourth column by (x). In this case, the figures in column D are multiplied by 100.

7. Use Excel to calculate the median of the number you plan to chart. The median is

in the formula correlate with the range of all the charted data points. In this case,
the median is 6.96679
8. Highlight the data labels (in this case, column A) and the figures to be charted
(column E)

202
 SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

9. Select Insert > Charts > Line Chart to insert a line chart of the attribute or attribute
calculation.

10. Select Insert > Shape > Draw Line.

203
 MEASURE

11. Draw a line on your run chart approximately where the median would be. Use Excel
tools to select a color and thickness for the line that you desire.

The completed run chart can be used to present information to the Six Sigma team or
include a graphical representation of baseline process performance in a measure phase
tollgate presentation. Again, it should be noted that manual creation of a run chart is not
required for most Black Belts and Green Belts, who will have access to statistical analysis
software.

Use the checklist below to determine whether a team is ready to move from the measure
phase to the analyze phase of a DMAIC project.
204
 SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

o The team has agreed upon the key measurements and come up with a baseline
measurement of process performance.
o The team has analyzed measurement systems and identified any issues that might
contribute to analysis errors.
o Where possible, the team has corrected measurement systems to remove error
risks.
o The team has calculated process variation and sigma level.
o The team has conducted appropriate sampling to allow for statistically valid
conclusions in the next phase.
o The sponsor or champion has reviewed and signed off on all elements of the
measure phase.

205
long and requires a great deal of work. Without a strong measure phase, the team cannot
move on to analyze data and make data-based decisions that drive improve and control
phases. Analyze phases also require a lot of work, but that work is usually performed by
Black Belts and Green Belts, who report findings to the Six Sigma team and ask for feedback
about analysis and verification of analysis.

In this
s and units. Units 4, 5, 7 and 8
provide in-depth information about the statistical tools referenced throughout this chapter.

One of the fundamental activities of the analyze phase performed by the entire team with
help from identified subject matter experts is the root cause analysis. Root cause analysis is
used to identify root causes for problems or defects when a team has reached the analyze
phase without a clear idea of primary causation. Some of the tools described for identifying
root causes in this chapter could be used in either define or measure phases at the
discretion of the Black Belt leading the team; the FEMA described in Chapter 13 on measure
could likewise be used in the analyze phase as part of root cause analysis.

A popular method for brainstorming and analyzing causation in a process is the fishbone
diagram. The fishbone diagram can be completed by a single Six Sigma expert, but it
typically has more value when it is completed by the entire Six Sigma team. The diagram
lets teams concentrate on a brainstorming process that generates ideas about possible
problem causes, organizes those possibilities in a logical way, and lets teams visualize the
information to identify priorities, trends, and relationships between ideas. When used as a
team activity, the fishbone diagram encourages participation and input from all team
members, which increases the chance of laying the foundational work for a viable and
original solution.

206
 SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

The cause and effect diagram is called the fishbone diagram because you begin with what
looks like a simple drawing of a fish skeleton. Reference the diagram below and follow the
instructions to create a fishbone diagram as part of a team brainstorming exercise. You can
also use these instructions now to practice a fishbone diagram based on a process or
problem you have experience with.

1. Sketch a basic fishbone shape on a whiteboard or large flipchart. Write a

have to conduct a fishbone diagram only on the problem statement from the define
phase. Teams might also conduct the diagram on a specific defect or issue found
during the define or measure phase. For example, if the problem statement
discusses the amount of waste in a restaurant, the team might have discovered
during the measure phase that bread is being thrown away at a rate higher than all
other food items. One of the activities handled in the analyze phase might be a
fishbone diagram specifically about the reason for so much bread waste.
2. Draw a fish spine and four major connectors. Label each connector as People,
Process, Materials, and Procedure. Some Black Belts also include two other major
connectors: Equipment and Environment.
3. Explain the categories of the fishbone diagram to the team. Note that there are
some places, especially with particular processes, where the various categories will
overlap. Some ideas generated by the team as they complete the fishbone diagram

a. People references anyone who carries out or interacts with a process.

b. Process or machine refers to the process by which inputs become outputs.
207
 c. Procedure or method refers to the way things are done, whether by written
documents or unwritten rules.
d. Materials are the inputs, such as raw goods, into the process.
e. Equipment includes the technology or machines required to handle the
work.
f. Environment is the immediate area surrounding the process.
4. Begin with each category on the fishbone diagram, asking the team how something
in that category might be responsible for a problem or defect.
5. Use sticky notes to write down ideas and place them on the fishbone diagram so
you can move ideas around later. You can also write directly on the diagram.
6. Couple cause-and-effect brainstorming with the 5 Whys exercise described in
Chapter 6. For each branch of the fishbone diagram,
to ensure the most granular detail possible.
7. Once the team has run out of ideas for the first category, repeat steps 4 through 6
for all other categories.
8. Take some time as a team to review the diagram, discussing the placement of
potential causes, and moving them to appropriate categories and subsections to
create an organized visual representation.
9. Remove or cross-
10. As a team, decide which root causes seem most likely or highest priority. Circle
those causes as high-priority possibilities for further investigation.

To provide a better idea of how a fishbone diagram works, consider the example image
h how the team came up with the information recorded on the
procedure/method line of this diagram.

208
 SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

The team above was working to solve a problem of burnt cakes in a food-service bakery.
When discussing the method by which the cakes are being baked, the team first came up
the with the reason that the cakes were being baked at inconsistent times. Perhaps,

oven too long. The idea was written down.

The instructions for baking cakes are vague, said another team member this time, a

the team determine that new cakes were added to the menu without the overall
instructions for bakery staff being updated.

After digging deeper into the inconsistent baking times, the team again asked themselves
how methods could be responsible for burnt cakes. Someone suggested that the
temperature in the oven was too hot, and the team tied that suggestion to the same root
cause as the inconsistent bake times. Upon final review, someone noted that the suggestion
akery was equipped
with alarms that sounded when baking time was done. The team crossed that idea off the
diagram.

In this case, the Six Sigma team might prioritize the fact that instructions are not available
for all types of items being prepared in the bakery. Because this would likely be a simple
and common-sense improvement to make, the Black Belt might even assign someone to
begin working on the improve phase as soon as the cause was verified. Many times, the
root cause is not as obvious and the solution for the cause even less obvious, requiring
additional analysis and validation before moving forward.

Once teams identify possible root causes, they must verify that the causes are valid. Root
cause verification can be completed via a variety of methods, including statistical analysis,
design of experiments, logical questioning, observing a process, gathering additional data,
analyzing data via graphical representation, and mapping processes at a more granular level
than accomplished in the define phase. While this chapter touches briefly on statistical
analysis and graphical representation, those topics, as well as experiments and process
mapping, are covered in later units.

209
Whatever method is used to validate root cause assumptions, the Six Sigma team should
document it. Documentation regarding root cause verification is usually completed on a
matrix that includes the problem, possible root causes, the verification method, why the
verification method was chosen, results of the verification, and, in some cases, whether a
senior Six Sigma leader, such as a Master Black Belt, agrees. A template for such a matrix is
included below, but teams can also create similar documents in Excel or Word.

Problem Possible Method of Reason for Verified? Notes

Root Verification Verification
Causes Method

The root cause verification matrix for the burnt cake example might be completed as
follows:

210
 SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

Problem Possible Root Method of Reason for Verified? Notes

Causes Verification Verification
Method

Cakes in Temperature Run chart of Allows team to Yes

the too hot temperature visually
Delaware against determine
bakery are required whether
coming temperature temperatures
out burnt exceed
10 percent requirements at
of the any point during
time. bake process

Bake times Box plot of Provides visual Yes

inconsistent bake times representation of
per type of the variation per
cake cake type; bake
times should not
vary widely by
cake, so the
boxes should be
flat; lets teams
determine if
certain cake
types are more of
a problem.

Instructions Process Easy way to Yes

not provided observation determine if
to staff bakery staff have
the instructions
necessary to
complete work
without defects

211
Six Sigma experts and teams can use a variety of graphical analysis tools to help generate
ideas about root causes or understand how inputs and outputs really impact each other.
Some of those graphical analysis tools require statistical analysis software, and those will be

can create easily with Excel.

The first graphical tool for validating root causes is the Pareto chart, which was covered in
chapter 5. Chapter 5 discussed the Pareto Principle, or 80/20 rule, which says that 20
percent of the causes lead to 80 percent of the results. Because of this, a Pareto chart is a
good starting point for root cause brainstorming teams can start with the few inputs or

fishbone diagram, asking deeper and dee

Pareto chart.

Consider the Pareto chart illustrating reasons for medical claims denials from Chapter 5.

In this case, the team might decide to concentrate on the top two claims denial reasons:
duplicate claim and timely filing. The team could use various methods for digging deeper

212
 SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

into root causes for these two claims denial reasons. They might perform a fishbone
diagram to discover why duplicate claims are being generated. To understand the timely
filing problem, teams might gather additional data for graphical analysis.

Timely filing means that the claim was not originally filed with an insurance company prior
to the deadline for claims submission. Different insurance companies have various timely
filing requirements, and the countdown usually starts at the time of service to the patient or
the time of discharge from a facility. The team might want to understand which payers are
denying claims for timely filing, so they collect data as follows on how many timely filing
denials are associated with each payer. Because Pareto analysis is concerned with the top
few, you can lump the many others together in a single entry and, for the purposes of the
Pareto analysis, ignore them. A medical provider might bill claims to dozens of providers;
including every provider on the data table and Pareto chart would be a waste of both time
and space for this particular exercise.

Payer Denials
A 3512
B 2779
C 1575
D 1142
E 945
F 847
G 702
H 502
Others 1241

Converted to a basic Pareto chart, the data is illustrated in the graph below.

213
these companies have shorter timely filing guidelines than the other companies or billing
staff is unaware of the proper timely filing requirements for those payers. Six Sigma teams
can begin asking questions specific to these payers as they continue analyzing data and
discovering root causes.

Box plots are another graphical representation that can be handled with Excel. In later

testing and other analysis. When differences between distributions are marked, however,
or when outliers are fairly obvious within data, the image of a box plot tells its own story
without requiring advanced statistical knowledge.

Box plots are often called Box-and-Whisker graphs. To understand how to read a box plot,
consider the data table and graph below. The data table shows the time in minutes in which
various runners completed a one-mile race. The results are divided into the categories
Children and Adults.

214
SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

Children Adults
9.9 4.1
10.2 4.1
11.6 4.5
12.7 4.6
13.8 4.7
13.4 7.5
13.4 8.3
13.9 8.9
15.2 9.7
15.8 9.8
15.7 10.1
16.2 11.6
16.7 13.5
18.9 13.7
19.4 14.8
20.1 14.9
29 15

215
The above image shows a box plot of the data table, generated in Excel. Even without
further explanation, you can likely tell that the children completed the race on average
slower than the adults. The blue box, which represents the children, is shorter than the
orange box, indicating that the middle 50 percent of children completed the mile-long run
in times that had less variation than that of the middle 50 percent of adults. To understand

A box plot begins with the upper and lower hinge the top and bottom of the box. The top
represents the 75th percentile; the bottom represents the 25th percentile of the data. The
line within the box represents the 50th percentile. Within the box are the 50 percent of data
points between the 25th and 75th percentiles.

Each box plot receives upper and lower whiskers indicating the range of most of the other
data within a set. In this case, Excel creates whiskers that extend to the top and bottom of a
range barring any statistical outliers. Some statistical analysis software or methodologies
use other methods to create the whiskers with very similar results.

Finally, since all plot points must be represented on a box plot graph, outliers are indicated

minutes. That particular point is a statistical outlier; Six Sigma teams who note outliers on
box plots should consider the data that is shown as an outlier. If an explanation can be

216
 SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

found for the outlier, it can be ignored. For example, if the child who took 29 minutes to
complete the mile was much younger than the other children or was walking with an injury,
the data point is explained and can be excluded from further analysis.

In addition to calling out outliers, box plots let you compare two distributions graphically to
see if, as in the above example, there are obvious differences between the data sets. Box
plots are useful in comparing how various attributes impact a process. Six Sigma teams
might compare process results for different operators, different times of day, different

alter one attribute or input; the cause of any

difference between data sets was if a statistical difference does seem likely on a box plot.

Use some information for a work process you are familiar with, or use the sample data
provided, to create box plots in Excel following the steps provided below.

A department manager believes that the staff on her teams would be more productive if
they were able to work with two computer monitors. Because outfitting an entire

that her assumption is correct. The manager equips a few stations with dual monitors and
lets different team members work at the stations. She records the amount of work done
within hourly increments at stations that have dual monitors as well as stations that have
single monitors. Her data is featured in the table below.

One Two
Monitor Monitors
9 10
8 4
4 9
7 7
2 8
6 7
1 9
8 9
5 14
4 10
3 12
7 7

217
 9 9
5 4
1 9
5 7
6 8
7 14
4 13
2 10
9 12
5 12
4 8
8 6
2 10

1. Copy the data from the table above into Excel.

2. Highlight all of the data cells, including the header row.
3. Click Insert > Statistic Chart > More Statistical Charts

4. Select Box and Whisker and click OK.

218
 SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

5. Using normal Excel chart editing functions, edit the title and data labels of your
chart as desired.

219
While a Six Sigma Black Belt would be able to back up the conclusion with hypothesis testing
or other analytics, the manager might get her request for monitors approved with nothing
at the second monitor did increase production
capability for staff. Another thing worth noting is that the distributions for each of the boxes
and whiskers is similar, which is somewhat expected. High producers are still going to
produce the most, and low producers are still going to produce less than high produces,
even if everyone is producing slightly more with the new set-up.

Note that the two examples used in this chapter used data sets that were different enough
to be visually noticeable on a box plot
other graphical representations are often only the starting point for analysis.

Because statistical analysis is covered in several future units, this section lists some common
statistical analysis tools with definitions.

220
 SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

Hypothesis testing lets Six Sigma experts draw conclusions about the population based on
statistical analysis performed on a sample. Because the conclusions are based on samples
and not the entire population, there is always some risk of error. You might have seen or

is the value for the

error risk.

population is known as the alpha-risk and the beta-risk. Using information about the sample
and alpha and beta risks, statisticians calculate what is called the p-value. The p-value is a
probability estimate that tells statisticians how likely an assumption or conclusion drawn on
sample data will be incorrect.

Statistical software removes a lot of the manual calculations from the process of setting up
and running hypothesis tests. With Minitab, for example, Six Sigma experts can conduct
hypothesis tests on prepared data with a few mouse clicks. They do have to know which
types of hypothesis tests to use in which situations.

Regression and correlation analysis helps Six Sigma experts understand how variables within
a process might be related. Regression analysis helps teams define the relationship between
one independent variable possibly an input and one dependent variable possibly an
output. Does the temperature in the oven have a relationship to whether the cake is baked
correctly, and how close are the two things related? Does the number of hours a person
works have an impact on his or her productivity can the team show a correlation between
lower production as employees approach the end of a shift? These are the types of
questions that regression analysis can answer.

To work with regression analysis, both of the variables being studied have to be in
numerical format. To conduct a regression analysis regarding the relationship between oven
temperature and whether a cake is baked correctly, a Six Sigma team baked cakes at varying

hardly cooked at all; a rating of 5 indicated a perfectly baked cake. At 10, the cake was
completely burned. The temperatures and corresponding ratings are seen in the data table
below.

221
 Oven Temperature Doneness of Cake
200 1
225 1.2
250 2.2
275 2.4
300 4
325 5
350 5.2
375 6.2
400 8
425 8.9
450 9.5
475 10

Using this data, the team creates a scatter diagram with a trend line, as seen below.

Just looking at this graph, you can tell that a relationship does exist. The data plots are
gathered tightly around the trend line, which indicates that as temperature rises so does
the doneness of the cake.

222
 SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

determine relationships between variables, especially when those relationships are

complex, or multiple variables are present. Because the analyze phase sets the stage for the
improve phase, teams have to be as certain as possible in their analytical conclusions before
they decide on solutions for implementation. A design of experiments can provide the more
granular details and analysis required for that level of certainty.

Design of experiments, or DoE, is performed via statistical analysis software such as

Minitab. Teams can set up experiments for one factor or multiple factors.

o Primary root causes have been identified.

o Team has prioritized root causes.
o Champion or sponsor agrees with team priorities moving into the improve
phase.
o Where possible, root cause assumptions are backed by statistical data.
o Relationships between variables within a process are understood.
o Where possible, variable relationships have been confirmed with statistical
analysis.

223
During the improve phase of a DMAIC project, Six Sigma teams brainstorm possible
solutions for the root causes identified in the analyze phase and rank those solutions
according to costs, how effective the solution would be, and how likely the solution could
be implemented. Analytical rankings are used to prioritize and select solutions for
implementation. Teams pilot solutions through beta tests or small roll outs, collect data on
the solution, and verify that the solution is working as expected via statistical analysis. Once
the team is confident that the solution works to address the problem, it plans and
implements a full rollout of the solution.

A solutions selection matrix is an analytical tool that lets teams propose and rank solutions
for any of the root causes identified in the analyze phase. While teams can work
simultaneously on multiple solutions if multiple prioritized root causes were found, Black
Belt and other team leaders must remain cognizant of timelines, resources, scope, and the
purpose of the project. If a single solution provides enough positive impact to reach project
goals, then other solutions might be saved for future improvement projects. If one solution
would reach results, but another solution would be especially easy to implement and
provide additional positive results, the team is likely to decide to implement both.

A solutions selection matrix can be created in Excel, and teams should work on the

as well as relevant subject matter experts and stakeholders during solutions brainstorming.
This ensures the solutions the team comes up with are more likely to be a realistic fit for the
process and business; once solutions are selected using the matrix, teams will also likely
have to get sponsor, champion, or leadership council support before a partial or full
implementation is possible. This is especially true where solutions require expenditure or
will impact processes and people outside of the project

An example solutions selection matrix is pictured below.

224
 SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

The matrix is completed by:

1. Entering the problem statement in the first column. This should be the final
problem statement that was arrived at during the measure phase if the team
decided that the statement should be altered after gathering data. Otherwise, this
can be the problem statement from the define phase.
2. Entering a priority validated root cause from the analyze phase. If the team is going
to attempt to solve more than one root cause during the improve phase, it should
create a solutions selection matrix for each root cause.
3. Brainstorming potential solutions in column three. During the brainstorming
process, teams should not question or attempt to analyze solutions, but should
record any solution suggested that seems at all viable. The only solutions that
might be ignored are those that are clearly out of scope or impossible, but the
Black Belt leading the brainstorming exercise will have to use his or her own
discernment about such suggestions.
4. Noting, at a very high level, the practical methods by which a solution could be
implemented. In one example used in the analyze chapter, a commercial bakery
had a problem that was caused by lack of baking instructions. In this case, the Six
Sigma team might propose that staff be provided with proper instructions. The
method by which that solution occurs is written documentation and training.
5. Rating solutions. After a list of possible solutions and practical methods is created,
the team rates each possibility on effectiveness, feasibility, and cost-benefit. Each
category is given a rating between 1 and 10.
a. Effectiveness is the measure of how well a solution will eliminate a root
cause for a problem, with 1 being not effective and 10 being highly
effective.

225
 IMPROVE

b. Feasibility is the measure of effort required to implement the

improvement, with 1 being not feasible because of the effort or resources
required and 10 being highly feasible.
c. Cost-benefit is an estimated measure of how the costs of a project
compare to the savings expected. This rating is not a formal cost benefit
analysis, but is a high-level estimation. If savings are expected to outweigh
the costs associated with a project, the team ranks the solution as high.
Otherwise, the team ranks the solution low.
6. The scores for effectiveness, feasibility, and cost-benefit are multiplied to calculate
an overall score. The overall score can be used to prioritize solutions and select the
solution that features the best overall effectiveness, feasibility, and cost-benefit
rating.

Consider a possible solutions selection matrix created regarding the medical claims denial
example used in the analyze chapter.

226
 SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

During the analyze chapter, a Pareto analysis indicated that timely filing issues were a root
cause for high claims denials. In the solutions selection matrix above, a team has identified
and prioritized four possible solutions after discovering that claims were not filed on time
ation in time.

The first solution is to require the front desk to collect as much information as possible. The
team gave this solution a mid-
fact that clinical information can still be missing from the claim. But, the process would be
effective at gathering demographic information and would not require large expense or
effort to implement, so the team rated it high for the other categories.

Next, the team considered creating and requiring a patient email or portal system so billing
staff could communicate quickly with patients when information was missing. Because not
all patients have email and many would be unlikely to use such a system, the team ranked
this solution as low for feasibility and effectiveness.

The third solution considered by the team was requiring clinicians to include all information
necessary for billing as they charted during the visit with the patient. The team thought this
solution would be fairly effective, but that clinicians would be unlikely to take the time to be
so thorough when dealing with patient issues. The solution might also slow physician visits,
resulting in a negative impact to revenue.

Finally, the team considered a solution that put certain staff members in charge of claims
with missing information. Because those staff members could concentrate on missing claims
and would work through a new workflow built by the team, they would be more likely to
file claims on time. Overall, the team ranked this solution highest, but they decided to
implement both solutions one and four because solution one required so little effort and
would actually help drive the success of solution four.

When teams include the appropriate individuals in the process, a solutions selection matrix
is very adept at identifying the best possible solutions for implementation. Leadership
councils and executive sponsors often want more information about the costs and benefits
expected for a solution, though, so teams should be prepared to create a cost benefit
analysis. Black Belts often have many of the elements required for such an analysis after the
measure and analyze phases, but teams might also need to work with accounting, finance,
or business planning departments to gain accurate financial details required.

227
 IMPROVE

The goal of a cost benefits analysis is to compare the costs of implementing a solution with
the monetary benefits expected from the solution. Costs include expenses such as software
development or purchase, equipment purchase, building development or renovation,
additional labor or hiring, training expenses, additional supplies, and any losses associated
with disruption as the solution is implemented. Benefits might include an increase in
product margin, increase in revenue, cost savings or avoidance, and intangible benefits such
as increase to staff morale or customer retention.

might be completed by a certified accountant, but they can create an idea of cost benefit
relationships via the payback method of analysis. This is the simplest way to approach such

To conduct a payback analysis, Six Sigma teams must have an estimated cost for the project

include increases in revenue. Cost savings, new customers, or mitigation of customer loss
can all be considered as contributing to benefits each year. The team also needs to
understand the estimated operating costs of an improvement for each year.

The formula for this analysis is:

(Cost of implementing solution) / (Annual financial benefits annual costs)

If a project costs $50,000 to implement and $2,000 per year in extra labor, and the team
expects the project to generate $15,000 in financial benefits each year, then the calculation
is:

$50,000 / ($15,000 - $2,000)

$50,000/$13,000

Or, approximately 3.84 years until the project pays for itself.

Obviously, organizational leaders are looking for payback calculations that are as short as
possible, but if the solution will solve a major problem or set a foundation for extraordinary

table.

228
 SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

A more concise way of calculating cost benefits is known as the Net Present Value, or NPV.
NPV adjusts benefits and costs as time passes because cash flow in the future is not as
valuable as current cash flow due to inflation and other economic factors. The discount rate
for various endeavors can be calculated using expected return, interest rates, or inflation
rates. Often, corporate finance departments can provide Six Sigma teams with a discount
rate used for NPV in the company. A very basic NPV model is shown below.

This year First Year Second Year Third Year

Total

Benefits

Increased $10,000 $20,000 $30,000

revenue

Cost Savings $5,000 $10,000 $15,000

Costs

Capital -$20,000 -$20,000

Equipment -$15,000 -$15,000

Training -$5,000 -$5,000

Labor -$2,000 -$2,000 -$2,000

Total Benefits -$50,000 $13,000 $27,000 $3,000

Cost Annual

NPV $12,350 $25,650 $2,850

(Discounted 5
percent)

229
 IMPROVE

In this model, the team spent $50,000 implementing a solution and expects a $2,000 labor
expense associated with the solution each year. The estimated benefits from the solution
are recorded for the first two years, and all of the benefits and costs are added up for a
third-year view. The NPV is discounted by 5 percent, and the final number is seen in the
bottom right cell: $2,850. The goal is a positive NPV, so this project fits that goal.

Once a solution is selected and work done to bring it into production at a minimal level, the
Six Sigma team is ready for a pilot. A pilot is a limited trial of a solution in a live
environment. No matter how much analysis was completed or how well test cases were

when people are involved in the process.

Benefits of a limited pilot include:

Use of resources are limited, which reduces waste if the solution turns out to
be incorrect or not effective in resolving the problem
Confirmation that expected results occur
Allows troubleshooting of a new solution on a smaller scale to minimize
disruption during full transition
Lets employees outside of the Six Sigma team provide feedback on the solution
and implementation to make the final rollout more successful

made
without piloting as long as teams document the changes well and measure results for
verification. When changes are large in scope, could cause expensive or expansive
consequences, or would be difficult to reverse, teams should begin with a pilot. The same is
true for any solution that might be expensive to implement at a full scale.

Pilots can occur on a limited scale or for a limited time. Limited scale pilots incorporate a
specific region, team, group of people, or machinery. Limited time pilots implement a
temporary change; at the end of a scheduled time, the team makes a decision about
whether the change should be made permanent.

Pilots can occur with either processes or products. Process pilots might feature testing
specific locations, testing results with some customers, working with some employees to
test new processes, or conducting dry runs of a process without impact to the end-user.

230
 SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

Product pilots are conducted using test markets, product models, or alpha and beta testing
of the product with certain end-users.

To create a pilot, a Six Sigma team must first select the audience for the pilot. Internal
process pilots can be performed by a select team or a select few employees. External
product pilots can be performed using a subset of customers. For the best pilot results,
avoid biasing results by selecting the best possible performers or customers who are most
likely to work hard to ensure a product succeeds. When the solution is implemented on a
full scale, it will be used by everyone, so you want to ensure it works for everyone.

When possible, pilot at a very small level and then expand the pilot to a larger audience.
This is the premise behind alpha and beta testing. A very small set of loyal customers tests
the product first, because you know they will provide feedback. Next, an expanded set of
customers tests the product after teams have made changes associated with feedback from
the first group of users. Finally, if the limited tests are successful, teams choose to roll out
the product to the entire audience.

Six Sigma experts can use all the tools associated with the analyze phase to test whether
solutions have a positive impact during testing or pilot programs. Hypothesis testing can be
used to compare data from before the solution to data after the solution, determining if
there is a statistically significant and positive change. Graphical analysis can be very helpful
in demonstrating for executive leaders how a solution has positively impacted a problem by
reducing defects, improving production or efficiency, or reducing costs.

Once solutions have been verified through tests and analysis, teams can begin the work of
implementing changes on a large scale. Teams should create and work from an action plan
during this vital and active stage of the DMAIC process to ensure that no plans or
requirements fall through the cracks. During the active part of the improve phase, the
project leader will likely delegate numerous activities to members of the team, and the
team will also rely on input and assistance from those outside of the Six Sigma team.
Keeping an action plan document helps everyone on the team see where they are in the
process, what they are responsible for, and what date work must be completed by.

If the Six Sigma team is working with a project manager, he or she is likely responsible for
action plan documentation and follow-up. Teams can also use a basic spreadsheet or Word

231
 IMPROVE

document, which should be saved in a common location, to keep track of what work is to be
done and who will do it.

Because every project is unique, improve action plans are also unique. Most action plans
will contain common tasks such as documentation, training, and transition.

First, the Six Sigma team should have documented all of its work so far during the DMAIC
process. If asked, the Black Belt or designated team member should be able to present
analysis in the form of data tables, statistical calculations and explanations, and graphical
analysis. All brainstorming activities and diagrams should be saved in a central file location
where all team members can view them and appropriate team members can edit them as
needed. Access to these documents helps team members work efficiently on additional
documentation required for implementation.

Almost any organization will require new processes or changes to processes to be recorded
in standard operating procedures. Depending on the organizational structure of a business,
Six Sigma teams might be responsible for drafting such documents or they might need to
work with knowledge management resources to create SOPs in keeping with corporate
branding and templates. Teams might also create general communications letting other
staff members know about the upcoming changes and the reasons for those changes as well
as general reference documents such as cheat sheets and Frequently Asked Questions.

Strong documentation is key to the next part of implementing improvements, which is staff

improvement they make. Instead, teams begin the training process by working with training
subject matter experts or delegates in the department impacted by the upcoming changes.
The Six Sigma team trains these individuals, who then go on to train other individuals who
will be impacted by the process. At some point, the process training should be integrated
into regular organizational training by trainers or knowledge management departments.

During the improve phase of a DMAIC project, teams should begin to consider the need to
transition a process back to the business and traditional process owner. Transition is part of
the control phase, but teams should move from improve to control with a good
understanding of how the process should be measured and monitored. Strong

232
 SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

documentation and training during the improve phase helps cement the success of the
control phase.

o Solutions were reviewed and prioritized.

o One or two top solutions were selected for action.
o Solutions were implemented on a limited basis.
o Data from limited trials was analyzed and solutions appear to work as
expected.
o Cost-benefits analysis was performed.
o Sponsor, champion, or executive steering committee signed off on
implementing the solution completely.
o All team members agree the solution should be implemented.
o The solution is fully documented through SOPs and training materials.
o Critical staff received training on the solution and are prepared to pass that
training on to others.

233
The last stage of a DMAIC project is control. During the control phase, teams build monitors
that let them ensure the process continues to work successfully after changes are
implemented across the regular business process. At the same time, Six Sigma teams work
to transition the process back to the process owner.

Up until this point in the DMAIC process, Six Sigma teams have worked with statistical
analysis tools, and a Black Belt or other Six Sigma expert has been present to walk team
members through analysis and interpretation. While many organizations train process
owner
owner and his or her team will be familiar with the statistical controls that Six Sigma experts
have been using. Because of that, appropriate documentation via a control plan and
education regarding tools such as control charts might be necessary to ensure business
teams can maintain a process and identify when it is out of control and needs remediation.

At this time, Six Sigma teams might want to revisit the FMEA tool originally introduced in
chapter 13. Six Sigma teams initially use the Failure Modes and Effect Analysis to identify
potential failures in a process and causes of those failures. In chapter 13, we discussed how
the FMEA listed potential failure points and ranked them according to severity, occurrence,
and detection, calculating a total risk priority number.

At the end of the improve phase or beginning of the control phase, Six Sigma teams should
revisit the FMEA, noting what recommended actions were completed and recalculating risk
priority numbers for the improved process. There are two reasons for revisiting the FMEA.
First, the team is able to see that positive and significant changes have occurred because of
the solutions adapted during the improve phase. For any root cause that matched a solution
implemented, the team would hope to see a smaller risk priority number.

Second, an updated FMEA helps the team identify the next problem or root cause that
might be addressed. Remember, Six Sigma is a continuous improvement initiative. The team
might have implemented a solution and met an improvement goal, but further

234
 SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

improvements can always be made. Control is a time to review the process and suggest
possible improvements for future projects.

To facilitate continued success, Six Sigma teams should create a written control plan for the
process owner. The purpose of a control plan is to help the process owner and business
team track and respond to key performance indicators so that the process remains
improved. The control plan should be a concise, easy-to-reference document that tells the
business team when to monitor, how to monitor, what range of data is acceptable to the
monitor, and how to respond with corrective action if the range measured is not
acceptable.

Control plans can be spreadsheets, specialized digital documents, or hard-copy documents

posted at a work station. Common elements of a control plan include:

o Company, division, or department name

o Name of person who created the plan
o Date the plan was created
o Name of the person who last edited the plan
o Date the plan was last edited
o Project and/or process name or identifier
o Process owner
o List of process steps where control action is required
o CTQ or metric associated with each action required
o Limit specifications, or the acceptable range of measurements
o The unit of measurement
o The method of measurement
o The necessary sample size
o The frequency of measurement
o The person responsible for measurement
o Where the information is recorded
o Correction actions
o Associated policy and procedure documents

In discussing quality in chapter 8, we introduced the example of a company that makes

chocolate bars and noted that the amount of sugar in the chocolate bar recipe was critical

235
 CONTROL

improving customer satisfaction with a new chocolate bar product, they might have
implemented a solution that ensures the proper amount of sugar is added to the mixture at
the right temperature to incorporate the ingredient appropriately.

A control plan for the new chocolate bar solution might look something like the document
below.

Company: XYZ Sweets Control Plan Created by: Joe Black

Belt

Process: Sugar addition, raw goods Control Plan Created on: Jan. 4,
mixture 2012

Process Owner: Sue Processor

Process Step Addition of sugar to Heating of batch

batch

CTQ/Metric Total amount added to Mean temperature

batch during mixing

Limit specification LSL: 4.90 cups LSL: 105 F

USL: 5.10 cups USL: 110 F

Unit of measurement Cups Degrees F

Method of 6-cup sugar test bowl Read integrated digital

measurement thermometer on mixing
machine

Sample Size One batch 3 reading, 2 minutes

apart, during mixing

Frequency Every 2 hours Every 2 hours

Employee Mixer operator Mixer Operator

Record data in Mix operation log Mix operation log

spreadsheet spreadsheet

236
 SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

Corrective action Manually measure Turn off machine, waste

correct amount for inappropriately heated
current batch to allow batch, and report
for processing, calibrate temperature calibration
sugar disbursement issue to maintenance.
machine following SOP
100.54, test sugar
disbursement for first
batch after calibration
to ensure problem is
resolved. Report issue
to supervisor.

The above example control plan provides instructions for two specific steps in the process
with easy-to-understand measurement and monitor requirements. To reduce the chance of
errors, the Six Sigma team has even specified a special measuring tool for measuring the
sugar in the test batch so that every operator performing the monitor measures using the
same tool.

At the end of the control document, the team provides steps for corrective action. The first
step can be corrected by the operator, who has the ability to calibrate the machine him or
which
means the process has to be stopped so that someone can attend to the issue. Note that it
is always preferable, when possible, to build corrective action at the process level, such as
was done with the sugar measurement. This minimizes downtime, puts employees more in
control of the processes they own, and helps employees stay involved with the quality
process.

The control plan above assumes that manual measurements must be taken or recorded.
Optimally, Six Sigma teams should look for ways to automate measurements, which means
data can be continuously gathered and converted into statistical process controls such as

ructions for gathering the data. Instead,

employees and process owners can be instructed to review automated data or control
charts and take action if necessary.

237
 CONTROL

USL. LSL is the lower specification limit and the USL is the upper specification limit. These
are the upper and lower limits of the acceptable range.

In addition to providing a control plan, Six Sigma teams can implement specific visual
controls in a workplace to help business teams maintain a controlled process. Some of these
tools were covered in chapter 4 on Lean process management, including 5S. Other visual
controls teams might implement include signs, posted matrixes and instructions, auditing
boards that let teams keep track of individual or group performance over time, color coding,
and safety signals.

Standard operating procedures can often be distilled to visual representations on posters. A

coffee shop, for example, might provide employees with a visual representation of what
ingredients are used to create various complex drink flavors. Such a poster ensures that
employees can prepare drinks quickly while reducing errors in ingredient inclusion.

Other SOP visualizations might include safety procedures in a medical environment, such as
visual reminders for hand washing and short pictorial representations for how to operate
equipment such as hospital beds. In an office environment, pictorial instructions are found
on copy machines, where pictures indicate how paper should be loaded and visual gifs are
often displayed on LED screens to help employees remove jammed paper. These are some
examples that Six Sigma teams can follow when creating documents that will help business
staff accept ownership of an improved process and maintain the improvements made.

One of the most common methods Six Sigma teams use to monitor a process is the control
chart. A number of types of control charts exist, and Six Sigma experts must choose the
right control chart for the type of data and analytical purpose.

Now, w
process is out of control.

A basic control chart has the following elements:

A line chart of data with plot points for specific data points
An x-bar line representing the average of the data points

238
 SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

Lines above and below the x-bar line representing 1, 2, and 3 standard
deviations from the median in either direction
An upper control limit (UCL) line at 3 standard deviations above the median
A lower control limit (LCL) line at 3 standard deviations below the median

Above is an example of a control chart. The middle line, which is black, is the x-bar line. The
x-bar line is bounded by green lines on both sides, indicating 1 and 2 standard deviations
away from the mid-line. Those lines are bounded by orange lines on either side: at the top,
the upper control limit, and on bottom, the lower control limit. Between the various lines
are areas of the control chart, designated as C, B, and A going in either direction. These
distinct areas are important for understanding if a process might be out of control. They are
also called Zones 1, 2, and 3.

A control chart is best displayed using an automated reporting system or dashboard, where
process owners or responsible employees can view it as needed. If automated data
collection and control charting is not possible, then a business analyst can be tasked with
collecting data and presenting it in this format periodically, though periodic graphical
analysis is less likely to catch a problem of control within a process.

Eight tests exist that can quickly tell someone viewing a control chart if a process is out of
control.

Test one: A single point on the control chart appears outside of the upper or
lower control limits. If this occurs, process owners should take immediate

239
 CONTROL

action, because it is evidence of a major problem within a process. While there

is a very remote possibility that shifts outside of three standard deviations can
occur randomly, the likelihood is only 3 in 1,000.
Test two: Nine points in a row appear on one side of the center line. This
indicates that a change occurred in the process; if the process owner knows
what change caused the shift and it was intentional change, nothing needs to
be done and the control chart will right itself over time with the new data.
Otherwise, the process owner should investigate the process.
Test three: Six points on the control chart increase or decrease in a row,
indicating the process is becoming less or more efficient or is generating fewer
or more errors. Process owners should investigate unless there is a known
reason for the trend.
Test four: Fourteen points on the control chart in a row alternate moving up
and down. This could indicate variation in machines, employees, shifts, or over
correction.
Test five: Two out of three points in a row on the control chart are in the upper
A section or in the lower A section. This might indicate some type of special
cause creating sudden high variation.
Test six: Four out of five points in a row on the control chart are located in the
upper B section (or beyond) or the lower B section (or beyond). This can
indicate a problem of major causation or a shift problem similar to that of test
four.
Test seven: Fifteen points in a row are located within the C section above or
below the centerline. This can indicate that control limits are no longer relevant
to the process; if a team has improved variation of a process, they should
recalculate control limits to new parameters. Alternatively, this might occur
temporarily when short-term variation is high or low relative to the rest of the
points on a control chart.
Test eight: Eight points in a row on the control chart are located on either side
of the center line, but none are located in the C section above or below the line.
This could indicate an issue of mixed resources or processes; a team might think
they are measuring a single process when they are actually measuring two
process, for example. Alternatively, it could indicate a major difference in
processing for two employees or teams.

these
-versed in
240
 SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

statistical process control to learn about the eight tests, and business teams with the
benefit of control charts can spend more time working on production or correcting issues
and less time collecting and documenting measurements.

of meeting customer requirements. We touched on this concept in unit 1: controlled

and
the outputs center around a customer requirement. This is why both control limits and
specification limits are important.

Consider the example used earlier in this chapter about sugar in the chocolate mix. The

control chart to show that the process is in control if the measurements range from 3.5 to
3.6 cups of sugar per batch, but the process owner should know that those measurements

To understand how a process is performing against specification limits, Six Sigma teams can
calculate sigma level and process capability.

Sigma level is the number of standard deviations between the current process center, as
measured by the median, and the nearest specification limit (not control limit.) The
equation for sigma level is the smaller of the following calculations:

For example, if a process has an USL of 5 and a LSL of 3, a standard deviation of .25 and a
median of 4.2, then you would calculate from the USL, since the median is closer to the USL.

241
 CONTROL

Process capability is calculated by dividing the sigma level by 3. In the case of the example,
the capability is 1.06.

Process capability is denoted as Cpk. A process capability of 1.33 is equal to a sigma level of
4, which is what most experts agree is the minimal level at which most customers will be
satisfied. Under statistical process control, many organizations aim for a process capability
of 2.0 with minimal acceptable process capability at 1.5.

When Six Sigma teams deem improvements and the related process to be capable and in

they should take time to celebrate and reflect on the outcome of the project. This is usually
done following the final tollgate review with a sponsor or champion, and can be a quick
meeting to close loose ends, recognize the work done by the team, and discuss lessons
learned within the process.

The celebration and reflection meeting is also a great time for team members to bring up
ideas for possible improvement projects. While improvements and the related problems
and causes are still fresh on the minds, they can effectively brainstorm ideas for
next steps. As with any brainstorming session, no idea should be automatically held off the
table because it seems silly, would be too hard to implement, is too costly, or seems too big.

information that the Black Belt can later share with Six Sigma leadership panels.

The team has calculated the performance and capability of the new process
The team has written a control plan and communicated it to the process owner
The team has created a monitor for the process, either through procedures for
manual data collection or automated generation of control charts
The team has provided the process owner and business team with all tools and
information required to maintain improvements
The sponsor, champion, or executive leadership has been informed about the
state of the improvements
The team met to reflect on the project and generate a list of ideas for future
improvements
242
Graphical analysis is a critical part of the Six Sigma approach. Whenever data or ideas can be
displayed in a visual format, obvious data trends can be quickly identified and
communicated. Visual analysis is extremely helpful when presenting ideas to auxiliary staff
or executives, who might not be trained in intermediate or advanced statistics. The Black
Belt or other Six Sigma experts have the ability to tell the same story that statistics tell, but
in a format that can be understood by anyone. Visual depictions are tools that make it
easier for Six Sigma professionals or sponsors to retell the story to others, which can be
helpful for training, building cultural buy-in for a process change, or even resource requests.

ussed a number of graphical analysis

tools, including:

Pareto charts in chapters 5 and 14

Run charts in chapter 13
Box plots in chapter 14
Introduction to control charts in chapter 16
Introduction to scatter diagrams in chapter 14

16, you learned about the components of an X-bar control chart and how to tell if a control

243
 INTERMEDIATE GRAPHICAL ANALYSIS

steps for manually creating an X-

a data analysis add-in for various versions of Excel. The add-in provides functionality that
will become relevant in the next few chapters.

As with many other areas of the Six Sigma methodology, it can be easy to go overboard
when dealing with graphical analysis particularly when presenting information to others.

more complex. Often, though, that means spending non-value added time creating one
visual tool when an easier tool would do the job. If you recall from chapter 4, this would be
an instance of over-processing to avoid.

Consider the two charts below to understand when over-processing might be occurring with
regard to graphical analysis.

244
 SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

Both charts are a graphical representation of the same information. The first is a simple bar
chart that was created within a few clicks in Excel. The second is a Pareto chart, also created
in Excel but requiring many more clicks of the mouse. In this particular case, a Pareto chart
is not necessary to convey the required information. The data is diverse enough that anyone
can see from looking at the bar chart that sales person E accounts for a lot of the errors. In
this particular case, a Pareto chart is
created if extra work is required.

conclusion is obvious. Looking at the raw data below, you can quickly see that employee E
has triple the errors as the next-highest employee. Even when the raw data provides for
easy analysis, however, many Six Sigma experts do take the time to create a graphical
representation for the purpose of presentations. A graph is more visually appealing and
quicker to read than a data table.

245
 INTERMEDIATE GRAPHICAL ANALYSIS

Sales
Employee Errors
A 10
B 5
C 2
D 9
E 34
F 7
G 3

are not so obvious to note. More complex analysis might be required to discover outliers,

always correct. This is especially true when dealing with the relationship or correlation
between factors in a process.

As a Six Sigma expert, it takes time to be able to quickly choose which type of graphical

for graphical analysis relies equally on the questions being asked as much as it does on the
data itself. In the example above, the basic bar chart is a good choice for a team that wants
to know where errors might be coming from. If the team was comparing errors to shift
times being worked, the bar chart would not be helpful.

not helpful or could be considered muda of over-production

when dealing with analysis. At the analysis stage of a DMAIC project, Black Belts and other
Six Sigma experts often work with numerous graphical analysis tools as they attempt to
understand data. The need to view data in various ways is one reason statistical analysis
software is helpful: such software takes some of the manual work out of creating these
charts and graphs. Trial-and-error work with all types of Six Sigma analysis tools also helps
you learn to identify which tools fit each situation best.

far, covering
benefits, limitations, and how to create the tools in Excel without statistical analysis
software when necessary.

246
 SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

One of the biggest benefits of bar charts is that they are recognizable and easy to read.
Almost any employee in a company can glean information from a properly formatted bar
chart without instruction or guidance, making them a great choice for general use
presentations and training materials. Bar charts are also easy to create, so they are a
preferred method of illustration in presentations and reports. Other benefits include:

The ability to summarize large data sets in a simple visual format

The ability to clarify trends
The fact that most people can easily estimate important values on a well-formatted
bar chart
The ability to visually check data and identify areas where data might be skewed
The ability to easily display data sets that range above and below zero on the same
chart

Bar charts typically require nominal or ordinal data data that is classified according to

Pareto-style chart so the reader can see instantly how the population relates to the
categories. Ordinal data usually involves its own logical order for presentation. For example,
if individuals are asked to rank satisfaction with a product as very satisfied, somewhat
satisfied, neutral, unsatisfied, and very unsatisfied, then you usually would not want to
change the order of that presentation on a bar chart.

Bar charts are not without limitations. Depending on what is being presented, additional
narrative might be required to explain the chart. Bar charts also often fail to reveal key
fic design of the chart, and they rarely

manipulate colors, order, and layout of a bar chart to influence the message that your
audience takes away. Six Sigma experts should always choose the best format for displaying
fundamental truths, but should never format a chart to create an impression that might not
be in keeping with the statistical data behind the graphical representation.

Create your own bar charts in Excel by starting with data tables of relevant nominal or
ordinal data. Copy the data table below into Excel to practice bar chart creation, or use data
relevant to your own processes or projects. The sample data table provides the total

247
 INTERMEDIATE GRAPHICAL ANALYSIS

number of phone calls experienced by a regional customer service call-center team for each
hour in a particular day.

Phone
Hour calls
8:00
AM 78
9:00
AM 89
10:00
AM 107
11:00
AM 118
12:00
PM 149
1:00
PM 147
2:00
PM 105
3:00
PM 90
4:00
PM 97
5:00
PM 85
6:00
PM 178
7:00
PM 198
8:00
PM 145
9:00
PM 57

1. Highlight the columns that contain the data you want to chart as well as the labels
for that data. In this case, the data is the number of calls and the labels are the
hours.

248
 SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

2. Select Insert > Chart > Bar Chart

3. For this example, the simplest form of bar chart is appropriate.

4. Use Excel formatting tools as desired to customize the chart title, colors, and labels.

249
 INTERMEDIATE GRAPHICAL ANALYSIS

Most versions of Excel offer several types of bar charts, and you can experiment with these
various formats to see how they display your data. Some notes about common formats are
included below.

Technically, what is commonly referred to as a bar chart

bar chart here is a column chart. The visual columns representing each data category rise
vertically. A literal bar chart displays the same information horizontally. Most versions of
Excel let you choose between the two displays. The information from the example data
table is presented below in a horizontal bar chart.

250
 SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

The charts displayed so far in this chapter are all 2-dimensional. Excel can also help you
create 3-dimensional charts to add visual appeal to a presentation. 3-D charts are helpful if
you are presenting a number of similar-looking charts in a row, as it helps to differentiate
between information in your presentation. It might also be helpful to create more visually

same data used above in a 3-D chart.

251
 INTERMEDIATE GRAPHICAL ANALYSIS

Stacked bar charts let you display the total nominal or ordinal data for each category while
also breaking that information into color coded categories. For example, if the Six Sigma
team analyzing phone calls per hour wanted to display data that included how many calls
were handled by separate teams within a department, the bar chart might look something
like the image below.

252
 SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

To create a stacked bar chart, you must have ordinal or nominal data that is broken into
categories. The categories must be the same for each data set so that the total of the sub-
category numbers equals the total of the main category. For example, if you have three
bags of marbles, you might have data that says:

Bag Number of marbles

1 10

2 3

3 15

Upon further analysis, you realize you have marbles in three colors: red, blue, and yellow.

Bag 1 Bag 2 Bag 3

Red 7 1 5

Yellow 3 1 8

Blue 0 1 2

note that the totals of the subcategories under each bag add up to the amount of marbles
in each bag. Following these guidelines including all subcategories for each section and
ensuring the subcategories total correctly helps you create accurate stacked bar charts.

Examine the data table below, which displays the breakdown of calls by teams in the call
center.

253
 INTERMEDIATE GRAPHICAL ANALYSIS

Phone
Hour calls Team A Team B
8:00
AM 78 35 43
9:00
AM 89 40 49
10:00
AM 107 57 50
11:00
AM 118 90 28
12:00
PM 149 42 107
1:00
PM 147 77 70
2:00
PM 105 57 48
3:00
PM 90 45 45
4:00
PM 97 72 25
5:00
PM 85 64 21
6:00
PM 178 98 80
7:00
PM 198 89 109
8:00
PM 145 77 68
9:00
PM 57 54 3

1. Highlight the cells containing the data you want to include in your stacked bar
chart as well as the cells containing the data labels.

254
 SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

2. Select Insert > Charts > Bar Charts > Stacked Bar Chart

3.

The pie chart is another visual tool that almost any employee in a business environment will
be familiar with, making it a good choice for displaying certain types of information. Ordinal
and nominal information can also be displayed in a pie chart with the main purpose of
visually representing how each category relates to the whole. Conventional wisdom says to

be the case. Excel converts values to percentages to create pie charts. Consider the pie
chart below, which graphs the numbers 8, 5, and 4. Excel considers each of these numbers
255
 INTERMEDIATE GRAPHICAL ANALYSIS

raw data 8/17.

Benefits of pie charts include:

The ability to summarize large amounts of specific data types in a visual

format.
Simplicity; the pie chart is one of the simplest types of graphs
The ability to identify obvious problems with data or calculations
Pie charts usually require very little extra explanation when labeled
correctly
Pie charts display the relevance of subset data within a total data set

The fact that pie charts are used constantly in business environments is both an advantage
and disadvantage for Six Sigma teams and experts. When data can be depicted in a pie
chart, the Six Sigma expert usually has to create less narrative to get a potential point

statistical representation might because business employees are so used to seeing them.
Other disadvantages of pie charts include the fact that they can be manipulated in much the
same way as bar charts, they usually fail to easily display changes over time, and it can be
difficult to visualize exact values when presented with a pie chart.

256
 SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

Practice creating a pie chart in Excel using the following data table, which records how
many minutes are spent on average for each step of a process. You can also use data
relevant to your own process or project to create a practice pie chart if desired.

Process
Step Minutes
A 14
B 18
C 41
D 64
E 12
F 4
G 3
H 7
I 23
J 18

1. Highlight the column of data you want to chart as well as the data labels. In the
example, we are charting the number of minutes and the process numbers are the
data labels.

2. Click Insert > Chart > Pie Chart

257
 INTERMEDIATE GRAPHICAL ANALYSIS

3.

From the pie chart created for our sample data, we can see that approximately half of the
total process time is related to only two steps in the 10-step process. We can also quickly

draw any conclusions about actual time from only this chart.

As with bar charts, you can take advantage of Excel options for creating three-dimensional
pie charts. A particularly helpful function is the Pie of Pie function in Excel, which lets you
create an overall pie chart and carve out a second pie chart to delve deeper into certain
areas of the larger structure. Consider a Pie of Pie chart below, which represents the same
data from the pie chart above.

258
 SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

The original pie chart depicted 10 steps. The illustration above shows a large pie chart with
a secondary chart. The large pie chart appears to show 7 steps but this pie chart shows
the same information as the one above. However, four steps from the process are called
out in a smaller chart so you can see how they relate to each other better. Those four steps
combined make up the 7th, dark-blue pie piece in the larger pie. This makes it easier to
pinpoint relevance of smaller data categories within the whole or to emphasize a certain
area of the chart.

You do have to be careful,

based solely on the size of the pie pieces. The gray pie piece in the secondary chart above is

longer. The smaller pie should be considered a large-scale view of a small piece of the
overall graph. In this case, that small piece takes into account four of the steps.

We intro

create a scatter diagram in Excel with any analysis add- l use an add-
on for Excel to conduct some statistical analysis while creating scatter diagrams.

Scatter diagrams are beneficial because they can help teams visually see the relationship
between two factors in a process. Does temperature decrease over time

259
 INTERMEDIATE GRAPHICAL ANALYSIS

productivity increase with his or her salary? These are examples of questions that might be

Two variables can be closely related

without one causing changes in the other.

Scatter diagrams typically help teams see whether there is no correlation, weak correlation,
or positive or negative correlation. Positive correlation occurs when variable 2 is related to
an increase in variable 1, or vice versa. Negative correlation occurs when variable 2 is
related to a decrease in variable 1, or vice versa. For example, it is commonly noted that the
rate of crime and the average per-capita income of geographic areas in the United States
are related. Areas that demonstrate higher crime rates often also demonstrate lower
income statistics. The statement can be made the other direction: areas that demonstrate
lower income statistics often demonstrate higher crime rates. That is not to say that either
of these things necessarily causes the other.

Drawbacks of scatter diagrams are that they are not as familiar to business employees as
ve an
understanding of statistical analysis will mistake correlation for causation, which can lead to
incorrect decisions. Black Belts and other Six Sigma experts must be cognizant of these risks
so they can appropriately explain scatter diagrams and provide further explanation when
necessary.

Use the data table below, which includes the average number of errors per hour a process
creates as output is increased per hour to create a scatter diagram in Excel. Copy the
information from the table below into Excel, or use data from your own process or project if
desired.

Parts Per
Hour Defects
100 3
150 4
200 5
250 5
300 7
350 7
400 9
450 8
260
 SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

500 10

1. Highlight the data you want to chart. To create a scatter diagram, you will need to
highlight two sets of data that you want to compare.

1. Select Insert > Chart > Scatter

2.

261
 INTERMEDIATE GRAPHICAL ANALYSIS

From the scatter diagram created, we can see that the two variables production volume
and errors
production actually causes om this
graph. This assumption could be further tested using more advanced statistical techniques.

always the case. Six Sigma experts should always test correlation assumptions to ensure
they are correct statistically before presenting a scatter diagram to others as an illustration

The X-bar control chart plots the mean of a sample over time or the mean of samples
taken over time, in the case of an active process. X-bar control charts are one of the most
frequently used control charts. Some control charts can also be created in Excel using these
steps; more advanced control charts require statistical analysis software.

1. Create the following template in an empty Excel spreadsheet.

262
 SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

2. Enter your data points in column A.

3. In cell G1, enter the formula =STDEV(A2:A15)
a. The cell references A2:A15 should be edited to correspond to your actual data
list.
b. This calculates the standard deviation of your data set.
4. In cell G5, enter the formula =AVERAGE(A2:A15)
a. The cell references A2:A15 should be edited to correspond to your actual data
list.
b. This calculates the mean of your data set.
5. In cell G2, enter the formula: =G5+3*(G1)
a. This calculates an upper control limit that is three standard deviations above
the mean.
6. In cell G3, enter the formula: =G5-3*(G1)
a. This calculates a lower control limit that is three standard deviations below the
mean.
7. In the UCL column (B), direct Excel to create a column of numbers where every number
is equal to the upper control limit by copying =$G$2 into each cell in a row that has an
X-bar data point in column A.
8. In the LCL column (C), direct Excel to create a column of numbers where every number
is equal to the lower control limit by copying =$G$3 into each cell in a row that has an
X-bar data point in column A.

263
 INTERMEDIATE GRAPHICAL ANALYSIS

9. In the Mean column, direct Excel to create a column of numbers where every number is
equal to the mean by copying =$G$5 into each cell in a row that has an X-bar data point
in column A.

(Practice steps 1 through 9 with the following data.)

X-Bar Data Points

The result should be a spreadsheet that looks just like the one below.

10. Create a control chart by highlighting the data in columns A through D.

264
 SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

11. Select Insert > Charts > Line Chart > 2-D Line Chart with Markers

12.
265
 INTERMEDIATE GRAPHICAL ANALYSIS

13. Alter the bounds for the axis to remove excess white space in your control chart. In this

14. Format the chart as desired for color, titles, and data labels. You now have a very basic
control chart.

nes that denote one and two

standard deviations above and below the mean. In chapter 16, you learned that those lines

266
 SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

help you apply the tests that determine if a process is in control or not. You can easily add
those lines by following the steps below.

Add Standard Deviation Lines

1. Add calculations to your spreadsheet for:

a. The mean plus one standard deviation
b. The mean plus two standard deviations
c. The mean minus one standard deviation
d. The mean minus two standard deviations
2. Add four data columns, one for each of the calculations in step one. Copy the
numbers from each of the calculations down the columns as you did with the upper
and lower control limits and the mean in the original instructions above.
3. The spreadsheet should now appear similar to the image below.

4. Highlight all of the information in columns A through H.

5. Select Insert > Charts > Line Chart > 2-D Line Chart with Markers.
6. Change the Y-Axis to an appropriate range.
7. Edit the background standard deviation lines to be less intrusive by clicking on each
and first selecting the gradient line option.

267
 INTERMEDIATE GRAPHICAL ANALYSIS

8.
the background lines.

9. Edit the final chart as desired with titles and labels.

268
 SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

This process might seem very time consuming, but it only takes a few minutes once you are
well-versed in the elements of a control chart. a good idea to know how to create

analysis software.

You can complete these calculations by hand, but that can be time consuming and require
extremely advanced statistical skill sets. We will start with manual calculations because it
helps to understand the
that complete most of the calculations. Not only does statistical analysis software minimize
the time it takes to conduct such analysis, but it also reduces the chance of calculation
errors and increases the accuracy of analysis and conclusions offered by Six Sigma experts.

For lessons covered in the next few chapters, you can rely on the free data analysis tool
available from Microsoft for Excel. To add this free tool in Excel 2013 and higher, follow the
steps below.

269
 INTERMEDIATE GRAPHICAL ANALYSIS

1. Select File

2. Select options

270
 SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

3. Select Add-ins

271
 INTERMEDIATE GRAPHICAL ANALYSIS

4. Locate the Analysis ToolPak under inactive application add-ins. Note: if the Analysis
ToolPak is located under Active Application Add-ins, then it is already active and you

5. -

6. In the new dialogue box, ensure the option for Analysis ToolPak is checked.

272
 SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

7. Click OK
8.
menu bar and looking for the Data Analysis option.

-In opt
when your version of Excel was installed. If you are using Excel that was installed from a

custom install of just the Analysis ToolPak file.

Note that there are two files with the Analysis ToolPak name. The file called Analysis
ToolPak VBA is not the file that provides analytical capability.

The Analysis ToolPak is also available in earlier versions of Excel. In 2003 and earlier versions
-in menu item under Tools. In 2007, the Add-in option is found
by selecting the Office button and then Excel options. In 2010, select File, Options, and
Manage Add-ins.

273
normal probability curve to draw inferential statistics. Now is a good time to know the
difference between descriptive statistics and inferential statistics.

Descriptive statistics are numbers used to describe a set of data. Given a set of data about
the employees in a company, for example, someone might say,
, yees live

they simply discuss the nature of the data at hand. Averages, means, and even standard
deviations associated with that data are all descriptive in nature.

Inferential statistics involve calculations and resulting numbers that draw more general
conclusions from data. For example, if you have data about a sample of 150 employees in a
city not all the employees from a single business, but random employees from different
companies you might be able to use inferential statistics to draw conclusions about all the
working individuals in that city based on the data in your sample.

Descriptive statistics can relate to population data or sample data. Inferential statistics are
used to draw conclusions about the population from sample data. If you randomly select 20
people from an overall population and record their height, then the mean of the recorded
heights is a descriptive statistic about the sample size. Descriptive statistics about the
sample are used in calculations for inferential statistics about a population.

274
 SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

Probability is the likelihood that a certain event or outcome will occur given a specific set of
data or a specific action. If you flip a coin one time, the probability of it coming up heads is 1
in 2, or 50 percent.

When discussing basic probability, we talk about outcomes and events. The outcome is the
result of one trial: you flip the coin once or you randomly select a single item from the
sample. The event is a specific outcome or outcomes. The coin coming up heads is an event.
A person measuring 5 feet tall is an event if you are talking about selecting from a
population or sample.

Basic probability can be calculated by the formula:

In the case of the coin, there are two possible events, or outcomes, that can occur when
you flip it. Only one of those events is coming up heads. Therefore, ½ = .5, or 50 percent.

In the case of rolling a six-sided die, there are six possible events. You can roll a 1, 2, 3, 4, 5,
or 6. The probability of rolling a 5 is 1/6, or .167. The probability of rolling either a 4 or a 5,
however, is 2/6, or .33.

While probability measurements can be expressed as a percentage, they are calculated and
usually communicated in statistics as a decimal point. Since probabilities are presented as
fractions, the lowest probability possible is 0 and the highest is 1. A probability of 0 means
an impossible event; a probability of 1 means a certain event.

Twenty marbles are placed in a bag. They are all the same size, but are different colors as
follows:

275
 NORMAL PROBABILITY DISTRIBUTIONS

- 3 red
- 5 black
- 8 blue
- 4 green

What is the probability that a green marble would be pulled from the bag?

4 / 20 = 0.2

What is the probability a blue or a black marble would be pulled from the bag?

13 / 20 = 0.65

What is the probability a yellow marble would be pulled from the bag?

0 / 20 = 0

Note: This is an impossible task, because there are no yellow marbles in the bag.

are somewhat similar. If a process produces 100 products an hour and two of those
products are defective, the probability of receiving a defective product within that hour is

products that were produced that hour to know that 2 out of 100 are defective. That
information is also only useful for stating probabilities with relation to that particular hour

Instead, Six Sigma experts randomly sample the population of all outputs of a process for a
certain amount of time. They create descriptive statistics about that sample and use the
descriptive statistics to further study and understand the nature of the data they are
working with. Once they know the nature of the data, they can work with probability
distributions and curves to determine probability via calculations and statistical analysis. In

276
 SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

histograms. Histograms are another graphical analysis tool. Technically, histograms are bar

or population are divided between the bins and are graphed accordingly.

Histograms are used to analyze continuous or variable data that is finite in nature or comes
from population sampling. Histograms help Six Sigma experts understand how the spread,
or distribution, of the data is shaped and where the center of that data might be. The center

always located at the center of a histogram. A histogram also shows the range of the data
the variation between the highest value of the data set and the lowest value of the data set.

Consider the two histograms below. In the first, the data elements are concentrated toward
the center. The histogram approximates what is called the bell curve, or normal curve. This
shape is likely to indicate that the data used to create the histogram is normal data, which
will be discussed further later in this chapter.

The second histogram depicts data that is concentrated at either end of the range. When
two specific concentrations, or humps, are noted on a histogram, this is called bi-modal
distribution. Often, when you see this shape of data, it means that you are measuring two
processes that you believe are a single process. For example, if two different people are
performing a piece of work, they might have different outcomes that result in two different
humps on the histogram. If bi-modal data does come from a single process, then the data is
not normal.

277
 NORMAL PROBABILITY DISTRIBUTIONS

Histograms can also have several other common shapes, especially when presenting non-
normal data or data that has an error or issue. If the histogram appears very random, as in
rement
process. The measurement might not be adequate to capture the true picture of the
process, or the definition of the operation and how to collect data might be at fault. The
same is true if the histogram depicts an almost even distribution of data elements among
bins.

278
 SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

A skewed distribution, seen in the second diagram above, features data that is clustered at
only one end of the graph. Skewed histograms are common with certain types of data,
-off time. For example, in a call center, you
might measure the time it takes representatives to complete a certain type of phone call. A
natural end-point for that data is 0 it will always take more than 0 seconds to complete a
phone call. sually, it will
in the lower end of the range. Perhaps this particular phone call follows a script and the
majority of the calls end at the 7 to 8-minute mark. A histogram for length of calls might
look something like the chart below.

279
 NORMAL PROBABILITY DISTRIBUTIONS

Consider the basic example below to understand histograms further.

A local distribution company wants to understand more about how well delivery drivers are
meeting estimated arrival times. The team pulls a random sample of deliveries from a

time of delivery and the actual time of delivery. They subtracted the actual time from the
estimated time to determine how far off one way or the other the deliveries were. The
result was the following data table.

Difference in
Estimate Actual minutes

8:00 AM 7:55 AM -5

8:30 AM 8:20 AM -10

8:50 AM 9:07 AM 17

9:15 AM 9:23 AM 8

9:45 AM 9:46 AM 1

10:05 AM 10:03 AM -2

10:30 AM 10:33 AM 3

10:55 AM 11:05 AM 10

11:30 AM 11:27 AM -3

12:45 PM 12:32 PM -13

1:30 PM 1:35 PM 5

1:45 PM 1:57 PM 12

2:20 PM 3:01 PM 41

3:00 PM 3:48 PM 48

3:30 PM 3:05 PM -25

280
 SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

8:50 AM 8:45 AM -5

9:15 AM 9:58 AM 43

9:45 AM 10:21 AM 36

10:05 AM 10:09 AM 4

10:30 AM 10:03 AM -27

1:45 PM 1:46 PM 1

2:20 PM 2:54 PM 34

3:00 PM 3:04 PM 4

3:30 PM 3:23 PM -7

11:30 AM 11:16 AM -14

To create a histogram, the team sorted the selection by difference in minutes and
discovered the range as -27 to 48. The team decided to create a histogram from -30 to 50
with a bin size of 10. The resulting data is as follows:

Bin Frequency

-30 0

-20 2

-10 3

0 5

10 8

20 2

30 0

40 2

50 3

281
 NORMAL PROBABILITY DISTRIBUTIONS

To create the above table, the team simply counted how many data points fell in each bin.
The data points -27 and -25 fell into the bin -20 (to -29); that is 2 data points. Eight data
points fell into the 10 bin. When charted on a bar chart, this information creates a
histogram.

When dealing with histograms, the number of bins used is important. If you have too few

data set in question, you also lose informative functionality and your graph, rather than
appearing as a rough curve, becomes something that looks like a comb. Usually, the more
data you have, the more bins you can include on your histogram.

Consider the following four histograms that all represent the same set of data. The data
features 150 random numbers ranging from 38.62 to 62.89.

282
SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

283
 NORMAL PROBABILITY DISTRIBUTIONS

The bin sizes get progressively smaller as the number of bins increase. In the first chart, bin
sizes are 5; in the last chart, bin sizes are 0.5. For the purposes of understanding what type
of data is represented in this histogram, the second or third charts are probably the most
284
 SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

in this chapter.

in Excel
using the Analysis ToolPak that was discussed in the last chapter.

1. Open Excel and copy the following data table into a blank workbook.

56.1

38.2

47.1

48.1

60.1

45.8

33.4

49.2

53.1

41.8

19.2

49.3

49.0

61.8

42.4

53.2

285
NORMAL PROBABILITY DISTRIBUTIONS

61.6

20.8

55.2

57.5

48.8

28.9

33.8

53.2

58.0

47.3

51.4

61.5

58.9

60.4

30.7

52.5

40.7

44.8

54.6

61.6

31.0

52.7

47.5

286
 SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

26.3

71.3

61.8

51.9

43.9

50.0

47.4

61.9

51.8

50.4

61.0

2. Copy the following bin designations into the second column in the same worksheet.

10

15

20

25

30

35

40

45

50

55

287
 NORMAL PROBABILITY DISTRIBUTIONS

60

65

70

75

3. Select Data > Data Analysis

4. Select Histogram and click OK.

5. Click in the Input Range box, and then highlight the cells for the input range (the 50
data cells you copied from above).
6. Click in the Bin Range box, and then highlight the cells that include the bin data you
copied from above.
7. Check the New Worksheet Ply: option.
8. Check the Chart Output option.
9. Ensure that your Histogram dialog box appears as below.

288
 SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

10. Click OK.

Excel automatically generates a bin and frequency table as well as a histogram graph in a
new worksheet. The histogram design defaults to a standard bar graph, but you can create
the look of the histograms previously shown by changing one setting on the design.

289
 NORMAL PROBABILITY DISTRIBUTIONS

1. Click on the histogram.

2. Select Design from the top menu.

3. Under Quick Layout, choose the layout where the bars are stacked next to each
other rather than with space between them.

Your result should look like the graph below.

290
 SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

Experiment with your own data, creating histograms with various bin sizes to begin to
develop an understanding of how data converts to this type of chart.

Normal distribution, also called Gaussian distribution, is probably the most important
distribution related to continuous data from a statistical analysis standpoint. A normal, or
Gaussian, distribution is depicted below.

291
 NORMAL PROBABILITY DISTRIBUTIONS

Normal data is shaped symmetrically surrounding the mean, represented above by the x-
bar line. A normal curve is beneficial for determining the probability that a given data point
in a population will fall inside a certain range within the distribution.

In a perfect normal distribution, 68.26 percent of all data points fall within plus or minus
one standard deviation from the mean; this area is represented by the red arrow on the
visual above. The blue arrow, which covers plus or minus two standard deviations from the
mean, indicates the area under the curve that includes 95.46 percent of the data points.
The purple area, which covers plus and minus three standard deviations from the mean,
indicates the portion of the curve that covers 99.73 percent of the data. Less than 1 percent
of data will fall under the curve outside of three standard deviations on either side.

Distributions even normal distributions vary a bit. To determine the exact probabilities
of various data points, advanced statistics are required; Excel and other programs perform
the calculations for you, making it easier to conduct analysis. Before we discuss probability

place.

Testing whether data is normal is critical to many steps in statistical analysis, because the
results of many tests can be invalid if
The most basic form of many of these tests are designed to work with normal data.

292
 SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

how visual inspection can help you determine whether data is normal or
not. The histogram below, taken from earlier in the chapter, presents what seems to be
obviously normal data. Data points converge around the center, and the histogram is
roughly symmetrical in nature.

But what about this histogram, which is the one created in the practice exercise in the
section on creating histograms in Excel?

293
 NORMAL PROBABILITY DISTRIBUTIONS

histogram shapes previously discussed. It seems to have elements of the normal curve, but
it could also be skewed data. This is why we perform statistical tests to determine
normality.

evious

the same data from the section above, which is:

19.2

20.8

26.3

28.9

30.7

31

33.4

33.8

38.2

40.7

41.8

42.4

43.9

44.8

45.8

47.1

47.3

294
SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

47.4

47.5

48.1

48.8

49

49.2

49.3

50

50.4

51.4

51.8

51.9

52.5

52.7

53.1

53.2

53.2

54.6

55.2

56.1

57.5

58

295
 NORMAL PROBABILITY DISTRIBUTIONS

58.9

60.1

60.4

61

61.5

61.6

61.6

61.8

61.8

61.9

71.3

Creating a histogram using the Analysis ToolPak generates a chart and a data table, as seen
below.

296
 SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

Use the Descriptive Statistics option in the Analysis ToolPak to quickly generate descriptive
statistics for your data set.

1. Select Data > Data Analysis > Descriptive Statistics

2. Click OK
3. Click in the Input Range box and select your input range using the mouse.

297
 NORMAL PROBABILITY DISTRIBUTIONS

4. In this case, the data is grouped by columns. In most statistical analysis, that will be
the case, but if you have data grouped by rows, you should change the Grouped By
selection.
5. Select to output information in a new worksheet.
6. Ensure at least the Summary statistics box is checked. You can also check the
Confidence level for mean and the Kth largest and smallest boxes, though that
hi-Squared Goodness-of-Fit test, which is the test
we are running to test for normality of the data. If you check these extra boxes,
be using at
this time.

298
 SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

7. Click OK.
8. Excel returns descriptive summary statistics for your data set.

299
 NORMAL PROBABILITY DISTRIBUTIONS

For the Chi-Squared Goodness-of-Fit test, you will need to note the sample size (or count),
the same standard deviation, and the sample mean. The sample size is the number of items

was, you can refer to the count listed in the descriptive statistics. The other two figures are
taken from the descriptive statistics above:

- Sample mean: 48.778

- Sample standard deviation: 11.444

The Chi-Squared Goodness-of-Fit test is actually a hypothesis test. That means you are
testing the data with regard to a null hypothesis and an alternative hypothesis. The two
hypotheses for the Chi-Squared Goodness-of-Fit test are:

Null hypothesis: The data is normal.

Alternative hypothesis: The data is not normal.

If one is not true, then the other is. In statistical terms, we talk in terms of accepting or
rejecting the null hypothesis. If we reject the null, we accept the alternative.

Basically, the Chi-Squared Goodness-of-Fit test takes the number of samples in each bin on
the histogram and compares that to the number of samples you might expect to find in
each bin given a normal curve. Using the actual number of samples in each bin and the
expected number of samples, we can calculate what is called the Chi-Square Statistic in
Excel. That number then lets us calculate a p-Value. In this case, it is the size of the p-Value
that lets us decide whether to accept or reject the hypothesis that the data is normal.

For the purpose of the Chi-Squared Goodness-of-Fit test in this situation, if the p-Value is
greater than 0.05, we will accept the null hypothesis that the data is normally distributed.

The Observed Bins

Having created a histogram via the Analysis ToolPak, you already have access to the
observed bin distribution. That information is housed in the data table Excel creates to
make the histogram.

300
 SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

Bin Frequency

10 0

15 0

20 1

25 1

30 2

35 4

40 1

45 5

50 11

55 10

60 5

65 9

70 0

75 1

More 0

We can use statistics related to the normal curve to calculate how we might expect bins to
behave given the median and standard deviation of our sample.

To give you an idea of what is going on with the statistical calculations involved in
determining expected size of bins, consider the graphic below.

301
 NORMAL PROBABILITY DISTRIBUTIONS

This graphic roughly depicts the bins from our histogram drawn on the normal curve.

possible to determine the area under the curve within a specific bin. Ultimately, that is done
by calculating the total area and subtracting portions.

We begin with a calculation known as the Cumulative Distribution Function, or CDF. The
CDF measures the total area under a curve to the left of the point we are measuring from.
For example, the total area under the curve above that is to the left of 45 is 50 percent.

the CDF for each individual bin. For example, the CDF for the bin located between 40 and 45
would equal the CDF of 45 minus the CDF of 40.

One problem with this rough depiction is that the curve drawn above centers on 45, and we

account for the slight shift.

Excel can calculate CDF with the formula:

=NORMDIST(x value, Sample Mean, Sample Standard Deviation, TRUE)

Set up the tables for calculating the CDF of each bin by copying the bin designations onto
the descriptive statistics worksheet that Excel previously created for you and creating two
columns, one for total CDF and one for bin CDF. Use the image below as an example.

302
 SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

Enter the formula for calculating CDF into column E, referencing the same mean and
standard deviation for each row and using the numbers in D as X.

303
 NORMAL PROBABILITY DISTRIBUTIONS

For the first row in our case, the bin marked 10 -- the bin-only area is equal to the CDF
-only area is
the CDF minus the CDF for the bin designation above. So, you would enter =E2 in the first
data row for column F. The second data row would be calculated as E3-E2; the next would
be E4-E3, and so forth. The result is the percentage of the curve in each bin.

Calculating the expected number of samples in each bin is as easy as multiplying the
percentages of each bin by the sample size. Again, you can see from the descriptive
statistics that the count for this set of data was 50.

304
 SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

To calculate the Chi- in each

bin and the actual or observed number. Copy the observed numbers over from your
histogram worksheet.

305
 NORMAL PROBABILITY DISTRIBUTIONS

Apply the following formula to each row:

Calculate the final numbers for each row as desired in Excel.

Add up the final numbers to get the Chi-Squared statistic, denoted by X2. For our example,
X2 is 16.7333.

To use the Chi-Squared statistic to find the p-Value, we also need one more item for the
Excel formula to work: we need what is called the degrees of freedom.

Degrees of freedom = #bins 1 - #calculated parameters

We have 14 bins. The parameters we used to arrive at the Chi-Squared statistic that we
calculated from our sample were the mean and standard deviation: two parameters. For
our example:

306
 SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

Degrees of freedom = 14 1 2 = 11

Now that we have both the degrees of freedom (df), and the Chi-Squared value, we can use
Excel to calculate the p-Value. Simply enter the formula below, inputting the correct values.

In the case of our example, the resulting p-Value is 0.116. Because the p-Value is greater
than 0.05, we accept the null hypothesis. Our data is normal.

Once you verify that your data is normal, you can use many statistical tests to draw
conclusions about the data and the population as a whole from your sample. One of the

data.

roughout this chapter is

a sample selection of the weights of boxes in a delivery channel. The 50 samples were taken
from the same shipping center over the course of a week and were randomly selected.
Given that we know the data is normal, what is the probability that any package selected
from the distribution center will weigh over 50 pounds?

Excel provides the following formula:

307
 NORMAL PROBABILITY DISTRIBUTIONS

When using this formula:

- X = the value you are testing

- Mean = sample mean from the descriptive statistics
- Standard_dev = sample standard deviation from the descriptive statistics
- Cumulative = FALSE if you want to test the probability of the exact occurrence;
TRUE if you want to test the probability of any value left of x on the curve (or, in
most cases, less than x)

Entering the formula in Excel for our data set with a FALSE cumulative tells us there is a 3.47
percent chance of a randomly selected box weighing 50 pounds. This is a real and important
piece of information that might be beneficial to a Six Sigma team.

Entering the formula in Excel for our data set with a TRUE cumulative tells us there is a
54.25 percent chance of a randomly selected box weighing less than 50 pounds. Subtracting
from 100 percent leaves a 45.75 percent chance that a box will weigh over 50 pounds.

Why might this be important? In this particular case, a shipping manager might be working
to hire shipping reps. Because there is a strong likelihood that the employees will handle
boxes over 50 pounds, the manager knows to include that requirement in job descriptions.
Or, a Six Sigma team might be working on a process to increase production in a warehouse;
the knowledge that many of the boxes are heavy enough to require special handling is
relevant to how the team approaches the process.

Six Sigma teams can use Excel to generate a lot of information about the data. Excel also
provides a formula for the inverse of the above calculation. What if the team wanted to
know about the weight of the heaviest 80 percent of boxes in the shipping center?

The formula is:

Probability must be entered as a decimal point, not a percent. In the case of 0.8 in our
example data set, all of the boxes in the heaviest 80 percent are above 58.40 pounds.

Because we are drawing statistical conclusions from samples, there is some error inherent
in the process. Error is usually calculated for in advanced analysis.

308
 SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

How to apply statistical data and conclusions to a business or project approach is covered in

information that is applicable to business processes.

309
Correlation and regression are statistical concepts that help Six Sigma teams understand
whether two factors within a process are related. You can use these tools to calculate
whether an input is related to an output in some way or even whether two inputs might be
related. The concept of correlation and regression was introduced in chapter 14 in the

coefficient in Excel, and what that number means, as well as how to create linear regression
models in Excel to test relationships between a dependent and independent variable.

It's worth covering a point made in chapter 14 again. Strong correlation or regression
between two variables does not indicate a causal relationship. Even when strong regression
is found be
dependent result is caused by the independent variable. This is a critical concept to
understand, because attributing causation because correlation is strong can mean making
poor decisions regarding a process. A team might decide to alter or remove one element
because it has a strong correlation to an output; they might then be surprised that the

Why work with correlation and regression at all?

Correlation and regression is still important to understand, because it helps teams

understand more about processes, analyze how elements of a process are related, and
make predictions about a process. When two elements are strongly related along a linear
line or even along a curve teams can create equations linking the two elements. This lets
Six Sigma teams predict how the dependent element will perform if a change is made in the
independent variable.

B
assume normal data in all cases, but it is possible that very abnormal data can skew results.
Before analyzing data, Six Sigma experts should always understand what type of data they
are working with.

310
 SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

Typically, when talking about correlation, we mean the linear association that exists
between two variables. In chapter 14, we introduced the concept that correlation can be
strong or weak; it can also be positive or negative.

Positive Correlation: One variable increase as another variable increases.

Negative Correlation: One variable decreases as another variable increases.

Correlation can sometimes be roughly approximated by viewing a graphical analysis of data

in the form of a scatter diagram or X/Y plot. The steps for creating such diagrams were
covered in chapter 17. To understand how various types of correlation might appear
graphically, consider the example scatter diagrams below.

In the scatter diagram above, temperature has a strong positive correlation to time. This
means that temperature in the process has a tendency to increase as time increases. The
relationship is strong, because you could draw a best fit line, as seen in the image to the
left, and it would touch almost every data point with minimal deviation of data points from
the line.

311
 CORRELATION AND REGRESSION

The diagram above features a strong negative correlation. This means that temperature in
the process has a tendency to decrease as time increases. Again, you can draw a best fit line
through the points of data and all of them remain very close to the trend line, as seen in the
graph to the right.

312
 SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

In contrast, the graph to the left represents a set of data where correlation might be
present, but it is weaker than either of the two examples above. Weak correlation is
represented on a scatter diagram where a trend line does seem to exist and where roughly

Typically, the data points located above and below the trend lines are somewhat
symmetrical, meaning that one side d

correlation can be easy to see, as in this case, or more difficult to see on a graph, warranting
the use of statistical analysis to make a strong case for correlation.

313
 CORRELATION AND REGRESSION

The graph above shows a scatter diagram that indicates no or almost no correlation
between time and temperature in a process. The data points are scattered randomly over
the graph, and the trend line is almost horizontal. A horizontal trend line is usually an
indication of a lack of correlation.

image capturing a small sample of data could show correlation or no correlation when
correlation did exist in the overall sample or data set. Graphical analysis is useful to gain an
idea about data relationships and very useful for illustrating relationships to others but
statistical analysis helps teams understand relationships with more certainty.

For the purpose of correlation, the statistical calculation we are concerned with is known as
the correlation coefficient, or R value. The correlation coefficient is a number between -1
and 1. As the correlation coefficient approaches either -1 or 1, a relationship between two
variables is considered more likely or stronger.

If the correlation coefficient is equal to 1, then the relationship is certain: all

the data points on the set are found on the positive trend line.
The same is true for a correlation coefficient that is equal to -1, except that all
the data points are found on the negative trend line.
A correlation coefficient equal to 0 means there is certainly no relationship
between the variables, and there is no trend line.
314
 SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

Most of the time, when you calculate correlation coefficients, the number will
be between 0 and 1 or 0 and -1.

Correlation coefficients are calculated using several formulas. One of the most common

below:

Another calculation uses the sum of squares. The formula is shown below.

Where:

of these formulas rely on the specific (x,y) data arrived at by combining data pairs. For
example, consider the following data set:

315
 CORRELATION AND REGRESSION

Months on Job Average Production

2 10

2 11

3 12

3 11

3 13

4 11

4 15

4 15

5 16

5 15

5 18

If average production is considered the Y axis and months on the job the X axis, paired
points of this data set would be written as:

2, 10
2, 11
3, 12
3, 11
And so forth

Those numbers can then be included in the equations above; the sigma symbol (which looks

the (sum of all x)*(sum of all y) to arrive at the top number, for example.

It is possible to calculate the correlation coefficient using one of these formulas, but luckily

can use statistical analysis software or graphing calculators to arrive at the correlation

316
 SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

quick ways of determining correlation coefficients using Excel.

For both ex
Excel worksheet.

Time Temperature A Temperature B Temperature C Temperature D

1 30 70 35 30
2 35 68 48 31
3 37 63 75 38
4 40 62 25 24
5 41 55 50 35
6 48 47 34 49
7 57 42 48 45
8 55 41 31 59
9 59 38 24 42
10 62 36 41 65
11 63 31 71 64
12 65 25 55 59

One of the easiest ways to calculate the correlation coefficient for two sets of possibly
related data is to use Excel CORREL formula. Follow the steps below to calculate the
correlation coefficient between time and Temperature A.

1. In any cell in the worksheet where you copied the data table above, the formula
=CORREL

317
 CORRELATION AND REGRESSION

2. Array 1 is the set of cells containing data for time. Array 2 is the set of cells
containing data for Temperature A. You can select these cell ranges with your
mouse, separating them with a comma in your formula. The end result should
appear as follows.

3. Hit enter.
4. Excel will calculate the correlation coefficient. In this case, it is 0.981167422

The correlation coefficient calculated above is for the data set that created the very first
scatter diagram for this chapter. If you recall, the scatter diagram appeared to show that

318
 SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

the data had a very strong positive correlation; a correlation coefficient (or R value) of 0.98
is certainly in line with that assessment.

above. Repeat the steps above, but change your Array 2 range to reference the data set for
temperature C.

The R value for time and temperature C is calculated as 0.107, which is much closer to zero.
This data set created the scatter diagram above that appeared to have no correlation. The R
value corresponds with the conclusions drawn from the scatter diagram. In most
applications, correlation is considered to occur at some level if the R value is 0.4 or more
or -0.4 or less.

You can also use the Analysis ToolPak to calculate correlation coefficients. Follow the steps
below to calculate the R value for time and Temperature B.

1. Format your spreadsheet so that the Time and Temperature B columns are next to
each other.
2. Select Data and then Data Analysis from the top menu.

3. Highlight Correlation and select OK.

319
 CORRELATION AND REGRESSION

4. Click in the input box and select the inputs for the correlation analysis. Select the
data from both the time and temperature B columns.

5. Note that your data is grouped by columns. If you selected the header rows in the

6. Opt to output in a new worksheet.

7. Click OK.
8. Excel will calculate the correlation coefficient.

320
 SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

In the above example, the R value shows a very strong negative correlation between time
and temperature. This data set is the same set used to make the negative correlation
scatter diagram above, so the R value is consistent with that graphical analysis.

Repeat the steps above to find the correlation coefficient of time and temperature D.

The R value returned shows a weaker positive correlation than in our first calculation using
the CORREL function, but we would still consider these data correlated. The data for
temperature D is the same data that created the third scatter diagram in the graphical
analysis portion of this chapter.

Once you know your data has a linear correlation, you can conduct further analysis through
regression testing. Regression lets you create a linear model that helps you predict how one
variable might behave given changes in the other variable. Note again that the changes
only that the variables are intertwined in some
way.

When working with regression, we talk about the coefficient of determination, or the r2
value. The r2 value is, literally, the square of R, or the square of the correlation coefficient.

321
 CORRELATION AND REGRESSION

So, from our last example in the correlation coefficient section, R = 0.860624. The r2 value in
this case would be 0.74.

The r2 value is a number between 0 and 1. The number tells us about the strength of the
linear relationship between two variables x and y. While the R value tells us how likely it is
that two values are related in some way, the r2 value tells us how much the fluctuation in
one variable (the dependent variable, or y) is related to changes in another variable (the
independent variable, or x).

In the example used above, r2 is 0.74. That means that approximately 74 percent of the
variation in the temperature is related to the change in time; the other 26 percent variation
in temperature is unexplained.

Regression analysis lets us create a best fit line and equation for our data, and the r 2 value
tells us how confident we can be in using that line to make predictions about data.

To understand how regression works, consider the same data set used in the correlation
coefficient examples above:

Time Temperature A Temperature B Temperature C Temperature D

1 30 70 35 30
2 35 68 48 31
3 37 63 75 38
4 40 62 25 24
5 41 55 50 35
6 48 47 34 49
7 57 42 48 45
8 55 41 31 59
9 59 38 24 42
10 62 36 41 65
11 63 31 71 64
12 65 25 55 59

relationship on a regression model using the Analysis ToolPak in Excel.

322
 SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

Begin by creating a regression worksheet using the data. Follow the steps below.

1. Copy the data table into a new worksheet in Excel.

2. Select Data and then Data Analysis from the top menu.

3. Select Regression and click OK.

4. Select Temperate A as your Y input range and Time as your X input range.

323
 CORRELATION AND REGRESSION

5. If you selected header data when inputting ranges, check that labels are included.
6. Check the box for new worksheet.
7. Check the box for line fit plots.
8. Click OK.

Excel generates a new workbook with a number of statistical outputs. The relevant values
for t

324
 SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

The Multiple R value is the correlation coefficient. You can see that the coefficient in this
case matches the correlation coefficient calculated for this data set in the last section. The
r2 value for this data set is 0.96, which means almost all the variation in temperature in this
data set is related to a change in time.

Because you selected the line fit plots box when creating your regression analysis, Excel has
also generated a scatter diagram and a best fit plot.

325
 CORRELATION AND REGRESSION

The chart shows all of the actual data points in blue and predicted data points along a best
fit line in orange. Those data points are calculated by Excel using a formula created from the
data. You can display the formula by making design changes to the graph. Follow the steps
below to display the formula.

1. Click on the graph in Excel.

2. Select Design.
3. Select Add Chart Element.

4. Select Trendline, More Trendline Options.

326
 SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

5. Select the option for the predicted trend line and click OK.

327
 CORRELATION AND REGRESSION

6. Click the box to display the equation on the chart.

7. Note the trend equation on the graph.

328
 SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

In this case, the equation is y = 3.3497x + 27.561.

The r2 value, if you recall, is 0.96, which means that our linear equation is going to be very
good at determining the approximate temperature in our process as a relation to time.

At 5 minutes, for example, our equation predicts that the temperature is 44.30.

Y = 3.3497(5) + 27.561

At 25 minutes, our equation predicts that the temperature is 111.30.

To see how regression tests might look with low correlation, repeat the steps above with
the data for temperature C.

329
 CORRELATION AND REGRESSION

Again, the correlation coefficient is very low. The r2 value is also very low; according to this
model, only about 1 percent of the variation in temperature is related to time.

If you format it to do so, Excel will create a predicted line and a linear trend equation. Given
the extremely low r2 value, though, you know that this equation is not useful in making any
real predictions about the temperature for a given time. You only have to look at the graph
to see this is true. For a time around 3, the equation predicts a temperature of just over 40.

330
 SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

that might point to a problem of data collection or a temporary process issue; the data
throughout the chart is skewed far away from the line.

Correlation and regression are useful tools, usually deployed by teams in the measure,

useful in all situations where they might apply. One of the things that a Six Sigma expert
develops over time is the ability to decide which tools might be relevant given certain data
and situations. Because Excel and other statistical analysis software make it easy for Six

there is no harm in creating correlation or regression analyses for data to explore

connections between variables. This is true as long as you are careful about what

types of statistical analysis.

Both correlation and regression work with continuous or ratio data, and both sets of
have x variables that are
the names of people and y variables that are the number of outputs.

Suzy 10

Jim 11

Rhonda 15

Misha 14

Aisha 13

The above chart is an example of the type of data that would not apply to regression and
correlation analysis, but might be analyzed using tools such as Pareto charts or pie charts.

Correlation requires two quantifiable data elements. Regression requires data that can be
written as function:

331
 CORRELATION AND REGRESSION

Some examples of data that can be written as a function include:

Distance
o Y, the distance, is function of time, speed, acceleration, etc.
Hardness of a substance
o Y, the hardness, is a function of temperature, alloy, etc.
Experienced (feels like) temperature
o Y, the temperature it feels like, is a function of the actual temperature,
wind, humidity, etc.
BMI (body mass index)
o Y, the BMI, is a function of height and weight

Correlation and regression can be applied by Six Sigma experts at a number of points in a
DMAIC process. During the analyze phase, teams can use these tools to help verify
relationships between inputs or to help bolster understanding of root cause analysis. While
correlation and re

factor, then a strong regression model helps validate that assumption. This is especially true
if a team couples regression modeling with trial-and-error hypothesis testing.

For example, a Six Sigma team might conduct brainstorming and fishbone diagrams, coming
s Six
Sigma experts conduct regression testing, which shows that variation in B is definitely
related strongly to variation in A. The team can remove A from the process or alter A in
some way and measure the process again. The team can then use hypothesis testingto
determine if their changes created a statistical difference in the output.

An even more important function of regression in the DMAIC process is determining

optimum input values so teams have a goal for control. Consider the example data used
throughout this chapter. What if the team exploring this sample data was working with a
process that involved mixing ingredients at a specific temperature range to create certain
results? If the continuous time a mixing machine runs causes temperature fluctuations,
those fluctuations would create variation in the final product. If the team discovered that
the temperature range 40 to 50 degrees correlated to the most optimal product output,
then the team could use the equation provided by the regression model to determine at
what time range the mixing machine should be operated.

Consider the data for Temperature A again:

332
SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

Temperature
Time A
1 30
2 35
3 37
4 40
5 41
6 48
7 57
8 55
9 59
10 62
11 63
12 65

333
 CORRELATION AND REGRESSION

This data had both a strong correlation coefficient and a strong r2 value, which means the
team can rely on the equation for the best fit line.

In this case, to find the time values relating to a range between 40 and 50 degrees, the team
would solve for x when y = 40 and when y = 50.

40 = 3.3497x + 27.561

40 27.561 = 3.3497x

12.439 = 3.3497x

3.71 = x when y = 40

50 = 3.3497x + 27.561

50 27.561 = 3.3497x

22.439 = 3.3497x

6.69 = x when y = 50

Using these results, the team decides that mixing must be done when the machine has been
on between 3.71 and 6.69 minutes. They create a change in the process that involves
starting the mixing procedure after the machine has warmed up for 4 minutes.

334
In chapter 18, we discussed the normal probability curve, which is also called the Gaussian
he most important and

of data is not going to fall along the normal probability distribution. This chapter covers
non-normal probability distributions, how to recognize some of the more common
distributions through graphical analysis, and some information about using statistical
analysis with such distributions.

Some of the simpler statistical calculations for non-normal probability distributions are
included in this chapter, along with details for how to handle those calculations manually.
Calculations for all of the distributions mentioned are not included, because they can be
very complex and almost no one
how to calculate probabilities using a wide variety of distributions in Minitab.

information about
the normal curve covered in chapter 18. First, remember that the normal distribution is
related to continuous data. This means that your data is associated with random variables
that can take the form of any of an infinite number of points along an interval. Temperature
over time is a continuous type of data. Yes, you might never see the temperature of a room
fall below 40 degrees F, but within a range such as 40 degrees and 80 degrees are an infinite
number of continuous data points if you measured to a high enough degree of accuracy.
335
 NON-NORMAL PROBABILITY DISTRIBUTIONS

Remember that discrete data is not continuous in nature. Discrete data is typically about
categorical occurrences. In some cases, discrete data can be converted to continuous data
to allow Six Sigma teams to use more statistical analysis tools when working with the

statistical analysis on discrete data. Discrete data usually involves either a binomial
distribution or a Poisson distribution; both will be covered later in this chapter.

The normal curve, when presented in graphical format or on a histogram, is typically

symmetrical in nature. The data elements tend to concentrate along the center of the
curve, with elements fanning out slowly on either side as the curve drops ever toward zero.

While a visual inspection of data through graphical analysis can help determine if data might
be normal, there are several continuous distribution curves that look a lot like normal
-Squared Goodness-of-Fit Test

Chi-Squared Goodness-of-Fit Test with the Chi-Squared distribution, which makes an

appearance later in this chapter. One is a statistical test and the other is a probability
distribution.

The normal curve:

Is symmetrical
Is most commonly required for inferential statistics
Is described statistically using both the mean and standard deviation
o The curve centers on the mean
o The number of standard deviations away from the mean changes the
probability of the amount of data under the curve

The exponential distribution creates a histogram or trend line that is exponential in nature.
It decreases exponentially as you travel across the x-axis. See the image below for an
example.

336
 SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

In an exponential distribution, the data is not organized around the mean as in the normal
probability curve, but at one end of the x-axis. The data used to create the histogram above
had a mean of 24.79 but a median of 16. Often, when working with non-normal data, the
median is a more applicable data point than the mean for graphical analysis. If you locate
the median on the x-axis of an exponential histogram, you know that 50 percent of the
other data points fall before it and 50 percent after it.

Exponential distributions are often used when working with data sets that include arrival
times, the average or mean time between failures in a process, wait-line theories, or the
distance in time or space between interesting points of occurrence in data, processes, or
experiments. The exponential distribution is particularly valuable when working with certain
Poisson distributions. If you recall, Poisson distributions are discrete rather than continuous,
which means Six Sigma teams are limited in what statistical tools can be used to analyze the
data. However, the exponential distribution can often be used to describe the rate of
change within the data that is distributed in a Poisson distribution, providing a method for
creating continuous data that can be analyzed.

The exponential curve:

Never appears symmetrical

Is described statistically by the mean of the data and a value known as lambda

337
 NON-NORMAL PROBABILITY DISTRIBUTIONS

Here is another graphical representation of an exponential distribution generated in

Minitab, which is a statistical analysis software.

Distribution Plot
Exponential, Scale=1, Thresh=0

1.0

0.8

0.6

0.4

0.2

0.0
0 2 4 6 8 10 12 14 16
X

As with the normal distribution, you can run a statistical test in Excel to determine if your
data is exponential. The test is basically the same Chi-Squared Goodness-of-Fit test we used
in chapter 18. The same idea is used: you first predict the number of data points you would
find in each bin in a perfect exponential curve given the statistical data regarding your set.
You then compare the actual breakdown of your data against expected numbers to find the
Chi-Squared value, which is used to evaluate your hypothesis that the data is exponential.

Because we are going to introduce using Minitab for statistical analysis functions in the next
few units, and because Minitab handles Chi-Squared Goodness-of-Fit testing and many
for testing each
type of data distribution in Excel. Because most of the tests that can be performed using the
free Analysis ToolPak in Excel require normally distributed data, the steps for testing for
normal data in Excel in chapter 18 should be sufficient for much of your Excel-only based
work.

338
 SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

At first, the lognormal distribution might appear to be an exponential distribution, but on a

graphical analysis, it is obviously constrained by zero. The lognormal distribution on a
histogram is asymmetrical, with the trend line appearing more as a wave that moves across
the page. The location and size of the wave varies with each data set, and the shape can
even be very different from the one below.

Lognormal distributions are often used when working with data sets that describe time
durations, such as the time a process or machine is down, or distribution of assets or wealth
among the population. Anytime data has a positive skew, which means it the tail of the data
is to the right, then a lognormal distribution might be possible. Consider the example of
wealth within a population. If we only consider the amount of income each household
within a population reports, almost
impossible to have an income of less than zero. (Yes, you can have a net worth that is

339
 NON-NORMAL PROBABILITY DISTRIBUTIONS

negative, and you can have a cash flow that is negative if you consider income and
expenses, but for our exampl )

Now, in a random sample of income from across a large population a city, for example --
data is likely to be lognormal. As you move further right in the bins on a histogram
representing increasing dollar figures of income there are fewer households that will fall

represented on a histogram. While the effect is similar to the positive exponential

distribution seen above, there are some differences. The lognormal distribution of the

to a median point, and then a sloping trail to the right.

The lognormal distribution:

Typically describes a data set with values in large ranges

Can be described with both mean and standard deviation
Always has a positive skew

Here is another image of a lognormal distribution, generated from Minitab.

340
 SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

Just to make things more confusing, probabilities can fall into families of distributions, such
as the Weibull distributions. Data that fits a Weibull distribution might also fit another
distribution, including all of those discussed previously in this chapter. Note that Weibull
distributions relate to continuous data, so the data would not likely fit a discrete data
distribution.

Look at the Weibull distributions below, which were created in Minitab. Do these images
look familiar?

341
NON-NORMAL PROBABILITY DISTRIBUTIONS

342
SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

343
 NON-NORMAL PROBABILITY DISTRIBUTIONS

The curves appear to fit the lognormal, exponential, and normal distributions, respectively,
even though they each also fit the Weibull distribution.

Weibull distributions can describe many data types and are often used when working with
reliability applications and failure probabilities that change or vary with time. For example,
the chance that a plasma screen will be damaged by burn-in increases both over the time
the screen is used and over the time that a single still image is displayed on a screen. Data
associated with such a concern might feature a Weibull distribution.

As previously stated, you will commonly work with normal distributions when applying
statistical analysis. Even if the underlying distribution of your data set is not normal, if you
have a large, random sample of continuous data, you are usually able to perform statistical
analyses that are reserved for normal data thanks to the Central Limit Theorem. The Central
Limit Theorem states that the distribution of the mean of a large, identically distributed
number of independent variables will approximate the normal curve
be able to use software like Minitab to apply statistical analysis to a number of data sets.

at are not
-normal
continuous distributions and provide images.

In a graphical representation, the Cauchy distribution often looks like an elongated normal
curve with a tighter peak. While the Cauchy distribution does share some characteristics

defined mean or variance, which makes it not useful for many common statistical analysis
tools used in Six Sigma.

344
 SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

The logistic distribution also appears to approximate the normal curve, and is, in fact, used
in some science and math functions to approximate other symmetrical distributions. This is
because the CDF of a logistic distribution is more consistently calculable. Remember that
the CDF is the cumulative density function and calculates the probability that a given data
point is less than or equal to X, where X is any point on the x-axis of a probability curve.

345
 NON-NORMAL PROBABILITY DISTRIBUTIONS

The Laplace Distribution is often referred to as the bilateral exponential distribution or the
double-exponential distribution. You can see from the image below, the Laplace distribution
does seem to pair exponential distributions back to back.

346
 SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

A uniform distribution occurs when data points are divided evenly among bins. Exact
uniform distributions are almost never going to occur in randomly-sampled data, which
means Six Sigma teams that see such distributions in data should check the authenticity of
the data. Uniformity often points to an error of measurement, problem with the
measurement system, a non-random sample, or an issue with a process.

In some very specific cases, uniform distributions could point to success. For example, in a
manufacturing plant, one machine might make dowels that are used later in the
manufacturing process. If the dowels must measure 4 inches for functionality later, and a
random sampling of dowels across all time buckets for the machine show 4-inch dowels
being made, then uniformity equals success.

347
 NON-NORMAL PROBABILITY DISTRIBUTIONS

Like Weibull distributions, Beta distributions can take on a number of shapes, two of which
are pictured below. Beta distributions are considered extremely flexible, and can become
stand-ins for other distributions given certain statistical parameters.

348
SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

349
 NON-NORMAL PROBABILITY DISTRIBUTIONS

Gamma distributions are similar to Beta and Weibull distributions, in that they take on a

just more specific forms of the Gamma distribution. Gamma distributions are, however,
always skewed to the right.

350
SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

351
 NON-NORMAL PROBABILITY DISTRIBUTIONS

A triangular distribution is formed using the mode and the upper and lower limits of a data
set.

A binomial distribution is used when you are dealing with discrete data and there are only
two outcomes for each trial or sample. In short, binomial distributions relate to attribute
data: Yes or no, pass or fail, one or the other.

A binomial distribution assumes that the outcome of any given trial is independent of the
outcome of another trial. If you flip a coin, it will either be heads or tails. If you flip it again,

matter what happened with the first flip of the coin.

352
 SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

In most cases when considering attribute outcomes, the probability is not distributed evenly
between the two events, as it is with a coin flip. Instead, there is a probability of success
(notated as p) and a probability of failure (notated as 1 p). For example, if a machine
making parts creates defects in 5 percent of the parts, then you have two outcomes for
each part made. Either the part will be defective, or it will not be defective.

For this example, p (the outcome that the part will be defective) equals 0.05, and 1 p (the
outcome that the part will not be defective) is 0.95. If you select one random part, the
chance of getting a defective part is 0.05.

e seeming discrepancy in the above two paragraphs.

First, we said that the probability of success was notated as p. Then, we said that the
probability of pulling a defective part is p, or 0.05. In business terms, pulling a defective part
would not be deemed a success. However, in statistical terms, we are not talking about the
quality nature of the outcome. We are talking about whether an outcome will successfully
match the definition we have created. In this case, whether the outcome will be defective.)

The Binomial distribution of the data described above on a histogram might appear as
follows:

353
 NON-NORMAL PROBABILITY DISTRIBUTIONS

When you are dealing with binomial data sets, you can use the Binomial Distribution
Probability Function to determine the probability of the number of successes you will have
within a given number of trials given the sample data you have. The answers generated by
the Binomial Distribution Probability Function correlate to the histogram generated by
Minitab given the same information.

To understand what we mean by

this example, imagine a coin is tossed 10 times, and we consider heads to be success. The
chance of success (heads) for each coin flip is 0.5. Given that information, Minitab creates
the histogram below.

354
 SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

This graph shows us that, if we toss the coin 10 times, the chance of getting 1 success in all
those 10 tosses is very small as is the chance of getting 9 successes. The chance of getting
5 successes out of 10 is much higher but not actually 50 percent.

To apply this concept to a real-

which produces defective parts at a rate of 0.05. Again, the histogram below illustrates the
data that might occur if we took 20 random samples from the process.

355
 NON-NORMAL PROBABILITY DISTRIBUTIONS

The probability that we will have 0 successes (defective parts) in our sample of 20 is a bit
over 0.35. The probability that we will have 1 defective part is a bit higher, but the chance
small. The probability of more than 5
defective parts in our batch of 20 is almost nonexistent.

What does this tell a real-world Six Sigma team? Imagine a quality control team took 20
random samples from this process every day. For two weeks straight, the quality team
records no defects within those samples. If the Six Sigma team believes a process generates
0.05 defective parts, then a lack of defective parts in the samples shows that (1) something
has changed to improve the performance of the process, (2) the measurement system is
defective, or (3) the samples are not actually random.

If you increase the number of trials to 100, the binomial distribution remains.

356
 SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

Here, you can see that the greatest likelihood for this process is that, within 100 samples, 4
or 5 will be defective. It is unlikely, given a defect rate of 0.05, that a random sample of 100
would ever have more than 11 or 12 defective parts.

What we have done graphically, you can also do mathematically using the Binomial
Distribution Probability Function. The equation is provided below.

For this formula:

x is equal to the number of successes desired

n is equal to the number of trials

357
 NON-NORMAL PROBABILITY DISTRIBUTIONS

p is equal to the probability of success in each trial

q is equal to 1 p
The exclamation point indicates a factorial.
o 3! = 3x2x1 = 6
o 4! = 4x3x2x1 = 24
o 5! = 5x4x3x2x1 = 120

Using the Binomial Distribution Probability Function, we can calculate the probability that
20 trials will yield 4 defective parts in a process that delivers 0.05 defective parts.

358
 SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

The probability of pulling 4 defective parts is very low, as was shown on the distribution
graph.

What are some reasons this type of calculation might matter in the business world or for Six
Sigma teams?

Consider a company that sells chocolates. If 20 chocolates come in each box, what is the
chance that a customer might receive defective product in his or her box? What is the
chance that a customer might receive more than one defective chocolate? Teams can
calculate these probabilities after sampling data to determine how often a process
generates defective product.

What if a Six Sigma team randomly samples the chocolate process and determines that the
process operates at a 4-sigma level. Remember from Unit 1 that at 4 sigma, a process

chocolates per million created, or a chance of a chocolate being defective equal to 0.00621.

In a box of 20 chocolates, with p = 0.00621, Minitab generates the following binomial

histogram.

359
 NON-NORMAL PROBABILITY DISTRIBUTIONS

From the above histogram, you can see that the likelihood of a box having no defective
chocolates is approximately 88 percent. The likelihood of a box having one defective
chocolate is approximately 10 percent. The likelihood of a box having two or more defective
chocolates is almost nonexistent.

What does this mean for the business? It depends on the goals of the business or the Six
Sigma project. If the Six Sigma team was tasked with improving customer satisfaction or
decreasing the rate of return of chocolate boxes, then a 10 percent probability of a
defective chocolate in a box might be something the team wants to address. However, if
customer satisfaction is strong and return costs are low, then the potential complaints or
returns associated with a small number of boxes might be acceptable to the business
particularly if a Six Sigma team discovers that reducing the error rate would cost well over
the amount that it costs to handle the errors.

By applying the binomial distribution in this manner to pass/fail, yes/no, or defect/no

defect, Six Sigma teams can calculate data that is useful in a variety of real-world situations.
The ability to tie costs to defects makes it even more possible to evaluate this type of
information using binomial probability, because teams can speak in terms of the probability
that different processes or outcomes will cost varying amounts of money or save varying
amounts of money both of which are metrics company leadership will be interested in
hearing.

The Poisson distribution is often used when dealing with data that is distributed randomly
within time, distance, or other unit of measurement. For example, the distribution might be
relevant when dealing with calls per hour in a call center, traffic accidents per week on a
speci
a good indication that the Poisson distribution might be useful.

As you can see from the graph before, the Poisson distribution can look a little like the
normal curve, but it is not symmetrical. This curve, based off a mean of 4, tails slightly to the
right.

360
 SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

This graph of a Poisson distribution with a mean of 42 looks even more like the normal
curve. If you look very close, however, and count the bins to the left and right of the mean,
you see that the curve tails very slightly to the right. This is a good example of why
statistical analysis is needed in addition to graphical analysis when working with data. Data
can approximate the normal curve on a graph without being normal.

361
 NON-NORMAL PROBABILITY DISTRIBUTIONS

One interesting thing to note about the Poisson distribution is that it relates closely to the
exponential distribution. Remember, the Poisson distribution is discrete, not continuous.
The distribution is an analysis of how many of one thing occurs per another. For example,

362
SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

Errors Per Hour

5
8
5
5
4
7
4
12
3
6
9
9
7
3
6
5
7
6
4
5
5
2
4
7
6

363
 NON-NORMAL PROBABILITY DISTRIBUTIONS

Here is the histogram of the above data:

make in hour 1, hour 2, hour 3, and so forth. However, if the process is following a Poisson
distribution, then the time between each event in the process when looked at continuously
yields an exponential distribution. Understanding this relationship can help Six Sigma teams
convert Poisson distributions to exponential distributions when necessary for statistical
analysis. The use of this conversion within a Six Sigma environment is not likely to be
common for most teams, but could be useful in special applications
cover in the industry-specific sections at the end of this book.

The Poisson distribution might be applicable when the following statements can be used to
describe your data:

Data describes events that occur at a random interval measurement the word

364
 SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

The sample size (number of intervals) is at least 16. If you are measuring per hour,
you have
least 16 feet of data.
You know that the population size is 10 times the size of the sample or bigger. If you
irst 16 hours
the worker is on the job.
Event occurrences are independent of each other. In our data-entry example, if we
are counting each unnecessary keystroke as an error, we might count the
backspace. However, the backspace is dependent on the person making another

count the backspace as error 2, then the errors are not independent.
The probability of occurrences is less than 0.1. Consider the data-entry example
above. When using a 10-key number pad, the average rate of entry for employees
in the United States is approximately 8,000 per hour. If an employee is making 8
12 errors an hour, the probability of an error for each keystroke is well below 0.1.
The rate of occurrence per interval is constant. The average number of phone calls
per hour is 50; the average number of errors per hour is 6.

If data fits the Poisson distribution, you can use an equation to determine the probability of
certain events.

The Poisson equation is:

Within the Poisson equation:

e equals the Poisson constant, which is 2.71828

is the mean or average of the process
is the specific case or data point; remember that the exclamation point indicates
a factorial.

365
 NON-NORMAL PROBABILITY DISTRIBUTIONS

For example, if a Six Sigma team was working to improve processes in a call center, they
might have gathered data that indicates the average number of people on hold for a
customer service representative at any given time is 10. Given that information and a
Poisson distribution when considering number of holds per increments of time, the team
can use the Poisson equation to calculate probabilities related to various call scenarios.

What is the probability that there are 15 callers on hold at a given time?

366
 SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

The probability of having exactly 15 callers on hold at a given time in the call center is 0.034,
or approximately 3.4 percent.

a specific protocol or process problem that occurs only

not an extremely likely scenario. However, what if a Six Sigma team has noted that when
the number of callers on hold hits 15 or above, a regularly staffed call center team is
unlikely to catch up, resulting in continued high hold times and volumes. Knowing this, the

some resources from other areas to answering calls.

But how does a process owner or Six Sigma team plan appropriately? Do other resources
need to constantly be on hand? Is the probability of this situation occurring high or low?

To answer the question, you can use the Poisson equation to find the probability of there
being 0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, and 14 callers on hold at any given time.
Because the probability of all possible outcomes totals to 1, you would subtract the sum of
all those probabilities from 1 to determine the probability of the number of holds being 15
or more.

Mathematically, the concept is written as:

To manually calculate the above probability, you would have to use the Poisson equation
for each P(x) listed above. You can also use statistical analysis software, such as Minitab, to

chapters. For now, though, consider the chart below, which shows the probability of each x.

0 0.0000454

1 0.0004540

2 0.0022700

3 0.0075667

4 0.0189166

367
 NON-NORMAL PROBABILITY DISTRIBUTIONS

5 0.0378333

6 0.0630555

7 0.0900792

8 0.112599

9 0.125110

10 0.125110

11 0.113736

12 0.0947803

13 0.0729079

14 0.0520771

Summing the probabilities for each X provides a total probability that x will be less than 15
of 0.916541.

Alternatively, you can use statistical software to determine the cumulative probability the
area under the curve to the left of a certain point. The cumulative probabilities of x being
less than or equal to 14 is 0.916541.

Therefore, the probability of x being 15 or greater is 1 0.916541, or 0.083459.

The probability that there are 15 or more callers on hold at any given time is about 0.083, or
just over 8 percent. Given this probability, call center leadership can make a more educated
decision about whether they want to pay for more staff or take a risk at hitting the higher
call volume and having to play catch up for the rest of the shift.

Teams can use the Poisson distribution and calculations in this manner to understand the
probabilities of different events occurring. They can also tie financial data to such events to
make even better data-driven decisions.

368
 SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

The geometric distribution is used when there are two outcomes for a trial, trials are
independent, and there is a waiting time before the first occurrence. As with the binomial
distribution, the outcome of each trial must be independent. A real-world use of the
geometric distribution might include the number of inspections that occur before finding a
defective part or the number of resumes or applications a recruiter reviews before finding
the first candidate that meets minimal job requirements.

An example of the geometric distribution is seen below.

The negative binomial distribution is also used with attribute data fail/pass and other
situations where there are only two outcomes for each trial. Again, the outcomes must be
independent of each other; the probability of an occurrence must be the same across each
trial, just as with the binomial distribution. The negative binomial distribution is often used
369
 NON-NORMAL PROBABILITY DISTRIBUTIONS

when teams are trying to determine the probability of a certain number of passes or fails
before reaching the sth pass or fail.

Sigma teams apply raw data about processes to real-world concerns and situations. As
previously stated, many of the statistical analyses you will perform on data will be related to
the normal curve. But understanding a few equations related to distributions such as
Binomial or Poisson helps you expand your ability to create real-world assumptions about
data.

When you can determine the probability for certain events or data points, you can make
educated decisions about a process. What is the probability that a cost-heavy event will
occur? Is the cost associated with an event enough to create a reason for working to reduce
the likelihood of that event? If the probability of a low-cost defect is high, are the total
expenses associated with that defect likely to be higher than the total expenses associated

370
 SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

with a high-cost defect that has an extremely low probability? These are just some of the
questions that teams can use statistics and probability distributions to answer

371
Hypothesis tests are statistical tools used to draw conclusions based on data. With a basic
understanding of the process and process performance, as well as real-time data or as
close to real-time data as possible Six Sigma teams can draw statistical conclusions that
are extremely accurate if they (1) ensure that the measurement systems are good, (2) have
the right sample size, and (3) know how to set up the right type of hypothesis test.

talk about
different types of hypothesis tests, how to set them up, and how to read the results. In

-by-step guide for running

tests in the software program. You can also run some tests in Excel using the Analysis
Toolpak you previously downloaded.

Hypothesis tests cover three broad categories:

Testing whether the data you have fits a data model. In Chapter 18, we
conducted a hypothesis test to determine whether data fit the normal curve.
We used the Chi-Squared Goodness-of-Fit Test, but ultimately, it was a
hypothesis test.
Comparing a statistic to a hypothesis about the data or population.
Answering the question whether something changed within the data, often
after a team has modified an input or other part of the process. In the case of
most Six Sigma projects, the team probably wants to find out whether the
process or outcome is improved.

While the type of hypothesis test you use depends on the answers you are seeking and the
type of data you have, all of the tests follow essentially the same guidelines:

You begin with a statistic or criteria that you usually compute from your sample
data

372
 SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

You create a null hypothesis and an alternate hypothesis, in keeping with the type
of test you are dealing with
o Remember, in chapter 18 when we ran the Chi-Squared Goodness-of-Fit
Test:
The null hypothesis was that the data was normal
The alternate hypothesis was that the data was not normal
The statistic or criteria is compared against a reference criteria or distribution
How the calculated statistic compares to the reference criteria determines whether
you accept the null hypothesis or reject the null hypothesis in favor of the
alternative hypothesis

Hypothesis tests are a large part of inferential statistics, where we draw conclusions about
the overall process or population by analyzing the sample data and measurements. When
stating hypotheses, we are not making statements about the sample. We are making
statements about the population or entire process.

we can calculate the sample

mean.

Hypothesis tests have two main parts: the null hypothesis and the alternative hypothesis.

The null hypothesis is abbreviated as H0 and is usually a statement about the data that
reflects no effect or no difference. In chapter 18, we hypothesized that our data was

of our data and the distrib

The alternative hypothesis is abbreviated as Ha and is usually a statement that is likely to be

is
a statistical difference between the distribution of our data and the distribution of data on a

hypothesis in this case, that our data is not normal.

Typically, the null hypothesis is an equal statement of some type. The mean of the new
process is equal to the mean of the new process. The distribution of the data is equal to the
normal curve.
373
 HYPOTHESIS TESTING

The alternative hypothesis is typically written as a not equals, a greater than, or a less than
statement. The mean of the new process is greater than the mean of the old process. The
distribution of the data is not equal to the normal curve. How you write the alternative
hypothesis depends on the question you are asking and the type of hypothesis test you are
running.

Anytime you draw inferences about a population from sample data, there is at least some
likelihood of error. With hypothesis testing, errors come in two types.

Type I Error: The null hypothesis is rejected when it is actually true.

o Also called producer risk
o The probability of the risk is measured by alpha, where is a probability
between 0 and 1.
Type II Error: The null hypothesis is accepted when it is actually false.
o Also called the consumer risk
o The probability of the risk is measured by beta, where is a probability
between 0 and 1.

To describe the risk and set up our hypothesis test for what we deem to be an acceptable
risk, a confidence level must be picked. The most common confidence level used is 95
percent, or . Typically, the confidence level is set with the Type I error in mind, so
you use alpha for the confidence level. The value of then contributes to the sample size
requirements and the power; sample size will be covered in the next chapter. Additional
information about alpha and beta values is covered in the section in this chapter on running
individual hypothesis tests.

A school administrator believes that the students in a certain classroom are being impacted
by some environmental issue with that room. The administrator wants to see if the students
in that room perform statistically lower on tests than the students throughout the school;
she has already noted that different classes in that room are taught by other teachers who
also teach in the other rooms, which means she has ruled out a teacher influence.

The administrator gathers data. The rate of failure for classes taught in any other classroom
is 20 percent. Out of 142 students who have a class in the classroom in question, 38 are
failing that particular class.

The hypothesis test would be written out with both a real-world statement and a statistical
statement for both the null and alternative hypothesis.
374
 SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

To determine which hypothesis test to run, you must know:

Which type of data you have (continuous/variable or discrete/attribute)

The number of levels of interest for the input in question (1, 2, or more than two)
Distribution of data (normal or non-normal)
What you are testing (means, medians, variance, count, or proportions)

In this chapter, the hypothesis tests that are covered deal with 1 or 2 levels of interest.

In the following
through running some of these tests using statistical software.

Used when there is only one factor for x

Used when there is one level of interest for x (the input)
o Typically, there is one set of data (the sample) and a target
Used when comparing proportions (percentage, rates) between samples and a
target
Test is available in Minitab
o Requires samples in columns or summary of data to include number of
events, number of trials, and hypothesized proportion

The hypothesis test is set up as:

Note that in the above set-up, the -

Minitab allows you to adjust the alternative hypothesis to a greater than or less than
scenario if desired.

Example of the 1-Proportion Test:

375
 HYPOTHESIS TESTING

H0: There is no difference in the failure rate between the classes taught in the
specific classroom and those taught elsewhere.

Ha: The rate of failure is greater for those classes taught in the specific classroom.

Used when there is only one factor for x

Used when there are 2 levels of interest for x (the input)
o Typically, there are two sets of data regarding the input and you are
comparing those two sets of data
Used when comparing proportions (percentage, rates) between samples or
between samples and a target
Test is available in Minitab
o Requires samples in columns

The hypothesis test is set up as:

Note that in the above set- -

Minitab allows you to adjust the alternative hypothesis to a greater than or less than
scenario if desired.

Example of the 2-Proportion Test:

Reconsider the example above about the school administrator and the classroom. Perhaps
the school administrator determines that students in that classroom do in fact perform
statistically differently from students in other classrooms. The administrator makes a

376
 SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

decision to move the students in that classroom to another classroom. She then gathers
data about test performance in the new classroom.

Now, the administrator has two sets of test performance data for the same students. The
only variable that has changed is the classroom location. She wants to know, is there a
difference between performance in the old room versus the new room?

H0: There is no difference in the failure rate between the old room and the new
room.

Ha: The rate of failure is less in the new classroom.

Used when comparing means

Typically used when dealing with smaller samples or when standard deviation is
known.
o When dealing with larger samples or when standard deviation is not known,
use the 1-Sample Z-Test.
1-Sample T Test compares the mean of a sample to a target mean
Paired T Test compares the mean of a sample against the mean of another sample
for the same factor of X. Usually, the first sample is the baseline and the second
sample is taken after a change is made; the test is run to see if the change made a
difference.
1-Sample T Test is available in Minitab
o Requires standard deviation, mean, and sample size as well as the target, or
hypothesized mean.

The hypothesis test is set up as:

377
 HYPOTHESIS TESTING

Note that in the above set-up, the -

Minitab allows you to adjust the alternative hypothesis to a greater than or less than
scenario if desired.

Example of the 1-Sample T Test

An agriculture company growing corn produced an average of 168 bushels of corn per acre
each year for five years. The company made a change to its seeding process in an attempt
to increase the yield of its fields. The harvest following the seed change resulted in an
average of 175 bushels of corn per acre each year. Did the seeding change result in a
statistically different outcome in yield per acre?

H0: There is no difference between the outcome after the seeding change and the
outcome before it.

Ha: The seeding change resulted in an increase in the average amount yielded per
field.

Used when comparing standard deviation or variance

Compares the standard deviation or variance between two samples of the same x
factor (usually one sample is taken after a change to the process has been made) or
compares the standard deviation or variance to a target or hypothesized statistic.
Is available in Minitab
o Requires sample data in two columns or summary data to include sample
size with either sample standard deviation or sample variance and a
hypothesized standard deviation or variance.

The hypothesis test is set up as:

378
 SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

Note that in the above set-up, the alternative hypothesis is -

Minitab allows you to adjust the alternative hypothesis to a greater than or less than
scenario if desired.

Example of the Chi-Square Test (or 1-Variance Test)

A manufacturing plant makes the wheels that are put on dolly carts. The plant provides
three sizes of wheels to a partner plant, which makes the dolly itself and uses the wheels to
complete the process. The partner plant makes a complaint: The smallest wheels coming
from the factory have enough variation in size to cause stability problems with the final
product, resulting in defective dollies or wasted wheel parts.

The wheel-making plant launches a Six Sigma improvement project to reduce the variation
of the size in the smallest wheels. The team measures for a baseline and finds that the
standard deviation in the size of the wheels is 0.0409. The team makes changes to the
process and measures again to compare the new standard deviation to the old.

H0: There is no difference between the standard deviations; the changes did nothing
to statistically reduce the variation in the process.

Ha: The improvements reduced the standard deviation in the process.

Used when comparing means

Compares the means between two samples of the different x factors
Is available in Minitab and Excel
o Requires sample data in two columns or summary data to include sample
size, sample mean, and standard deviation for both samples
Requires you to note whether equal variances between the two samples are
assumed or not

379
 HYPOTHESIS TESTING

The 2-Sample T Test might seem like the 1-Sample T Test, because in the 1-Sample T Test,
we discussed how you might compare a sample from before a change to a sample from
after a change. However, the 2-Sample T Test is slightly different, because you are
comparing two samples from different populations. For example, you might compare the

temperature for children who did receive the vaccine. This is different than comparing

receive the vaccine, to this year, when they did receive the vaccine. The first example would
use the 2-Sample T Test; the second (year-over-year) example would use the 1-Sample T
Test.

The hypothesis test is set up as:

Note that in the above set- -

Minitab allows you to adjust the alternative hypothesis to a greater than or less than
scenario if desired.

Example of the 2-Sample T Test

Two attorneys practicing in the same law firm are turning in very different amounts of
billable hours, even though the partners note that the two lawyers have similar caseloads.

tasks; because the firm bills at a certain time for administrative-style tasks regardless of
how long those tasks take, if legal staff is taking too long for such tasks, they are spending

To test the theory, data is gathered about the time legal staff spends on tasks. Specifically, a
team looks at time spent on three tasks for which the legal staff is allowed to bill 15 minutes
each: certain types of simple administrative phone, copying, or filing tasks.

H0
f takes for such tasks.

380
 SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

Ha

The Chi-Square Test, which was described in the above section, can also be used to compare
standard deviation between a sample and a hypothesized standard deviation when data is
not normal.

Used when comparing medians

Compares the medians between a sample and a hypothesized sample or a new
sample to a previous sample before changes were made
Is available in Minitab
o Requires sample data in a column and hypothesized mean
Typically used when data is somewhat symmetrical; otherwise, you can use the 1-
Sample Sign test in Minitab for the same thing

The hypothesis test is set up as:

Note that in the above set-up, the alternative hypothesis -

Minitab allows you to adjust the alternative hypothesis to a greater than or less than
scenario if desired.

Example of the One-Sample Wilcox

A real estate company wants to enter a high-dollar market, so they want to build some
regional websites that cater to homebuyers looking for larger homes. Specifically, they want
to cater to homebuyers who are purchasing homes with 4 or more bedrooms. Before the
company can build websites and begin marketing, it needs to find neighborhoods where
homes for sale typically have 4 or more bedrooms.
381
 HYPOTHESIS TESTING

The real estate company gathers information about the median number of bedrooms in
homes in a specific area.

H0: The median number of bedrooms is equal to 3.

Ha: The median number of bedrooms is greater than 3 (4 or more).

Used when comparing medians

Compares the medians between samples of two factors of x
Is available in Minitab
o Requires sample data in columns

The hypothesis test is set up as:

Note that in the above set- -

Minitab allows you to adjust the alternative hypothesis to a greater than or less than
scenario if desired.

Example of the Mann-Whitney Test

Consider the real estate company example above again. Now, imagine that the real estate
company has limited resources, so they can only launch a campaign for one neighborhood
at a time. To maximize potential results of their first campaign, the company wants to
choose the neighborhood with the highest number of bedrooms in the highest number of
homes.

The company gathers information about homes in two neighborhoods. They then compare
the samples to determine whether one sample has a statistically higher median number of
bedrooms.

382
 SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

H0: The median number of bedrooms in neighborhood A is equal to the median

number of bedrooms in neighborhood B.

Ha: The median number of bedrooms in neighborhood A is higher than that in

neighborhood B.

At this point, a common question about hypothesis tests comes up: Why go through the
trouble to run statistical tests to determine if things are different?

Consider the example problem presented for the 1-Sample T Test:

An agriculture company growing corn produced an average of 168 bushels of corn per acre
each year for five years. The company made a change to its seeding process in an attempt
to increase the yield of its fields. The harvest following the seed change resulted in an
average of 175 bushels of corn per acre each year. Did the seeding change result in a
statistically different outcome in yield per acre?

The question is: Is the mean corn yield per acre statistically higher following the seed
change? Just looking at the numbers presented in this problem, we can see that the mean
following the seed change is 175. When you compare that to 168, it is obviously higher.
Why would we go through the trouble of setting up and running a hypothesis test?

When introducing the normal curve in chapter 18, we noted that it

see that a histogram appears to be symmetric and following the normal curve. Raw data can
be deceptive, which is why we ran the Chi-Squared Goodness-of-Fit test in Excel to ensure
the data was normal. The same principle is true when comparing statistics from samples.
Yes, 175 is higher than 168. But in the case of the corn yield is it statistically different?

statistically different that we can take action on this

difference between 10 and 23 might not be statistically different. In another, the difference
between 10 and 10.5 might be statistically different.

383
 HYPOTHESIS TESTING

To run a hypothesis test, you typically follow the same series of steps:

1. State the null and alternative hypotheses. The null hypothesis is always written
from the perspective that no change or difference occurs or is present. The
alternative hypothesis is always written from the perspective that a change or
difference is present either not equal, greater than, or less than.
2. Set the confidence level for alpha. Usually, the confidence level is set at 95 percent
for alpha, but other common settings are 99 and 99.9 percent. A confidence level of
95 percent means an alpha value of 0.05; a confidence interval level of 99 percent
means an alpha value of 0.01; a confidence interval of 99.9 percent means an alpha
value of 0.001.
3. Decide which hypothesis test you are going to use. Use the information in the above
section to find the appropriate test based on:
a. The type of data you have
b. The statistic you are dealing with (mean, variation, etc.)
c. How many sets of data you have (level of interest in x)
d. Remember, testing more than two sets of data is covered in Unit 7 on
Design of Experiments and ANOVA
4. Decide whether your sample size is fixed or whether you can select a sample sized
based on your beta setting. Setting sample size is covered in Chapter 22.
5. Run the test in Minitab. Specific steps for each hypothesis test as well as other
Minitab functions are covered in Unit 6.
6. Interpret the p-value against your alpha setting, which tells you whether or not to
reject the null hypothesis.
7. Translate the statistical analysis into real-world, business-relevant language.

For each statistical test, the calculations handled automatically for us in Minitab or other
software programs return several values. Some calculations, such as the Chi-Squared
tests, return specific statistics. In the case of the Chi-Squared tests, that statistic is the Chi-
Squared value. These statistics can be valuable in applications and analyses, but for the
purpose of hypothesis testing, we are primarily concerned with the p-value. The p-value is
returned for each test, and we compare the p-value to the alpha value we set before we ran
the test.

If the p-value returned for a test is less than the alpha value you set, then you reject the null
hypothesis and accept the alternative hypothesis. If the p-value returned is more than the

384
 SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

alpha value you set, then you fail to reject the null hypothesis and you do reject the
alternative hypothesis.

Above, we talked about setting a confidence interval for alpha. If the confidence interval is
95 percent, then the alpha value is 0.05. Given a p-value of 0.02, we would reject the null
hypothesis and accept the alternative. Given a p-value of 0.13, we would fail to reject the
null hypothesis and would reject the alternative.

example tests from the section above.

Problem: A school administrator believes that the students in a certain classroom are being
impacted by some environmental issue with that room. The administrator wants to see if
the students in that room perform statistically lower on tests than the students throughout
the school; she has already noted that different classes in that room are taught by other
teachers.

The administrator gathers data. The rate of failure for classes taught in any other classroom
is 20 percent. Out of 142 students who have a class in the classroom in question, 38 are
failing that particular class.

Step 1: State Null and Alternative Hypothesis

H0: There is no difference between the failure rate of students in the classroom and those in
other classrooms.

Ha: The failure rate of students in the classroom is statistically higher than the failure rate of
students in other classrooms.

Step 2: Set Confidence Level

Unless otherwise stated in this book, the confidence level will always be 95 percent, or
alpha = 0.05.

Step 3: Select Appropriate Test

385
 HYPOTHESIS TESTING

Our data takes the form of pass/fail, which is attribute or discrete data. We have one set of
data that we are comparing to the hypothesized rate of failure of 20 percent. Given that
information, the 1-Proportion Test is the appropriate test.

Step 4: Select Sample Size

data. Since Six Sigma teams usually gather their own data, they need to worry about the
proper sample size. How to determine sample size is covered in the next chapter.

Step 5: Run the Test and Analyze P-Value

The test is run in Minitab with the following information:

- Number of trials: 142 (there are 142 students)

- Number of events: 38
- Hypothesized proportion: 0.20 (20 percent of students in other classrooms meet the

- Alternative hypothesis: proportion > hypothesized proportion (we are testing whether
the fail rate of students within the classroom is higher than the 20 percent fail rate of
those outside the classroom)

Given all of the information above, Minitab returns a p-value of 0.031.

Since the p-value is less than the alpha value of 0.05, we reject the null hypothesis and
accept the alternative. Yes, the rate of failure is statistically higher for students in the
classroom.

386
 SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

Problem: An agriculture company growing corn produced an average of 168 bushels of corn
per acre each year for five years. The company made a change to its seeding process in an
attempt to increase the yield of its fields. The harvest following the seed change resulted in
an average of 175 bushels of corn per acre each year over 500 acres of fields. The standard
deviation is 2.2 bushels. Did the seeding change result in a statistically different outcome in
yield per acre?

Step 1: State Null and Alternative Hypothesis

H0: There is no difference in the mean yield after the seed change as compared to before
the seed change.

Ha: Following the seed change, the acres produce a statistically higher mean yield of corn.

Step 2: Set Confidence Level

Again, we set the confidence level for alpha at 95 percent, so alpha is 0.05.

Step 3: Select Appropriate Test

We are testing the mean of a sample against a hypothesized (or historic, or population)
mean. Assume for the purpose of this example that our data is normal. The correct test
would be the 1-Sample T Test or th

Step 4: Select Sample Size

data. Since Six Sigma teams usually gather their own data, they need to worry about the
proper sample size. How to determine sample size is covered in the next chapter.

Step 5: Run the Test and Analyze P-Value

The test is run in Minitab with the following information:

- Sample size: 500 (there were 500 acres of corn fields)

- Sample mean: 175 (following the seed change, the 500 acres yielded an average of 175
bushels)
- Standard deviation: 2.2 (provided in our problem)
- Hypothesized mean: 168 (prior to the seed change, the average yield was 168 bushels)

387
 HYPOTHESIS TESTING

- Alternative hypothesis: mean > hypothesized mean (we are testing whether the
average yield after the seed change is statistically higher than the yield prior to the
seed change).

Given all of the information above, Minitab returns a p-value of 0.000.

Since the p-value is less than the alpha value of 0.05, we reject the null hypothesis and
accept the alternative. Yes, the mean yield is statistically higher following the seed change.
It should be noted that the p-value in this case is exceptionally low. For our purposes, it is
zero. That means there is only a very tiny chance of a Type 1 hypothesis testing error in this
case. Why is this important? When making a decision that involves resources of money,
leadership is always happy to hear that analysts are especially confident in their
conclusions.

Problem: Two attorneys practicing in the same law firm are turning in very different
amounts of billable hours, even though the partners note that the two lawyers have similar

handle some tasks; because the firm bills at a certain time for administrative-style tasks
regardless of how long those tasks take, if legal staff is taking too long for such tasks, they

To test the theory, data is gathered about the time legal staff spends on tasks. Specifically, a
team looks at time spent on three tasks for which the legal staff is allowed to bill 15 minutes
each: certain types of simple administrative phone, copying, or filing tasks. The following
data is collected.

388
 SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

Minutes Spent on Quarter-Hour Tasks

Attorney A's Team Attorney B's Team
10 8
8 7
15 10
16 12
17 18
5 6
8 9
16 12
12 11
11 11
8 9
9 8
15 14
17 15
22 16
25 12
30 19
2 8
9 12
4 6
5 8
7 11
8 12
12 13
13 15
15 17
16 13
18 15
19 15
20 12
22 21
389
 HYPOTHESIS TESTING

25 32
15 18
14 11

Step 1: State Null and Alternative Hypothesis

H0 -

Ha to perform quarter-hour tasks than Attorney

Step 2: Set Confidence Level

Again, we set the confidence level for alpha at 95 percent, so alpha is 0.05.

Step 3: Select Appropriate Test

We are testing the mean of one sample against the mean of another. For this problem, we
will assume the data was tested to be normal. The data is time, so it is continuous. The
appropriate test is the 2-Sample T test.

Step 4: Select Sample Size

provided with the

data. Since Six Sigma teams usually gather their own data, they need to worry about the
proper sample size. How to determine sample size is covered in the next chapter.

Step 5: Run the Test and Analyze P-Value

The test is run in Minitab by indicating the two columns with the same data in it and setting
the alternative hypothesis to greater than.

Given all of the information above, Minitab return a p-value of 0.255

390
 SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

Since the p-value is greater than the alpha value of 0.05, we fail to reject the null hypothesis
and we do reject the alternative. The average time spent on the quarter-hour tasks is not
statistically different for each team. Minitab also calculates the means for us:

average of 12.82 minutes on each task. Both teams, in fact, spend less time on the tasks
than they can bill for the tasks, which means neither team is technically wasting any time.

Problem: A real estate company wants to enter a high-dollar market, so they want to build
some regional websites that cater to homebuyers looking for larger homes. Specifically,
they want to cater to homebuyers who are purchasing homes with 4 or more bedrooms.
Before the company can build websites and begin marketing, it needs to find
neighborhoods where homes for sale typically have 4 or more bedrooms.

The real estate company gathers information about the number of bedrooms in homes in a
neighborhood. That data is presented below.

Number of Bedrooms
4
4
4
4
4
5
5
5
391
 HYPOTHESIS TESTING

5
5
5
5
5
5
5
3
3
3
3
3
3
3
3
3
6
6
6
6
6
6
6
6
7
7
5
5
5
5
4
4
4
4

392
 SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

3
3
3
2
2
2
2
1
1
1

Step 1: State Null and Alternative Hypothesis

H0: The median number of bedrooms is equal to 3.

Ha: The median number of bedrooms is greater than 3.

Step 2: Set Confidence Level

Again, we set the confidence level for alpha at 95 percent, so alpha is 0.05.

Step 3: Select Appropriate Test

We are testing the median of a sample against a hypothesized median. We only have one
factor for X, so we will use the One Sample Wilcox.

Step 4: Select Sample Size

been provided with the

data. Since Six Sigma teams usually gather their own data, they need to worry about the
proper sample size. How to determine sample size is covered in the next chapter.

Step 5: Run the Test and Analyze P-Value

The test is run in Minitab by indicating the column where the sample data resides, notating
a hypothesized median of 3, and setting the alternative hypothesis to greater than 3.

Given all of the information above, Minitab returns a p-value of 0.000.

393
 HYPOTHESIS TESTING

Since the p-value is less than the alpha value of 0.05, we reject the null hypothesis and
accept the alternative. Yes, the median number of bedrooms for homes in the neighborhood
is 4 or more.

For Six Sigma teams, hypothesis testing is an activity typically found in the latter three
phases of a DMAIC project, and most specifically in Analyze or Improve. During Analyze,
teams might use hypothesis testing to help determine if certain inputs or factors could be a
root cause. Given historical data, teams who can separate factors within a process can use
that data to run a hypothesis test. For example, a team might believe that one employee is
responsible for defects in a process. By comparing defect rates when that employee is on
shift to defect rates when the employee is not working, teams can draw conclusions from
data.

During Improve, hypothesis tests are often used to validate solutions that are being tested
in beta environments or in small batches. The team compares the results of that solution
with previous data about the process to determine whether the solution has created a
statistically different result and whether that result is what the team wanted.

Because hypothesis tests can begin to sound ve

a business-friendly statement and end with a business-relevant statement regarding the
conclusions. Chapter 24 covers some tips for converting various Six Sigma and statistical
conclusions to business-friendly presentations and knowledge, which is especially helpful
when a Six Sigma team is proposing changes based on statistical analysis. The ability to
convert that analysis into business-friendly language helps teams explain to decision-makers
why a change is likely to help.

394
 SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

This chapter introduced a lot of new statistical terms, many of which will come up again in
future chapters. Before moving to the next chapter, take a few minutes to review the terms
that came up in this chapter.

Null Hypothesis: The assumed hypothesis; the statement that nothing has changed or no
statistical difference exists. This is like the innocent verdict in a criminal court: it is
assumed unless evidence tells us otherwise.

Alternative Hypothesis: The statement that something has changed or is statistically

different can be framed as not equal, greater than, or less than. If evidence means
rejecting the null hypothesis, then the alternative hypothesis is assumed to be true.

Alpha: The measurement of the risk of a Type I error the error that occurs if the null
hypothesis is rejected when it was actually true.

Beta The measurement of the risk of a Type II error the error that occurs if the null
hypothesis is not rejected when it was actually false.

Test Statistic A standard value that is used to calculate the p-value to determine

screen and can be valuable for more advanced statistical calculations. When using
Minitab for hypothesis testing, the p-value is automatically generated.

P-Value: The number typically used to compare to the alpha value to determine whether
or not to reject the null hypothesis. The null hypothesis is rejected if the p-value is less
than the alpha value.

395
Accuracy in inferential statistics requires that you have the right sample size. A retail
company with thousands of customers would never make a major product decision based
on feedback from five people, and Six Sigma teams should never make decisions based on
statistical analysis when the sample size is not big enough to create an acceptable margin of
error.

To understand the importance of sample size

sampling as related to inferential statistics.

Six Sigma teams take samples because they want to determine information about a
population. With some exceptions, it is very difficult, very expensive, or impossible to run
statistical calculations on the entire population of data. However, when we sample the
population and draw inferences from those samples about means, proportions, or
variances, there is always some risk of error. Some level of uncertainty will always exist
when we draw conclusions about a population from sample data. In many cases, a larger
sample size helps reduce the uncertainty of our conclusions.

The uncertainty associated with sampling is defined by something called the Confidence
Interval, which is also called the margin of error in som
seen survey or poll results reported with a margin of error: The fans are 93 percent in favor
of the new team colors, +/- 2 percent. The margin of error in that particularly survey was 2
percent, or 0.02.

In chapter 21, we discussed hypothesis testing and we set our alpha value at 0.05, which
means we were at least 95 percent confident in choosing to reject or not reject the null
hypothesis. We can make our hypothesis tests even more accurate by defining power and
sample size for our data particularly when we are pulling the data ourselves and can
choose how many samples to take.

396
 SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

Download Minitab
Minitab is one of several advanced statistical analysis software program
available, and it is a software that is commonly used by Six Sigma experts.
For this chapter, we will work extensively in Minitab to perform calculations
and analyses.

Screenshots are included for all of the analysis completed in Minitab going
forward. If you would like to follow along or experiment with your own data
in Minitab, you can download a trial version of the program from
www.minitab.com. As of early 2016, the free trial is offered for 30 days. If
you are part of an organization that uses Six Sigma for process
improvement, you might also be able to gain access to Minitab through
your employer.

The concept of hypothesis testing errors is key to selecting sample sizes for various
hypothesis testing. The two types of errors were introduced in Chapter 21, but warrant a
quick review before moving on with sample size calculations.

A Type I error occurs when you reject the null hypothesis during a hypothesis test when, in
fact, the null hypothesis is true. You might reject the null hypothesis that the mean of the
sample is statistically the same as the hypothesized mean, deciding instead that the sample
mean is statistically greater than the hypothesized mean. If the sample mean is, in fact,
statistically the same as the hypothesized mean, you have a Type I error.

Type I error risks are denoted by alpha.

As stated in chapter 21, the Type I error is also called the producer risk. In a manufacturing
environment, if the null hypothesis is true but it is rejected due to a Type I error, then
material that is within specification (material that is statistically the same as the target) will
be rejected. This is a producer risk it causes waste, extra cost, and lower employee
morale.

397
 SAMPLE SIZE

A Type II error occurs when you accept the null hypothesis when it is, in fact, not true. If you
accept the hypothesis that the sample mean is statistically the same as the hypothesized
mean when, in fact, the sample mean is statistically greater than the hypothesized mean,
then you have a Type II error.

Type II error risks are denoted by beta.

As stated in chapter 21, the Type II error is also called the consumer risk. In a manufacturing
environment, if the null hypothesis is false but it is accepted anyway due to a Type II error,
then material that is not within specification (material that is not statistically the same as
et is passed onto the
consumer. That is a consumer risk it could cause returns, unsatisfied customers, poor
brand reputation.

The appropriate sample size can reduce the risk of either type of error.

To perform sample size calculations or have Minitab perform them for you certain
information and assumptions are required.

Alpha The alpha level you set remember, the default in Minitab
is typically set at 0.05

Beta The beta level can be set by the experimenter and a sample
size calculated from that number. If the sample size is fixed,
then the experimenter usually sets alpha and calculates the
beta risk from the sample size.

Delta The practical difference the experimenter wants to detect

using the test.

Standard The estimated population standard deviation.

deviation

Type of Data Discrete or continuous?

398
 SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

Type of Test Which hypothesis test are you using (see Chapter 21 for
selecting the right test)

Before you can move onto calculating sample size, some practical considerations must be
made with regard to alpha, beta, and delta. In our review of Type I and Type II errors, we

work in all organizations or with all hypothesis tests.

Because the outcome of your hypothesis tests are going to be used to make real-world
decisions, you have to truly understand what the test and its outcomes means for the
business, employee, and consumer. That real-world understanding helps you set
appropriate Alpha and Beta values.

In most cases, you can set the alpha value at 0.05. However, in specific experiments,
accuracy is extremely important. In medical or drug trials, for example, alpha values could
be set to 0.99 or 0.999, making it very unlikely that a researcher would accept an alternative
hypothesis (that a drug caused a change, for example) if it was not correct.

When selecting alpha, Six Sigma teams should ask:

- What are the costs associated with an unnecessary change if the team makes a
mistake in rejecting the null hypothesis?
o In a manufacturing environment, what are the costs with rejecting
materials that actually fit specifications?
o In a non-manufacturing environment, what are the costs associated with
accepting the hypothesis that change did occur?
Are there dangers or costs associated with concluding that a
statistical change occurred? What are they?

When selecting a beta value, Six Sigma teams should ask:

- What is the potential costs of a Type II error if the team makes a mistake in not
rejecting the null hypothesis?
o What is the potential damage or cost if defective materials are passed to
the customer? If a defective spoon is passed along, the ultimate cost might
be minimal compared to a defective piece of medical equipment or car
engine.
399
 SAMPLE SIZE

o Are there costs associated with lost time or resources in correcting a

problem that comes from a Type II error?

When selecting a value for delta, Six Sigma teams should ask:

- How small does the difference have to be before it becomes practically insignificant to
the customer? In a pizza restaurant, for example, a customer who orders a 14-inch
pizza is going to care if he or she receives a 12-inch pizza. If the pizza that is delivered
measures 13.9 inches, the customer is unlikely to care.
- What is the smallest delta that provides the best chance at exposing all benefits or
information but is not so small as to be unfeasible. Working at smaller delta values
means larger sample sizes or smaller measurement requirements, which take time to
gather.
- Ultimately, Six Sigma teams should ask themselves: what margin of error is tolerable in
results? The type of test you are running, the decisions you are making, and the
processes you are dealing with help you make this determination.

The delta, or critical difference, should always be based on business needs. It is helpful in
Minitab to express the delta in terms of sigma, or standard deviation. We want to detect
differences of .5 or 2 , for example.

While values for various factors in sample size calculations should be guided by the business
needs and purposes of the test, here are some guidelines Six Sigma teams can follow for
baseline values or if no other information or needs are present.

If you have some information about the population, including standard deviation:

- Set alpha at 0.05.

- Set beta at 0.10 or 0.20 (which will give you a power of 0.90 or 0.80 a power of 0.90 is
less likely to produce a Type II error than a power of 0.80, but it will also require a much
larger sample size in most cases).
- Set delta with absolute values as required by business needs or, more often in Minitab,
as a function of standard deviation (.5

400
 SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

- Set alpha at 0.05.

- Set beta at 0.10 or 0.20 (which will give you a power of 0.90 or 0.80 a power of 0.90 is
less likely to produce a Type II error than a power of 0.80, but it will also require a much
larger sample size in most cases).
- Set delta greater than 1 and according to business needs, or, more often in Minitab, as
a function of standard deviation (.5

If you have some information about the population, including standard deviation:

- Set alpha at 0.05.

- Set beta at 0.10 or 0.20 (which will give you a power of 0.90 or 0.80 a power of 0.90 is
less likely to produce a Type II error than a power of 0.80, but it will also require a much
larger sample size in most cases).
- Set delta logically according to business needs, or, more often in Minitab, as a function
of standard deviation (.5

Note: General sampling on an unknown population, also called survey sampling, requires a
different approach to sample-size calculation. Typically, you begin by setting alpha to 0.05,
beta to 0.5, and delta to a proportion in keeping with a standard deviation of 1. This
provides an initial sample size so you can better understand the population, but you might
have to do additional sampling based on the statistics calculated from your first sample.

Just as there were numerous hypothesis tests to cover a range of data types and questions,
there are numerous sample size calculations all of which can be performed in Minitab.
Review the list below to determine what type of calculation to perform when determining
sample size for various hypothesis tests.

1-Sample T Test Sample Size Calculation

- Means testing
- Comparing to a target value
- You already have sample statistics about the population

401
 SAMPLE SIZE

1-Sample Z Test Sample Size Calculation

- Means testing
- Comparing to a target
- You do not have sample statistics about the population (standard deviation is not
known)

2-Sample T Test Sample Size Calculation

- Means testing
- Comparing means from two sets of data

2-Sample Variance Test Sample Size Calculation

- Variance testing
- Comparing variance from two sets of data

1-Sample Variance Test Sample Size Calculation

- Variance testing
- Comparing variance of one data set to a target

1-Sample Proportion Test Sample Size Calculation

- Proportion testing (rate, x per y)

- Comparing rate of one data set to a target

2-Sample Proportion Test Sample Size Calculation

- Proportion testing (rate, x per y)

- Comparing rate from two sets of data

Analysis of Variance (ANOVA) Sample Size Calculation

- Means testing
- More than 2 sets of data
- Only one factor for x

Design of Experiment (DOE) Sample Size Calculation

- Means testing
- More than 2 sets of data
402
 SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

- More than one factor for x

To understand more about running and analyzing sample size tests

through the process for most of the test types above.

To find the Power and Sample Size calculation tools in Minitab, select Stat > Power and
Sample Size

-Sample Z power and sample size calculation. Remember, you would

know the standard deviation for the process or population. This calculation provides you
a sample size that provides you some relevant information that
is as accurate as you define and that you can use as a basis for future sampling.
403
 SAMPLE SIZE

Selecting the 1-Sample Z calculation from the menu brings up the following dialogue box.

Minitab lets you calculate sample sizes or back into power values if a sample size is fixed,
which is why it asks you to fill in only two of the three first values.

several parameters for differences and power values for Minitab. This causes Minitab to
provide a number of answers one for each difference and power value combination so
we can make a more viable decision about sample size.

lues at 0.8
and 0.9 (or, beta to either 0.2 or 0.1 respectively). Standard deviation for this first pass is
assumed to be 1; if we had information about standard deviation, we could run a more
precise calculation or use the 1-Sample T Test calculation.

If you click options, you can change the confidence level or alternative hypothesis (to a
greater or less than scenario).

404
 SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

After we click OK on the Power and Sample Size box, Minitab returns the following
information.

1-Sample Z Test

Calculating power for mean = null + difference

Sample Target
Difference Size Power Actual Power
0.5 32 0.8 0.807430
0.5 43 0.9 0.906375
1.0 8 0.8 0.807430
1.0 11 0.9 0.912556
2.0 2 0.8 0.807430
2.0 3 0.9 0.933727

The data table returned by Minitab shows us the required sample size for each difference
and target power. You can see that for the lowest difference and highest target power, the
sample size required is only 43. But to determine a difference of 1 at a target power of
0.9, you only need 11 in the sample size.

Here, the team has to make a business-centric decision. Perhaps samples are easy and
inexpensive to capture, so the team selects the maximum sample size. But if each sample
takes an hour to obtain and costs the company $26, then the team might make the call to
select a smaller sample size.

405
 SAMPLE SIZE

Again, it is important to note with this particular test that more sampling might be required
to draw further conclusions.

To see the 1-
21.

An agriculture company growing corn produced an average of 168 bushels of corn per acre
each year for five years. The standard deviation is 2.2 bushels. The company made a change
to its seeding process in an attempt to increase the yield of its fields. Did the seeding
change result in a statistically different outcome in yield per acre?

How many acres would the company need to sample to determine if the new process made
a difference? This could help the company know how many acres to plant with the new
seed process.

To generate data for this problem, select the 1-Sample T calculation from the Power and
Sample Size menu in Minitab.

For this calculation, enter the differences as a multiple of standard deviation, the power
values as desired, and the standard deviation provided in the data.

406
 SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

Change the alternative hypothesis to greater than. The company wants to determine
whether the yield of the new process is greater than the average yield of the old process.

After you click OK, and then OK again, Minitab generates the following set of information.

We can see from the above table that the company would have to plant 168 acres to
achieve a target power of 0.9 and be able to discern differences of 0.5 standard deviations.
y might
weigh its options and resources and decide, instead, to plant 43 or 32 acres. By planting 43
acres, the company can detect differences as low as one standard deviation at 0.9 target

407
 SAMPLE SIZE

power. Given that the standard deviation is assumed to be 2.2 bushels, that is still a
relatively small detection range, and the company saves a ton of money by planting over
100 fewer acres for the test.

d
plant for a test? Finances, resources, or corporate leadership might collude to decide that
the test could be run on a total of 18 acres and only 18 acres.

Minitab can take the information about the test and calculate the target power. The test is
run the same way, but sample size is entered and power values are left blank.

408
 SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

You can see from the above data that, given a sample size of 18, the power for detecting a
difference of 0.5 beta. In this case, beta (the risk of
making a Type II error) is over 0.75.

approximately 4.4 in this case the power is

extremely high. The team in this case would likely conclude that planting on the 18 sample
acres would provide an appropriate sample for running the hypothesis test.

Proportion tests are a bit different for two major reasons:

- Proportion tests are run using attribute data. Attribute data almost always requires a
larger sample size for accurate results than continuous data does.
- Because you are dealing with attributes and rates
information about population parameters such as sigma levels or standard deviation.

To calculate the sample size required for accurate 1-Sample Proportion hypothesis testing,
you simply set alpha and beta and enter the proportions for both the null and alternative
hypothesis.

real-world example to better understand the calculations in Minitab.

A grocery store chain discovers it has a problem with prices being entered correctly into the
system when product is placed in the store or prices are updated. In population sampling,
the store found that approximately 8 percent of the prices are entered incorrectly. The
store provided ½ of the employees who do pricing entry with new scanners to attempt to
mitigate the problem. Before the store commits to a full-time lease on the new equipment
for all employees, it wants to confirm that the new scanners bring the error-rate down by 5
percent to 3 percent or less.

409
 SAMPLE SIZE

Select the 1-Proportion calculation from the Power and Sample Size menu in Minitab. The
comparison proportion is entered as 1 the probability of failure for a single trial. In this
case, 1 0.03, since the company wants to reduce the error rate to 3 percent.

The hypothesized proportion is entered as 1 the probability of a failure for a single trial, or
1 0.08, since the original process had an 8 percent failure rate.

Click OK and then OK, and Minitab generates the information seen below.

410
 SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

To run a valid 1-Proportion test, the company needs to gather between 140 and 177 data
samples.

Consider again the 2-Sample T test problem from the last chapter:

Two attorneys practicing in the same law firm are turning in very different amounts of
billable hours, even though the partners note that the two lawyers have similar caseloads.

tasks; because the firm bills at a certain time for administrative-style tasks regardless of
how long those tasks take, if legal staff is taking too long for such tasks, they are spending

To test the theory, data is gathered about the time legal staff spends on tasks. Specifically, a
team looks at time spent on three tasks for which the legal staff is allowed to bill 15 minutes
each: certain types of simple administrative phone, copying, or filing tasks. The following
data is collected.

Minutes Spent on Quarter-Hour Tasks

Attorney A's Team Attorney B's Team
10 8
8 7
15 10
16 12
17 18
411
 SAMPLE SIZE

5 6
8 9
16 12
12 11
11 11
8 9
9 8
15 14
17 15
22 16
25 12
30 19
2 8
9 12
4 6
5 8
7 11
8 12
12 13
13 15
15 17
16 13
18 15
19 15
20 12
22 21
25 32
15 18
14 11

For the purposes of this problem, the firm team collected a sample of 34 data elements
from each team. But is that enough to provide a strong conclusion in the 2-Sample T Test?

ta in question.
The standard deviation of the data is 5.83, and we want to test whether the mean for
412
 SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

-
Sample T Test in Minitab as pictured below.

Minitab generates the following information:

413
 SAMPLE SIZE

You can see above, the minimal sample size for each group that Minitab calculates is
106, and that is for a difference of 2 with target power of 0.8. While a 2-Sample T test can
be run on the data set provided above and we did, in fact, run that test in chapter 21 the
likelihood of a type II error for that test was fairly high.

Just how high was the likelihood for an error running the 2-Sample T test with a sample size
of 34 for each group? Minitab can tell us that answer.

Instead of entering power values in the power and sample size dialog box, enter the sample
size of 34 and click OK.

414
 SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

The highest power for a sample size of 34 given this data is 0.40, which means a beta of 0.60
more than 50 percent chance of a Type II error in the hypothesis test. In this particular
case, before the firm makes any decisions, they should gather a larger sample of data.

random sampling. Even if you perform a calculation to determine an appropriate sample

size and you choose to ere on the side of caution by sampling at the largest count size

415
 SAMPLE SIZE

returned by Minitab you can draw the wrong inferences because the sample was not
random.

416
The concept of control charts was introduced in the DMAIC chapter on control. Chapter 17
provided instructions for creating a specific type of control chart the x-bar chart in Excel

remember it took many steps and some basic graphical formatting to create a full-fledged x-

leadership wants to see an update on a process, and certainly not something day-to-day
analysts want to do for regular reporting.

Luckily, Minitab offers more automated x-bar control charting. The statistical software also
offers easy access to other types of control charting. After the chapters on hypothesis
testing and sample size calculations, it should come as no surprise that there are many
types of control charts, each of which are useful in monitoring different types of data and
ts and when to use
them before defining terms for each type of chart and walking through creating the charts
in Minitab.

417
 ADVANCED CONTROL CHARTS

- Data is variable (continuous)

- Data can be grouped into subgroups, letting you chart the mean of each group
o Production mean per hour
o Average answer speed of phones per hour
o Average number of customers each day
o Defects per hour
- Data count in each subgroup is less than 8 (although you can use the x-bar & R chart for
subgroup sizes up to 100 in Minitab)
- Presents two charts
o The x-bar control chart plots the mean of each subgroup
o The R chart plots the range

418
 SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

- Data is variable (continuous)

- Data can be grouped into subgroups, letting you chart the mean of each group
- Data count in each subgroup is more than 8
- Sigma can be easily calculated (otherwise, use x-bar &R)
- Presents two charts
o The x-bar control chart plots the mean of each subgroup
o The S chart plots the standard deviation

419
 ADVANCED CONTROL CHARTS

- Data is variable (continuous)

- Data cannot be easily grouped into reasonable subgroups, so you must track individual
data points instead of subgroup means
o Data is very difficult or expensive to obtain
o Production is very slow, so waiting for enough data to create subgroups
would take too much time (for example, tracking data about surgeries
performed in a small outpatient facility)
o Products have a low cycle time

420
 SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

- Data is discrete
- Data is presented as a percent (percent defective) rather than a count
-
-
specifications)
- Use when:
o it is impossible, very time-consuming, or not financially feasible to measure
or analyze numerical measurements
o sample or subgroup sizes are not equal, so the percent of defect is more
important that the number of defects per grou
numbers in a smaller group with numbers in a larger group accurately)
o data is rate-based because it comes from a binomial or attribute process:

on a control chart, but you can plot the percent of 1 or 0 in each sample.

421
 ADVANCED CONTROL CHARTS

- Data is discrete
- Data is presented as a percent (percent defective) rather than a count
- Does require a constant sample size to be of use; other than this, you can use it for
anything that you would use a p-chart for
- Plots the number of units that are nonconforming in each sample size

422
 SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

- Data is discrete
- Data is presented as a count (number of defects)
- sample/subgroup size
- Plots the number of defects per unit
- Use when data is about the defects themselves, not the overall defective product

423
 ADVANCED CONTROL CHARTS

- Data is discrete
- Data is presented as a count (number of defects)
- Does require a constant sample/subgroup size
- Plots the number of defects per sample
- Use when data is about the defects themselves, not the overall defective product

To better understand the X-Bar & R Chart process in Minitab, consider a real-world
scenario. A call center is tracking the number of calls on hold. The calls center randomly
424
 SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

records the number of calls on hold at a given moment five times each hour. That data is
presented below for a 10-hour shift.

Calls On
Hour Hold
1 14
1 12
1 15
1 11
1 10
2 25
2 10
2 13
2 15
2 9
3 15
3 10
3 13
3 10
3 11
4 12
4 13
4 10
4 15
4 15
5 9
5 11
5 13
5 8
5 16
6 15
6 13
6 11
6 13
6 12
425
 ADVANCED CONTROL CHARTS

7 10
7 10
7 12
7 11
7 10
8 13
8 13
8 9
8 12
8 6
9 12
9 15
9 14
9 11
9 12
10 15
10 14
10 14
10 8
10 9

Copy this information into Minitab if you are following along with the analysis on your own.

25 for additional information.

All of the control charts are found under Stat > Control Charts

426
 SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

Select Stat > Control Charts > Variables for Subgroups > Xbar- R

427
ADVANCED CONTROL CHARTS

428
 SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

429
 ADVANCED CONTROL CHARTS

In the Xbar-

430
SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

431
 ADVANCED CONTROL CHARTS

This option tells Minitab to perform the control tests covered in chapter 16. If a data point
or series of data points on the control chart fails any of these tests, Minitab will highlight
them in red, making it easy for Six Sigma experts to tell if a process might be out of control.

Click OK on the options dialogue box.

Click OK on the Xbar-R chart dialogue box.

Minitab generates a chart and a text description for any failed control tests.

432
 SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

TEST 1. One point more than 3.00 standard deviations from center line.
Test Failed at points: 2

You can see above, the control chart for the range failed at point 2 one point is beyond 3
standard deviations from the center line. All other points are within control parameters,
and the sample mean (x-bar) chart is also within control. That means that the average
number of calls on hold within each hour was controlled. However, without hour 2, the
specific data points ranged much more than in other hours, which means one or more
samples collected was very high.

Someone viewing this chart would not likely consider the process to be completely out of
control, but they would want to investigate the samples for hour 2. What happened to
make a drastic change in the range? Was a number recorded incorrectly? Did several

433
 ADVANCED CONTROL CHARTS

employees take short, unplanned breaks that caused a sudden increase in calls on hold?
Control charts can point to small problems such as these before they become bigger,
systemic problems.

The steps for creating an X-bar & S chart are the same as those described above, except you
would start with Stat > Control Charts > Variables for Subgroups > Xbar- S

The I-MR
uping or it is
important for some reason to view the chart at the data-point level.

For example, a medium-sized car dealership wants to measure the number of sales closed
g per hour

-MR chart is a
good option.

To work along with this example, copy the sales data below into Minitab.

Sales closed
daily
19
17
17
17
20
23
22
18
19
19
18
20
21
19
24

434
 SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

16
15
21
21
20
19
20

Select Stat > Control Charts > Variables for Individuals > I-MR

Complete steps similar to the steps completed for the X-bar & R chart:

1.
2.
3. Click on I-MR Options
4. Click on Tests
5.
6. Click OK
7. Click OK

Minitab generates the I-MR chart and data about whether any test was failed.

435
 ADVANCED CONTROL CHARTS

TEST 1. One point more than 3.00 standard deviations from center line.
Test Failed at points: 16

In this particular case, the number of sales per month, which is depicted on the top
individual chart, are within control. The moving range how much change there is between
the numbers for each day is in control except for point 16. If you look at the data, you see
that between point 16 and point 17, there is an 8-sale difference. This is the point with the
biggest difference in day-to-day data points, which likely accounts for the test failure.

A car dealership or Six Sigma team working with this data would probably not be too
concerned about that one point. When you consider it in practical terms, it could be that

numbers artificially and caused the sudden wider-than-normal spread. In this case, the

436
 SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

dealership or Six Sigma team would simply note the point and keep an eye on the control
chart for similar test failures in the future that might indicate some type of pattern or
problem with the process.

Remember, a p- -
used attribute-based control charts because rates are often easier to work with than actual
the same
sample or subgroup size. You can also combine multiple types of defects in the same report,
because p-charts are typically concerned with whether the product, part, or test passed or
failed not with the specifics about the defects that caused the pass or fail.

The type of data that is reported at a management level also tends to be complemented by
the p-Chart characteristics. Executive leadership often want to know audit results, first-run
yields, scrap rates, or defect rates, all of which can be illustrated with p-Charts.

Proportion data is usually one of the first available to Six Sigma teams even before a Six
Sigma team is associated with a process, many business operators and managers capture
pass/fail type data. This means that a Six Sigma team might be able to create a first baseline
control chart from proportion data and see rates improve as process improvements are
made.

To understand how to create a p-Chart in Minitab, consider the following real-world

scenario. A professional printing company discovers a possible problem with a printing
process used to print business cards. Customer are reporting high rates of defects. The
company begins monitoring the process more closely, reviewing 100 samples daily to
determine if there are any defects on each printed card. The results are recorded in the
data table below.

Copy this data into Minitab if you are following along with creating control charts.

437
 ADVANCED CONTROL CHARTS

Samples Defects
100 5
100 2
100 1
100 0
100 0
100 2
100 1
100 0
100 6
100 0
100 7
100 0
100 9
100 0
100 0
100 1
100 2
100 0
100 7
100 0

To create a p-Chart given this data, begin by going to Stat > Control Charts > Attribute
Charts > P

Follow the steps below to enter data in the dialogue box, as pictured below the steps.

438
 SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

1. Click in the box under Variables.

2.
3.
4.
5.
6. Click Select u
7. Click P-Chart Options.

8. Click Tests
9.

439
 ADVANCED CONTROL CHARTS

Note that there are fewer tests for special causes listed than there were for the
previous control charts created. That is because there are fewer control tests for
l times throughout this book that
continuous, or variable, data is typically better for statistical analysis than
attribute data, and this is one example of a reason for that.

10. Click OK.

11. Click OK.

Minitab generates the control chart of defect rates along with the test results from the
control tests.

440
 SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

TEST 1. One point more than 3.00 standard deviations from center line.
Test Failed at points: 11, 13, 19

In this case, there is likely a problem of control within the process, given the failure of three
points. The process owner or Six Sigma team associated with this process would want to
investigate both the samples and the process to find out if there is an ongoing process
problem or a sample-specific issue here.

The steps for creating other attribute control charts, including np, u, and c charts, are the
same as those for creating the p-chart.

With Minitab, creating control charts simply requires following the steps outlined above and
some practice determining which control chart is correct for the situation and setting up
441
 ADVANCED CONTROL CHARTS

data for the control chart. The harder part is interpreting control charts to make valid
ntrol charts for a
variety of real-world scenarios and practice interpreting them.

The technical support desk in an office tracks the number of tickets opened and closed each
hour. That data is found in the table below.

Incoming Complete
Tickets Tickets
10 8
12 10
15 12
7 9
2 5
8 9
9 8
11 12
1 2
16 14
18 17
5 6
8 7
14 13
7 6
6 7
17 12
8 10
5 6

Management wants to know whether the help desk process is performing well given only
this information. A Six Sigma expert can do a few things to provide a preliminary answer to
management.

442
 SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

First, you might create an I-MR chart for the Complete Tasks data. Why use an I-MR chart?
-bar chart, because we would only have a few
subgroups.

The I-MR chart is shown below, and you can see that there are no control failures for either
the moving range or the number of tickets closed per hour.

Someone might point out, though, that a control chart of the number of tickets completed

hour, and this is true. This control chart does not actually take into account the amount of
work that came in each hour and whether the help desk was clearing that work.

So, a Six Sigma expert might add another column to the data. This column keeps a running
total of how many tickets at that point of the day remain open.

443
 ADVANCED CONTROL CHARTS

Incoming Complete Tickets Left

Tickets Tickets Unresolved
10 8 2
12 10 4
15 12 7
7 9 5
2 5 2
8 9 1
9 8 2
11 12 1
1 2 0
16 14 2
18 17 3
5 6 2
8 7 3
14 13 4
7 6 5
6 7 4
17 12 9
8 10 7
5 6 6

Creating an I-MR chart of this figure yields very different results.

444
 SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

TEST 1. One point more than 3.00 standard deviations from center line.
Test Failed at points: 17

TEST 2. 9 points in a row on same side of center line.

Test Failed at points: 13

TEST 5. 2 out of 3 points more than 2 standard deviations from center line
(on one side of
CL).
Test Failed at points: 18

TEST 6. 4 out of 5 points more than 1 standard deviation from center line
(on one side of
CL).
Test Failed at points: 8, 9, 10, 12

445
 ADVANCED CONTROL CHARTS

Numerous tests are failed at numerous points on the chart. Is this a bad thing? Not
necessarily, given an understanding of the process. The first five points that fail do so
because they are below the centerline. That means the process was leaving fewer tickets
open per hour than might be expected. Several reasons could exist for such results:

- The team was just especially efficient at closing tickets those hours.
- The team received tickets that were abnormally easy to deal with and close during
those hours.
- The team closed tickets without really resolving them.
- The team simply received fewer tickets to work on.

The first two reasons are positive, and the third could point to a problem with the team.
The last reason could be positive or negative: Were there fewer tickets because everything
was working better than normal, or was there a problem with the ticketing software?

All of these questions are worth looking into they not only tell the team whether the
process is possibly out of control, but they also provide valuable business information.

The next two points that are out of control on this chart are due to upward trend
something a help desk manager would not want to see. He or she would want to continue
monitoring this chart to see if the process corrects itself or continues to trend upward.

Running a Hypothesis Test

Given all of this information, a Six Sigma expert might even run a hypothesis test to go along
with any control charts. He or she might run a 2-Sample T test to determine whether the
team is closing on average the same number of tickets that are being opened. If the answer
is yes, then the help desk is probably performing well. If the answer is no, then the team

The results of running the 2-Sample T test in Minitab are seen below.

446
 SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

Granted, with a sample size of 19, the chance of a Type II error is relatively high, but the p-
Value is also very strong in favor of accepting the null hypothesis that the mean is
statistically the same between the data for opened tickets and closing tickets. Given this
information, the Six Sigma team might accumulate more data before running another test
to verify these results.

At a company that produces yarn skeins for retail sales, each 244-yard skein features a
certain color or color pattern. The company allows slight defects in dye at a rate of 10 per
244-yard yarn length before the skein is considered defective overall.

A team has been tasked with controlling the quality of this process. This team reviews 10
lengths of sampled yarn each day before it is wound into a skein to determine the number
of defects. Over several days, the team gathers data and creates a control chart to indicate
whether its process is in control.

447
 ADVANCED CONTROL CHARTS

The team generates the following data table.

Skein Defects
Sampled Found
10 8
10 2
10 12
10 15
10 22
10 6
10 8
10 14
10 9
10 23
10 24
10 26
10 10
10 7
10 5
10 12
10 15
10 18
10 19
10 14
10 27
Because all the sample sizes are the same and the team is concerned with the number of
defects, it chooses to create a c-chart in Minitab.

448
 SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

TEST 1. One point more than 3.00 standard deviations from center line.
Test Failed at points: 2, 12, 21

Based on three points that fail the control tests, this process is not completely in control
and should be reviewed.

A banking call center monitors its employees for compliance with numerous federal
regulations as well as internal policies and procedures. The monitoring comes from a quality
assurance team who randomly listens to phone calls and records information about errors
that team members might make when on the phone. Team members can make more than
one error on each phone call; for example, a team member might forget to let the caller
know the call is being recorded and not provide the caller with the right information.
449
 ADVANCED CONTROL CHARTS

In a 10-hour shift, the quality assurance team monitored calls each hour and created the
following data table.

Calls
Hour Monitored Errors
1 15 3
2 13 5
3 18 4
4 10 0
5 12 2
6 9 2
7 11 4
8 12 5
9 8 0
10 7 1

The sample size, which is indicated by the number of calls monitored each hour, is not
consistent in this example. Reasons for inconsistent subgroups in this case could be that the
number of calls received each hour vary and the quality assurance team only reviews a
certain portion of calls. It could also be that some calls take longer and thus take longer to
review. Regardless of the reason for inconsistent subgroup sizes, it leads us to choose the u-
Chart, which is concerned with the number of errors per unit. In this case, that would be
calculated by dividing errors by calls monitored, creating rate or proportion data so that the
actual sa

When plugging data into Minitab to create this control chart, Errors are the Variables and
Calls Monitored are the Subgroup size.

Minitab generates the following control chart.

450
 SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

You can see that no tests are failed, which

note that the upper control limit seems to stair-step all over the place. This is because each
sample was a different size, so each upper control limit is calculated at a different value.

What if call center management was especially concerned with a specific error?

If call center management was concerned with PCI-compliance (compliance with the
regulations for the payment card industry), they might want to dig into the data to see
information about the number of errors related to PCI-compliance. They might ask the
quality assurance team to check a box when reviewing a call to denote whether the call was
compliant with PCI standards or not.

The new data table is seen below.

451
 ADVANCED CONTROL CHARTS

Calls PCI Non-

Hour Monitored Errors compliance
1 15 3 1
2 13 5 0
3 18 4 1
4 10 0 0
5 12 2 1
6 9 2 1
7 11 4 2
8 12 5 2
9 8 0 0
10 7 1 1

Instead of worrying about the number of errors per call, we want to understand the rate of
PCI errors occurring. We can use a p-chart for that purpose.

452
 SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

This chart is a great illustration of how an in-control process can still be a problematic
process. This process is in control, but four of the last five points are above the 0.10
proportion mark. That means that out of 10 samples, four had a fail rate of 10 percent or
more when it comes to PCI compliance.

In the banking world, PCI compliance issues can cost a great deal of money, which means
banking call centers will want to see PCI compliance error rates as close to zero as possible.
The process owner for this call center is likely to take action to reduce the mean proportion
of this process, even though it is in control.

ests
for special causes. This lets Minitab know to perform all the tests for special cause variation
special cause variation is that which is caused by something outside of the normal
expectations within a process.
453
 ADVANCED CONTROL CHARTS

Any process is going to have some amount of variation. The normal variation within a
process the variation that is caused by the process itself is called common cause
variation. For example, a phone operator working at a call center is going to make an
occasional speaking or typing error while performing job duties. That variation is accounted
for on a control chart. However, an operator who is working too many hours, has a bad
headset, or is new on the job might make more than the normal amount of errors. That
abnormality is likely to show up on a control chart and it is explained by the special causes
such as a malfunctioning headset.

Minitab offers a variety of other control charts. We have covered the most commonly used
control charts in this chapter, but highlight a few others briefly below.

Not to be confused with the aforementioned I-MR (Individual Moving Range) chart, the I &
MR (Within/Between) chart provides a three-way chart that indicates variation within a
group and between groups. You might use this chart when the variation between batches is
high compared to variation within batches or when variation is not randomly distributed.
For example, when working with a process that spans shifts, you might have subgroups
within shift groups, and differences might appear between the shifts.

You can find this chart under Stat > Control Charts > Variables for Subgroups

The EWMA chart delivers an individual and moving range graph. This chart is rarely used in
common statistical analysis, as it is difficult to work with. Most often, the EWMA chart is
used when dealing with automatic data systems.

The CUSUM chart deals with the cumulative sums of the deviations of each sample as they
relate to a target value. CUSUM is very adept at detecting a small shift from a target,
particularly in a process that is in control. Like the EWMA, CUSUM is not a commonly used
control chart for most business environments.

454
 yet but
and analysis to revisit some fundamental concepts of Six Sigma.

easy to get wrapped up in the numbers without remembering the business need. In fact,
the more you learn about statistical analysis, the easier it is to get caught up in the
challenge of the analysis and forget that the business is relying on you for answers about its
own challenges.

about the importance of including statistics in business presentations and some tips for
integrating statistics within those presentations.

Here is the first, and perhaps biggest, challenge faced by Six Sigma experts when delivering
information from statistical analysis
Six Sigma expert too or your leadership committee is well-versed in Six Sigma methods, the

Chi-Squared statistic or p-
care about the conclusions or results of your analysis, and they certainly care about the
results of the project as a whole. But you have to find a way to present your information in

and language.

A second challenge is that general business staff and leaders are unlikely to have the same
understanding of statistical analysis as a Six Sigma expert does. The Six Sigma presenter is
455
 APPLYING STATISTICS TO BUSINESS APPLICATIONS THROUGH SIX SIGMA

then tasked with presenting data when the backup for the data might not be understood.
As a Six Sigma expert, you have to be able to explain why the team chose to move in a
certain direction or how you know a process has improved, and not everyone is simply
going to take your word that the p-Value was appropriate for the decision you made.

A third challenge when presenting your analysis is that you can easily become distracted or
derailed by the statistics without getting your business point across. This can happen for
one of two reasons. First, if your sponsor or another member of your audience is
knowledgeable about Six Sigma and statistical analysis, he or she might begin asking
statistical questions instead of business questions. Why did you use the 1-Proportion test?
How did you choose the sample size? Do you really think a beta of 0.2 is sufficient, or should
you have gone with a beta of 0.1?

Some statistical questions might very well be valid. How did you choose the sample size?
That might be something relevant to cover in a presentation so that the audience
understands what assumptions were made by the team and what risks were considered
d made decisions

was set correctly. One way to keep this type of derailment from occurring is to keep a
sponsor with Six Sigma knowledge in the loop throughout the project; that way, he or she
can provide input on such things in real time if desired.

The other reason you might get lost in statistics when you are supposed to be presenting
business-relevant information is because your audience understand the statistics,

between educating the audience on necessary information regarding the Six Sigma process
and providing an in-depth discussion of statistical analysis. For example, you probably

important. You probably need to explain all the types of hypothesis tests or even, in
some cases, that you used a hypothesis test at all.

A final challenge in presenting statistical data is in a tendency to include too much

information or too many examples. You have a box plot, a bar graph, a pie chart, a
hypothesis test, and several calculations based on probability distributions. They all look
good to you. They all say a little something different about the data. They should all go in
the presentation, right? Wrong too many ways of illustrating the data overwhelms the
the
nuances in the data that a trained Six Sigma expert does, which means from their

456
 SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

presentation space or meeting time.

Even with a
presentation altogether. Whether you are presenting findings at a Measure or Analyze
tollgate review or you are presenting for final approval on improvements, statistics are what
provides your conclusions or results credibility. Incorporating your data into a presentation

statistics for credibility, but many people in your business

audience are likely not to follow the intricacies of the actual analysis, what do you do? This
tistical process control come
in. If you are working in a true Six Sigma environment, then you are never working only with
the numbers. Remember from Unit 1 that Six Sigma is more concerned with the business
and its goals and success than it is with the data and the analysis. In fact, the analysis and
data is only relevant in so much that it can be used as a tool to make or measure
improvements to business processes.

Because of this, the Six Sigma team should already be rooted in business terms and phrases.
Consider the act of drafting the problem statement, which we covered in chapter 5. The
problem statement combines statistical concepts with business goals and needs and it is a
good reference point for anyone presenting Six Sigma information throughout and following

ask yourself: does it relate to the problem statement or a solution for the problem
cate or present in a general
business environment. As you work with data and statistics, you might be surprised by how
your interest in the numbers grow. While interest in the numbers can lead you to valuable
discoveries and conclusions, you do need to avoid the temptation to include data or

Other ways you can create viable business presentations when dealing with statistics and
analysis include tailoring presentation to the audience, creating a story-style narrative,
keeping things short and simple, integrating appropriate pictures, and avoiding misuse of
technology, presentation, or statistical tools.

457
 APPLYING STATISTICS TO BUSINESS APPLICATIONS THROUGH SIX SIGMA

When presenting or communicating about a Six Sigma project or statistical analysis, tailor

consider a Six Sigma Black Belt who is presenting during an Improve tollgate review
meeting. The Black Belt is presenting to a leadership team, which includes the project
sponsor. At least two members of the leadership team are also Black Belts perhaps one is
even a Master Black belt.

Given this scenario, the audience is both invested in the information and capable of
understanding many of the statistical concepts behind the decisions made by the Six Sigma

data in the presentation your audience is also busy. But it does mean you should
anticipate at least a few questions about methodology or statistical analysis, and you can
incorporate a few technical terms or results into your presentation without worrying about
having to explain them.

When dealing with this type of audience,

appendix. You can put your analysis, data tables, and even specific Minitab results in the
appendix, where they can be quickly displayed if a specific question calls for it. In situations
that involve high costs or high risks, leadership teams are more likely to want to see the
background analysis because they need to fully understand all assumptions and risks before
making a decision on the recommended improvements.

In other situations, you might be called upon to present Six Sigma information or statistical

statistics. Examples include presenting to general management or presenting the reasons

behind changes to the staff in a department impacted by a Six Sigma process improvement.
When this happens, it can be helpful to rely on graphical representations of your analysis,
particularly if the information you are trying to share is very apparent on a certain type of
graph.

For example, a Six Sigma team in a government office has been task with improving the
production of a workflow process that involves reviewing applications for financial aid. The
team has made improvements to the process and now compares the average hourly
production rate of the process before the changes with the current average hourly
production rate. To do so, they complete a 2-Sample T Test to determine whether the new
data has a higher average mean than the old data. The Minitab results are presented below.

458
 SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

Knowing what you do about hypothesis tests and p-values, you can see quickly that we
reject the null hypothesis and accept the alternative. Yes, the mean averages of the process
have changed. It is, according to our statistical analysis, a more productive process.

But can you just copy that Minitab data into a PowerPoint slide? Probably not only a Six
Sigma or statistical analysis expert is going to be able to make that connection between p-
value and real-world conclusion.

Consider the graphical image below.

459
 APPLYING STATISTICS TO BUSINESS APPLICATIONS THROUGH SIX SIGMA

This is a box plot of the same data used to create the 2-Sample T Test above, but it provides
a friendlier version of the data for most business employees. You can quickly see that the
process C2 has a higher rate of production than the process C1
anything about box and whisper plots.

statistical analysis. When creating a presentation about your project or statistical data, you
also need to understand what the target audience wants to know and how they need to
know it.

High-level leadership typically wants to know whether the project was a success,
how you know it was a success, what savings or gains can be expected, and how you

Process owners and employees close to the process should know all of the above,
though you might have to broaden discussions about costs and savings to leave out
sensitive data that could be included in an executive-level presentation (such as
assumptions about employee pay). Process owners and employees are also likely to
want to know more details about the specific changes to the process why were
such changes implemented and how does it impact daily workflow? Where
leadership might be satisfied with a high-level process map, for example, process
owners want to see more granular maps.
Audience members from disparate departments are likely to ask questions that
make sense from their perspective, so know who will be in a tollgate or other type
of presentation meeting, and be prepared to answer questions specific to their
world.
o Compliance staff asks about adherence to regulations. Did the team
consider regulations when making decisions? What happens when
compliance required a decision that seemed opposite of where the
statistical inferences led the team? Did the team calculate risks and costs
associated with compliance failures, which are often greater than other
process failures?
o Accounting or finance staff asks about budget, revenue, and savings. Does
the team have a graph that shows results in a dollar format? What is the
likelihood of saving over a certain amount? Can the team associate a dollar
amount with various analysis, changes, or improvements?
o Marketing, sales, and customer service staff are likely to ask about
customer-facing and customer-centric information. How did the team
gather the voice of the customer? What assumptions about the customer
460
 SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

did the team make? When making process changes or analyzing data, how
did the team consider the customer?

process, project, or statistical analysis. The presentation would be enormous and take much

with plenty of time for questions, and your presentation time might be even more
constrained depending on organizational business practices. So, you should always answer
the biggest and most important questions about your improvements, project, or DMAIC
stage in the presentation and be prepared for other types of questions in the question and
answer session (this is where an appendix can be helpful).

Use PowerPoint or another presentation program to combine your information in a way

that tells a story that the audience can understand and care about. Keeping your audience
and the purpose of your presentation in mind, attempt to tell an overall story across your

Consider the example slide below, which presents the 2-Sample T Test information we used
previously in this chapter. The slide tells a story with the placement of information, the
inclusion of a graphic, and the question-and answer format. This is an easy template that
can be used in a majority of Six Sigma presentation slides, particularly when presenting
statistical analysis or data.

- First, ask a question as the title of the slide. Create a concise question, but try to write a
stand-
what process, and what hourly rates are being discussed? In the context of an entire

for an email communication or training document and executive leaders tend to pull
out one slide or graphical image to communicate information about a sponsored

- Next, include an image that answers your question or provides additional information
regarding the data or process. Pictures, graphs, and maps break up long blocks of text
in a presentation, making it easier for the audience to digest. By presenting data in a
variety of formats on one slide, you also capture the interest and understanding of a
wider audience. Graphical representations of data often get the point across quicker
and can help you tell a complex data story in a small amount of time.
461
 APPLYING STATISTICS TO BUSINESS APPLICATIONS THROUGH SIX SIGMA

- It can be a good idea to include some basic statistical information, such as a business-
friendly version of the null and alternative hypothesis. At the very least, everyone in the
audience can understand that these two statements are what the team were
evaluating, and only one of them can be true. Including the type of test run and the p-

informing sponsors or others who are also knowledgeable in Six Sigma.

- Finally, answer the question posed in the title in a short, text-based conclusion. The
conclusion should align with any graphical representation or statistical test data
included on the slide. You should avoid presenting contradicting data and conclusions
on a single slide unless you have a specific reason for doing so and an explanation of
why the contradiction is important.

Not all questions require all elements of the above slide. Sometimes, it makes sense to

that slide would be the problem statement as written

in the Define phase.

Sometimes, the only thing you need to include on the slide is a graph or chart. Consider the
slide below.

462
 SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

This slide asks what is causing defective orders to be sent out, and it answers the question

given the labels and data that the issue creating the biggest number of defects is in product
finishing. Someone who knows about Pareto charts could see within seconds that finishing

Some Six Sigma tools, such as the Pareto chart, are easy to explain to audience members
-existing understanding of data analysis or Six Sigma. Given a minute
or two of presentation time, you could sufficiently explain how the above chart answers the
eto (and other bar-type charts) are popular in
such presentations.

No matter how you are presenting your data, make sure each slide and each element on
the slide is clear and concise. This means writing in active language as much as possible,
avoiding long text explanation, and ensuring everything is spelled correctly and that you

easy to make an incorrect word choice that changes the meaning of your entire statement.

463
 APPLYING STATISTICS TO BUSINESS APPLICATIONS THROUGH SIX SIGMA

audience.

sentence structure. You should also be

on a single slide that words or images become unintelligible to the audience. In some cases,
you might need to ask a question and answer it over a series of slides.

Be cognizant of the fact that a projected slide image is never as clear as the image on your
computer. Because of this, you should always avoid layering light font or images over a light
background or layer dark font or
something as small as 12-point just to fit all the information -- remember, your audience

information on a slide the slide should act as the visual reference point for the audience
while you add information verbally.

need to make sure text and images within graphs are as clear and concise as possible.
Consider the two boxplots below.

Boxplot of C1, C2
14

12

10

0
C1 C2

464
 SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

Average Production Per Hour for Application Workflow

14

12

10

0
OldProcess NewProcess

who understands boxplots would simply be able to tell that two things were being charted
and that the average of one was higher than the average of the other.

The second boxplot is the same as the first, but the title and data labels have been edited.
Now, you can see that the graph provides information about the application workflow.
Specifically, the graph is illustrating the average production rates per hour for that
workflow, and the new workflow process seems to be more productive than the old one.

Creating graphical analyses that tell a small story all on their own with the help of data
labels and titles is important. First, as the Six Sigma presenter, labels provide cues that
help you verbally explain the slide or graph. After several slides with similar-looking graphs
or charts, even you can get confused about what specific data you are talking about.
Second, appropriate labels on such images let you reuse the image without the context of
the presentation or slide. This means you waste less time and can reuse your work when
communicating about your project or data effectively reducing muda of rework. Finally,

misinterpreting it or getting the wrong idea from it. If graphs are properly labeled, then it is
harder for someone seeing the information outside of your presentation to draw the wrong
conclusions.
465
 APPLYING STATISTICS TO BUSINESS APPLICATIONS THROUGH SIX SIGMA

You can also edit colors of data on charts in both Minitab and Excel, making it even easier
for an audience to draw appropriate conclusions from the illustration. Consider the Pareto
chart below. This is the same Pareto chart shown earlier in this chapter, but in this version,
the top three categories are presented in red. The color change emphasizes those
categories, which account for 80 percent of the errors, making it easier for the audience to
draw the conclusion that these are the issues that the team will concentrate on.

A final tip for creating business-

and resources you have for creating a presentations or communicating ideas.

to avoid animating every aspect of your presentation. Yes, there is a time for fade-in or fly-
in, but the time is not every single bullet point or image. Use animations sparingly if at all
for impact about very important conclusions or elements of your presentation. For the
most part, a strong Six Sigma presentation should rely on clarity and data over gimmicks
and animations.

466
 SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

that, Six Sigma analysts and teams tend to accumulate a lot of versions of data. Treat
Minitab graphs and analysis as you do PowerPoint animations. Use them as needed, but

per slide or data-related question. This forces you to choose the graph or chart that you feel
most powerfully or appropriately illustrates your data. There are exceptions to this rule, and
you might need to include two charts if one chart helps explain another or you are
comparing information about data or multiple data sets.

you make a data analysis error inadvertently or you force an analysis to fit a desired
outcome or conclusion. The first step is more of an error than a misuse; you can avoid
presenting such errors to others by double checking all conclusions and inferences and
having someone else knowledgeable about Six Sigma review your analysis before you
finalize a presentation. This is why many Six Sigma organizations have at least one Master
Black Belt on staff Black Belts and Green Belts can call on the Master Black Belt if they
have analysis questions or simply want to confirm they are on the right track with analysis.

No one is infallible, and mistakes can happen. As a Six Sigma expert, however, you can

you certainly should never cloud analysis by presenting it in a way that makes it hard for
your audience to see the appropriate conclusion because you want them to draw another
conclusion. For example, the trick used above to highlight a certain part of the Pareto chart
with color can be used inappropriately. You can draw attention to a certain part of a graph
or bar chart with color, increasing the chance that the audience will conclude that the
colored portion is of more relevance or more important than other data on the graph.
While this is sometimes true, Six Sigma teams and presenters must be careful when using
such tactics only create this scenario when the data supports such an emphasis.

Another example of creating bias in a presentation comes when you present analysis
information knowing the sample size was not big enough for strong conclusions, but you

others making decisions based on your data without understanding the risks that the
conclusions might be wrong.

Similarly, if you run four types of test in Minitab and three support one conclusion while
only a single test supports the conclusion the team originally wanted to present, you

review each of the tests to understand why there is a difference in one; you should also

467
 APPLYING STATISTICS TO BUSINESS APPLICATIONS THROUGH SIX SIGMA

ensure that the tests you are using are appropriate for the data, question, and sample size
at hand. If you verify the statistics are all correct, then you usually go with the conclusion
supported by the greater number of tests.

Given the decision-making processes and requirements in a corporate business world,

presentations are a requirement of the DMAIC process. Six Sigma teams and experts are
going to have to present their findings at some point, so presentation skills are something
you do need to work on.

That being said, it is important that you never let the need for a good presentation drive any
part of the process. Six Sigma teams should never make decisions about data gathering,
graphical analysis, process mapping, brainstorming, or any other Six Sigma function because
they think it will be good or look good for the presentation. Instead, the presentation of
information to the business must come after the work is done.

This truth is relevant any time you are presenting information to the business or business
staff not just when you are standing in front of a conference room with a PowerPoint
behind you. For example, Black Belts often communicate with sponsors or business leaders
regularly during the course of a DMAIC project. Perhaps the team provides a daily email or a
weekend update to the sponsor; whatever the mode of communication is, all of the tips in
this chapter are relevant.

468
 along in previous chapters through Minitab for sample size

within Minitab. This entire unit provides a how-to for some of the more common Six Sigma
statistical analysis tools used in Minitab. We are including Minitab because it is one of the
most widely used programs for this type of application, because it is, overall, fairly user
friendly, and because it is incredibly powerful for statistical analysis purposes.

In previous chapters, we walked through Excel calculations and analysis for some very basic
problems. The steps involved sometimes became very complex. Consider the Chi-Squared
Goodness-of-Fit Test performed in Excel in the chapter on normal probability distributions.
The calculations do work, and you can always use them absent a statistical software
program. However, by using Minitab, within a half-dozen mouse clicks you can arrive at the
same answer about the normality of data that all those Excel calculations required. Not only
do you save time, but you reduce the opportunities for a numerical or mathematical error.

If Minitab or other software is an option, why bother learning how to perform some of
these statistical analyses by hand or in Excel? As stated in the above paragraph, you might

important to understand the fundamentals behind some of these calculations so you

understand why Minitab provides
difficult to create the presentations and explanations discussed in chapter 24. Instead, you
would simply be relying on Minitab to make decisions for you, which is never a good way to
manage a Six Sigma project.

469
 INTRODUCTION TO MINITAB

Remember -numerical priorities. Statistics

can say something is a better option than another, but you knowing the process, the
goals, the resources, and the willingness of leadership to take risks or spend money might
decide against that option for good reason.

Before getting into some specific Minitab capabilities, take a moment to understand the
Minitab interface. At first glance, you might think this is just a spreadsheet, but the
interface actually has three prominent windows.

window on bottom. The worksheet window looks and behaves a lot like a worksheet in
Excel. You can have more than one worksheet window open in Minitab at a time, and you
can save individual worksheets as part of a Minitab project.

-based icons for

minimizing, maximizing, and closing the worksheet. You can also use your mouse to grab
the edges of the worksheet on all four sides to make it larger or smaller in the window as
you wish.

470
 SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

The worksheet is where you enter data that you are attempting to analyze. You can enter
data by typing it in yourself, but you can also copy data into Minitab from numerous other
spreadsheet programs, including Excel.

If you enter data into a column in Minitab, you can right click on that column to make
format changes to the column or the data within the column.

471
 INTRODUCTION TO MINITAB

Clear, delete, copy, cut, find and replace, sort, insert, and conditional formatting all work
very similar to the way these functions work in Excel. You might use these options when you
are ordering and preparing data for analysis or if you have analyzed data and you want to
create a more visually-pleasing format for a presentation. Minitab is more limited than Excel
when it comes to formatting and presenting information, which is why Six Sigma experts
often work simultaneously in both programs.

this option brings up the Format Column dialogue box.

The dialogue box lets you format numbers within a column to match certain parameters.
You can set Minitab to only display 2 decimal places, or set a column for percentage,
currency, or time. This helps create uniform data and reduce the confusion of the display,
especially since Minitab tends to default to a 3 or 4 decimal readout, which is not
appropriate for every analysis. If you are analyzing the number of workers in a shift, for

472
 SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

Like Excel, Minitab accepts both numerical and alpha-numerical entries. If any cell in a
column features alpha-numerical text, Minitab notes that the cells in that column are text

including text in Minitab would be to clarify subgroups or divisions of data and for use in
automatically labeling some charts.

When you first open Minitab, the session window is located above the spreadsheet-style
worksheet. The session window is critical to Minitab functionality because this is where all
of the analysis and calculations are displayed. Below is an example of results in the session
window. You can see that Minitab typically provides a bold header identifying what
calculation or analysis was run, followed by a summary of the test or analysis and the
results.

Here, for example, you see the results of a 1-Sample Z test. The null hypothesis is that the
mean is equal to 5. The alternative is that the mean is not equal to 5. A standard deviation
of 1 was assumed for this test.

Minitab provides several pieces of analysis, the most important in this case being the p-
Value of 0.893. In this case, the p-Value is above 0.05, so we would fail to reject the null
hypothesis.

473
 INTRODUCTION TO MINITAB

Throughout the next three c

learning to run and interpret various tests in Minitab.

Before moving onto the various menu options in Minitab, consider one more window in the
interface. The window might be minimized at the bottom of the screen. Click to maximize
this window for a moment.

474
SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

475
 INTRODUCTION TO MINITAB

This is the project window, where you can manage all of the data, worksheets, charts, and
graphs you create with Minitab. Minitab functions similar to Microsoft OneNote in this
regard. You can create a project file. Within that file, you can store additional files, including
workbooks, reports, session windows, and graphs. You can also view the history of your
open session; after a few hours of analysis, it can be difficult to remember what various
results are. In this window, you can find tests you already ran and click back to view them.

Minitab has a fairly expansive menu bar, and we will not cover all of the options and icons
within this unit. The file, edit, window, and help menu bar options are fairly self-explanatory
and function in keeping with similar menu options in common software programs. If you are
working in your own version of Minitab, you can experiment with those menu options and
the variety of icons below the menu on your own.

Calc, Stat, and Graph menus.

476
 SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

capabilities found under the Calc menu

option:

- Generate random data

- Calculate column or row statistics
- Calculate probabilities given various distributions

The random data tool lets you generate random data that conforms to a certain probability
distribution. This tool is helpful if you are testing different capabilities in Minitab or
practicing Six Sigma statistical analysis without any of your own data to start with. It can
also be helpful if you want to compare your own data to various types of distributions.

use for other functions within this chapter.

Click Calc > Random Data and then select a distribution you want to work with. For the
purpose of this exercise, select the normal distribution.

477
 INTRODUCTION TO MINITAB

In the Normal Distribution dialogue box, type:

- 30 for the number of rows of data to generate. This will generate a sample size of
30. You can create a sample size of a few data points or hundreds of data points.
- C1 for the column you want to store data in. You can type the alpha-numeric name
of any column or, if you already named some columns, you can select them from
the list on the left.
- 5 for the mean. You can specific any mean when you generate random data on your
own.
478
 SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

- 1.0 for the standard deviation. You can also specific any standard deviation when
you generate random data on your own.

Once your dialogue box looks like the one pictured above, click OK.

Minitab populates column C1 with data that fits the parameters you set. You can try
creating other types of data using this function, with various other parameters such as
different means and standard deviations, to see how inputting different information
changes the data that is returned.

Sometimes, you just need to know a specific statistic about your data quickly. The calculate
Column or Row Statistics functions lets you calculate the sum, mean, standard deviation,
min, max, range, median, sum of squares, or N total given data in a single column or row.
You can only calculate one of these statistics at a time, though, which creates some
limitations. In
descriptive statistics at one time.
479
 INTRODUCTION TO MINITAB

Using the random set of data created in the previous section, try out the column statistics
function.

Click Calc > Column Statistics.

Click to select the mean.

480
 SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

Click in the Input Variable box.

Without clicking in any other box first, click on the C1 listing in the large box to the left.

481
 INTRODUCTION TO MINITAB

move it into the Input Variable box.

in a box on the left. You have to select data by double clicking or by clicking the column
name and clicking select.

Click OK.

Minitab displays the mean of the data in the session window.

Note, the mean of your data will not necessarily be the same as the value calculated above,
since you generated random data that is likely somewhat different from the data that was
used to calculate the example in this book.

482
 SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

Calculating statistics for row-based data is done in the exact same fashion, except that you
would select a row name instead of a column name.

Take a moment to recall the chapters on both normal probability distributions and non-
normal probability distributions. In both of those chapters, we demonstrated calculations
that let us determine the probability of certain outcomes given a set of data and a certain
probability distribution.

If you recall, we mentioned that some of these calculations can get fairly complex, and we
only handled a few through manual calculations in those chapters. This is because software
such as Minitab lets you calculate the PDF and CDF by entering a few pieces of information

Brief Review
PDF = Probability Density Function the probability that a random variable, X, will
occur in the data. In a given population of people, what is the probability that a
randomly selected person will weigh 125 pounds?

CDF: Cumulative Density Function the probability that a randomly selected data point
will be less than or equal to a variable, x. In a given population of people, what is the
probability that a randomly selected person will weigh less than or equal to 125 pounds?

about your data.

To understand how to calculate PDF and CDF -world example.

A Six Sigma team working to increase the process efficiency of a library checkout system
randomly measured the length of time it takes for librarians to check out patrons. For this
purpose, assume the data is normal. The mean time of the sample was 5.5 minutes. The
standard deviation was 3.028 minutes. What is the probability that it takes exactly 5
minutes to check out a patron?

Begin the calculation by selection Calc > Probability Distributions > Normal

483
 INTRODUCTION TO MINITAB

an exact value.

,
want to find the answer for a specific number, not for a list of numbers in a column in your
worksheet.

484
 SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

Once your dialog box looks like the one pictured above, click OK.

being exactly 5 minutes.

What if you wanted to find out what the probability was that the time would be 5 minutes
or less?

Repeat the steps above, but choose cumulative probability instead.

485
 INTRODUCTION TO MINITAB

There is a 0.434 probability the time will be 5 minutes or less.

What is the probability that the time will be greater than 5 minutes? The total probability is
1, so:

1 0.434 = 0.566.

There is approximately 0.566 probability of the time being greater than 5 minutes.

486
 SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

20. In that scenario, a Six Sigma team was working with a process that produced boxes with
20 chocolates each. For each chocolate, there was a 0.05 probability of the chocolate being
defective. What is the likelihood of getting a box of chocolates where 2 of those chocolates
are defective?

This time, select Calc > Probability Distributions > Binomial.

Unlike the normal curve, the binomial distribution deals with number of trials and event
probability. If you recall from chapter 20, when discussing binomial probabilities, we talk
about the likelihood of one event or the other occurring for each trial.

For our chocolate boxes, there are effectively 20 trials in each box each chocolate can be
either defective or not defective. For each chocolate, there is a 0.05 probability that a
defect will occur. This is the event probability. We can plug these two numbers into the
Binomial Distribution dialog box and select the i
probability of getting a box with exactly 2 defective chocolates.

487
 INTRODUCTION TO MINITAB

Click OK.

chocolates.

What if we want to know the chance of getting over 5 defects? Repeat the steps above, but

Click OK.

488
 SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

Minitab returns an answer of 0.999, but remember that this is the cumulative probability. It
is the probability of having 5 or less defects in the box. The probability of having more than
5 is:

1 0.999, or 0.001.

One of the reasons for using Minitab is that it lets you analyze data with utmost efficiency.
While you can see that calculating a single probability or CDF is much faster than handling
the math on your own, what if you wanted information about a set of numbers at one time?
What if the chocolate team wanted to know the probability of getting 1, 2, 3, 4, 5, 6, 7, 8,
and 9 defects?

First, enter all of these numbers in a column in Minitab.

489
 INTRODUCTION TO MINITAB

re

numbers 1 through 9.

Click OK.

Minitab generates the probability for all the numbers you entered in the session window.

Probability Density Function

490
 SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

Before calculating probabilities, make sure you have the right distribution. If you have data
that follows a Binomial Distribution and you use Minitab to calculate answers based on the
normal curve, your answers will be incorrect.

Use the Binomial Distribution when you are dealing with binomial data and a consistent
event probability for every trial. Use the Normal Distribution when you are dealing with
data that is normal or approximates the normal curve. Remember, the Central Limit
Theorem says that, when dealing with continuous data, the larger the sample size, the
closer it will approximately the normal curve.

You can run a quick graphical analysis to determine if data you are working with is normal.

To understand how to perform this analysis, copy the following data into Minitab.

6.319075
3.556859
3.452996
3.300786
4.340399
5.242524
2.542238
4.429293
7.757632
7.582288
4.692596
4.450197
6.454293
3.666121
5.243754
2.025502
4.189321
2.79723
3.397714
5.507707

491
 INTRODUCTION TO MINITAB

Select Stat > Basic Statistics > Graphical Summary.

For Variables, select the column where you copied the data from above. Leave the
confidence level at 95.

Click OK.

492
 SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

Minitab generates a graphical and statistical analysis of the data.

There are several things on this analysis that tell us the data is probably normal:

- The graphical analysis does somewhat appear to follow a normal curve. Because we

on the image.
- The mean and median are fairly close, and the confidence interval of the mean is
within the confidence interval of the median. This is another indication we could be
dealing with a normal curve.
- Finally, the p-value for the Anderson-Darling Normality test is listed as 0.486. Since
the p-value is higher than the alpha value (remember, we left the confidence level
at 95, which means alpha is 0.05), we fail to reject the null hypothesis. The null
hypothesis in a normality test is always that the data is normal.

493
Summary Report for C2

A-Squared 7.97
P Value <0.005
Mean 1.0115
StDev 1.0871
Variance 1.1817
Skewness 2.6578
Kurtosis 10.1865
N 130
Minimum 0.0027
1st Quantité 0.3209
Median 0.6765
3rd Quantité 1.3484
Maximum 7.3700

0.8229 1.2001

0.5964 0.8487

0.9691 1.2381

494
One reason many Six Sigma experts use Minitab is the powerful and efficient graphical
analysis interface it offers. As previously mentioned, with Minitab or similar software
programs, you can quickly create graphs of various types to analyze your data. In the last
chapter, you learned how to create a graphical analysis that included a histogram, a
normality test, and several other pieces of information all within a few mouse clicks. The
same option is available in Minitab for scatterplots, histograms, boxplots, and dozens of
other graphs.

can be

advanced program such as Minitab to

charting functions and the Analysis Toolpak to conduct a lot of analysis and draw
appropriate statistical conclusions. That said, Excel is usually too limited for most Black Belts
especially those who are leading projects.

software built specifically for statistical analysis purposes. Can you do all of the calculations
in Excel that Minitab does for you? Yes. But you can also do all of those calculations with a
piece of paper, a basic calculator, and access to printed statistical tables. Either of those
methods involves spending more time to come to the same conclusion, which, as you
learned in the chapters on Lean, is muda or waste.

do a lot of work to create some charts in Excel. If you are looking at multiple charts to
reduces the time spent creating charts and
frees up time for analysis and consideration of data. This is helpful because, as mentioned in
chapter 17, each chart provides a different perspective on data. You might look at four or
five graphical representations of data along with statistical analysis before reaching a
conclusion you are confident in.

495
 GRAPHS AND QUALITY TOOLS IN MINITAB

Excel does offer some advantages when it comes to certain charts specifically with regard
to presenting data. It can be easier to customize formatting of bar and pie charts in Excel,
for example, which means some Six Sigma experts analyze data in Minitab but use Excel to
create some of the charts and graphs included in presentations. If you are working with
they need editable access to analysis
and graph files, then you might also have to provide information in Excel for this reason.

This chapter will cover some of the commonly used graphs in Minitab.

The graph menu is found in the top menu bar in Minitab. Within the menu, you have access
to over 20 graphical analysis options with more detailed options for each type of graph.
ome of

496
 SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

If you would like to explore graphing options alongside the steps presented in this chapter,
begin by copying the information below into a new Minitab worksheet. Ensure both of the
columns are marked by Minitab as cont
in the column header).

1 4.6
2 3.8
3 4.8
4 2.7
5 3.5
1 4.3
2 3.6
3 5.8
4 4.7
5 3.0
1 4.3
2 5.6
3 6.2
4 4.7
5 2.2
1 3.3
2 2.8
3 1.6
4 1.4
5 2.8

Scatter plots graph each data point on X-Y axis points and can help teams see how two
inputs or data elements might be related. Scatter plots are often the first step in regression
analysis, because you can tell quickly whether correlation is at all possible.

To create a scatter plot using the data from above, select Graph > Scatterplots > Simple and
click okay.

497
 GRAPHS AND QUALITY TOOLS IN MINITAB

Click in the first row under Y-variables and select column C2. Click in the first row under X-
variables and select column C1.

498
 SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

Click OK, and Minitab generates a scatterplot of the data. You can see that for x = 1, y values
in our data range between 3 an
relationship between the x and y values.

Now, copy the data below into columns C4 and C5 on the same worksheet.

1 1.5
2 2.2
3 3.3
4 4.2
5 5
1 0.9
2 2.3
3 3.7
4 3.9
5 4.5

499
 GRAPHS AND QUALITY TOOLS IN MINITAB

1 1.2
2 2.1
3 3.1
4 4.2
5 5

Select Graph > Scatterplots > Simple.

Change the x and y references so that x is C4 and y is C5. Click OK, and Minitab generates
the graph below. This graph is different from the one above this data does appear to be
correlated in some way.

in the Scatterplots diagram.

500
 SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

able

but it does seem that the two are correlated. Upon seeing such a result on a scatterplot, a
Six Sigma expert might run a regression test to determine how likely a correlation is
between the two factors. Running such analysis in Minitab will be covered in the next
chapter.

501
 GRAPHS AND QUALITY TOOLS IN MINITAB

If a correlation is noted during a statistical analysis, the statistical proof alone (the math

chapter 24, you might want to include illustrations of the statistical analysis, which is
another reason for using the scatterplot.

several times. Histograms help

Six Sigma experts understand the shape of the data, which also helps understand how data

502
 SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

is distributed. You can create histograms in Minitab by selecting Graph > Histogram.

along, select column C2 from your worksheet into the graph variables box and click OK.

503
 GRAPHS AND QUALITY TOOLS IN MINITAB

Minitab creates a histogram of the data in column C2. The column in question only has 20
data points, which can make it difficult to tell whether the data is distributed normally via a
graphical analysis. Remember the lessons on sample size and the Central Limit Theorem.

as you increase the size of the data pool, most data does approach the normal curve.

504
 SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

Sometimes, placing a fit curve over your data can help you get a better idea of its shape.

dialogue box.

505
 GRAPHS AND QUALITY TOOLS IN MINITAB

Minitab fits a normal curve over the histogram to help you visualize the distribution of your
data. In this case, we might be tempted to say the data is somewhat normal. Only the one

enough information here to draw a conclusion.

506
SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

507
 GRAPHS AND QUALITY TOOLS IN MINITAB

Create a histogram with fit curve of the data in column C1.

graphical summary for the data in column C2 so we can double check our
conclusions about normality. Remember, this summary is found under Stat > Basic Statistics
> Graphical Summary.

508
 SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

The p-value for the Anderson-Darling Normality test lets us fail to reject the null hypothesis
that the data is normal. The other information on this summary all tends to lean toward a
conclusion of normality too.

As you can see, the graphical summary provides the same information as the histogram
graph and the extra information that lets Six Sigma experts make a better determination
about data distribution. For that reason, you are more likely to run the graphical analysis for
general understanding of your data. The histogram function is still valuable for analyzing
certain pieces of information or for generating a single graph that you might want to include
in a presentation or report.

509
 GRAPHS AND QUALITY TOOLS IN MINITAB

A dotplot is similar to a histogram in helping you understand how data is shaped, but it plots
individual data points along a single line.

Access the dotplot function via Graph > Dotplot.

In the example of a dotplot graph below, you can see that four data points are measured as
9; three data points are measured as 5, 6, 7, and 8. Each measurement above 10 has only
one data point related to it. The simple dotplot shown below is usually used for smaller
sample sizes, and teams can even create rough dotplot sketches on white boards or paper
during a meeting if they want a quick visual depiction of data for the discussion.

510
 SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

The dotplots created above are simple dotplots. They depict one Y (one measurement, one
data set) over a single line. You can also depict single Y measurements with groups,
separated or stacked. The concept is similar to a stacked bar graph, where you delineate the
measurements via another variable or factor. For example, you might be creating a dotplot
of amount of work done per hour. If you have all of the numbers per person recorded for a
shift, you could group the information by person.

You can also work with more than one measurement, or Y, with simple, stacked, or grouped
data. For example, you might have data for two shifts with employees in each shift.

The dotplot below shows the amount of work done each hour for two employees in each
shift. You can immediately see that the employees in shift 2 seem to be doing more per
hour than the employees in shift 1 there are more data points located above the higher
numbers in the graph for the second shift.

511
 GRAPHS AND QUALITY TOOLS IN MINITAB

To select the type of dotplot you want to create, click on the appropriate image in the
Dotplots box.

Select the appropriate columns for your dotplot chart data. Graph variables are always your
data columns. Categorical variables for grouping are the columns with your categories
listed, such as the names of people, the labels for shifts, or the time. If you are working with
multiple Ys, you can select more than one data column into the graph variables box.

512
 SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

In chapter 17, you learned to create boxplots in Excel. Minitab creates box and whisker
diagrams without all the editing you have to do in Excel. As with dotplots, you can also
create boxplots with single sets of data, groups of data, and subgroups of data.

Group A Group B Group C

5 5 6
5 14 19
4 3 8
6 6 6
4 6 5
6 8 5
18 7 8
6 6 7

513
 GRAPHS AND QUALITY TOOLS IN MINITAB

5 12 8
4 5 6
6 6 6
6 5 10
5 6 7
4 4 9
5 6 6

Copy the above data into Minitab to create some boxplots.

Select Graph > Boxplot > One Y Simple.

Click OK.

514
 SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

Minitab returns the boxplot of the data above. We can see that the data has a mean of
about 5 and that there is an outlier of around 18 in the data. The outlier is indicated by the
dot toward the top of the graph.

If you wanted to compare the three groups of data, you could graph them all at once.

Select Graph > Boxplot > Multiple Ys Simple.

Select all three data columns into the Graph Variables box.

Click OK.

515
 GRAPHS AND QUALITY TOOLS IN MINITAB

The above graph provides more information than a single boxplot for group A. You can see
there are outliers in each group of data. You can also see that the mean of each group is
progressively higher the mean is denoted by the line in the center of each box. The boxes
are roughly the same size, indicating that the standard deviation, or variance, within each
data set is approximately the same.

What if your data was arranged differently than the table above, which has three groups of
separate data? What if you have two columns of data, and the first column is the group
indication and the second column is the measurement? The data for two groups might look
something like the table below.

516
 SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

Group Result
1 5
1 5
1 4
1 6
1 4
1 6
1 18
1 6
1 5
1 4
1 6
1 6
1 5
1 4
1 5
2 5
2 14
2 3
2 6
2 6
2 8
2 7
2 6
2 12
2 5
2 6
2 5
2 6
2 4
2 6

To create a boxplot for this type of data, select Graph > Boxplot > One Y with Groups.

517
 GRAPHS AND QUALITY TOOLS IN MINITAB

Select the Result column into the Graph Variables box. Select the Group column into the
Categorical Variables box. Note that you can only graph up to 4 categorical variables at
once.

Click OK.

Minitab returns the boxplot below. It is similar to the boxplot returned for the three
subgroups, although obviously there are only two groups here because there were only two
different categories listed in our data table.

518
 SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

Interval plots are helpful when you want to analyze the means or variance within a single
group of data or between subgroups of data. The creation of interval plots in Minitab
follows essentially the same steps as the creation of boxplots, and you can break interval
plots into the same groups.

Using the first data table from the boxplot section above, in Minitab select Graph > Interval
Plot > Multiple Ys Simple. Select all three of the data columns into the Graph Variables box.
Click OK.

519
 GRAPHS AND QUALITY TOOLS IN MINITAB

Minitab returns the interval plots below.

As with the similar boxplot, you can see that the mean for each group is slightly above the
one before it. You can also see that the spread of the data is roughly the same for each

You can create interval plots for all of the data used in the boxplot examples just as you
created the box plots. The only difference is that you select interval plots from the graph
menu.

Excel is a great tool for creating colorful, presentation-friendly bar and pie charts, but you
can also quickly create such charts in Minitab for analysis purposes. Use the phone call per
hour data below to create a bar chart in Minitab.

520
 SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

Phone
Hour calls
8:00
AM 78
9:00
AM 89
10:00
AM 107
11:00
AM 118
12:00
PM 149
1:00
PM 147
2:00
PM 105
3:00
PM 90
4:00
PM 97
5:00
PM 85
6:00
PM 178
7:00
PM 198
8:00
PM 145
9:00
PM 57

Select Graph > Bar Chart.

Select Simple under one column of values.

Click OK.

521
 GRAPHS AND QUALITY TOOLS IN MINITAB

You always select the column that has the data to be graphed here. Select the column
you want
displayed along the x-axis here.

522
 SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

Minitab returns a simple bar chart, as seen below.

The options for creating bar charts in Minitab are similar to those in Excel: you can create
stacked or grouped bar charts. But you can also create charts from data without having to
perform the same calculations that you might have to do in Excel. For example, if you select

count unique values for you. You can try this with the data table below, which lists the
number of errors in 13 work samples.

box. Select Simple and click OK.

Minitab counts the number of times each unique value 1, 2, 3, or 4 appears in the data
and creates a bar chart of those counts.

523
GRAPHS AND QUALITY TOOLS IN MINITAB

524
 SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

Minitab also offers some options for ordering data in your chart. Create the first chart of
calls per hour again, but in the Bar Chart dialogue box, click

Select the option to order the bar chart by decreasing Y.

525
 GRAPHS AND QUALITY TOOLS IN MINITAB

Click OK. Click OK again.

526
 SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

Minitab puts the x-axis data in order from biggest to smallest as you might see in a Pareto
chart. This is helpful to someone conducting analysis because you can get an idea of the
highest and lowest values in a data set. Minitab also lets you order in the other way, show
the information as a percent of the total, or accumulate the totals across the x-axis.

Pie Charts in Minitab follow similar steps as bar charts. To create a pie chart, copy the data
from the table below into Minitab.

Process
Step Minutes
A 14
B 18
C 41
D 64
E 12

527
 GRAPHS AND QUALITY TOOLS IN MINITAB

F 4
G 3
H 7
I 23
J 18

Select Graph > Pie Chart.

Click the option for charting values from a table.

Click OK.

528
 SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

Minitab returns the pie chart below.

As with bar charts, you can choose to graph the counts of unique values. Use the number of
errors data from the bar chart example to create a pie chart of counts of unique values.

Select Graph > Pie Chart.

Select the option for charting counts of unique variables.

ble.

529
 GRAPHS AND QUALITY TOOLS IN MINITAB

Click OK.

530
 SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

As with the bar chart example, Minitab counts the number of times each unique variable
exists in the data and presents it in pie chart format.

Also similar to the bar chart, you can choose to have Minitab order the data in your pie
chart. Recreate the pie chart for the first example in this section, which showed the number
of errors at each process step.

Select Graph > Pie Chart.

Click the option for charting values from a table.

Click Pie Options.

531
 GRAPHS AND QUALITY TOOLS IN MINITAB

Choose to order the data by increasing volume.

Click OK. Click OK again.

532
 SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

Minitab returns a pie chart with data in order from smallest to largest going clockwise
around the pie. When compared with the first pie chart of this data, this one is probably
easier to read and draw conclusions from, because you can more easily compare the size of
each section when they are ordered.

Not all the graphs in Minitab are located

run charts, and Pareto charts under the Stat menu. Creating control charts in Minitab was
covered in Unit 5.

Remember that a run chart is a graphical representation of data elements over time. In
Minitab, you can also create a run chart of subgroups means or medians over time. Create
run charts using the data table below, which depicts the production time of various samples
taken during specific hours of a shift. The quality team takes five samples each hour.

533
 GRAPHS AND QUALITY TOOLS IN MINITAB

Production
Hour Time
1 12
1 11
1 13
1 18
1 16
2 13
2 10
2 12
2 19
2 15
3 10
3 7
3 10
3 18
3 14
4 10
4 15
4 12
4 14
4 18
5 11
5 9
5 8
5 14
5 12

In Minitab, select Stat > Quality Tools > Run Chart.

Since the quality team samples 5 per hour, the subgroup size is 5.

534
 SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

Click OK.

535
 GRAPHS AND QUALITY TOOLS IN MINITAB

Graphed this way, Minitab plots the points of each subgroup on the graph in gray and then
plots the mean in blue. The run line shown is for the mean of each subgroup. You can see
that hours 1, 2, and 4 have roughly the same mean and hours 3 and 5 have lower means.

Alternatively, you can graph all of your data points on a single run chart.

In Minitab, select Stat > Quality Tools > Run Chart.

Select the column for production time into the box.

536
 SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

Click OK.

537
 GRAPHS AND QUALITY TOOLS IN MINITAB

Minitab returns a simple run chart of your data.

In chapter 14, we covered how to create Pareto charts in Excel, but you might recall that it
took some data manipulation to get the chart in order, add trend lines, and create label
percentages. In Excel, you had to have totals, or summaries, of the data. In chapter 14, one
of the Pareto charts created was of claims denials per payer, and the data was presented as
totals in a table.

with rawer
data. In the data table below, a quality team reviewed 150 denied claims to determine the
reason they were denied. You can copy and paste this data into Minitab if desired to create
a Pareto chart.

538
 SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

Claim Claim Denial Reason 33 Lacking Medical Reason

1 Timely Filing 34 Lacking Medical Reason
2 Timely Filing 35 Lacking Medical Reason
3 Timely Filing 36 Lacking Medical Reason
4 Timely Filing 37 Lacking Medical Reason
5 Timely Filing 38 Lacking Medical Reason
6 Timely Filing 39 Lacking Medical Reason
7 Timely Filing 40 Wrong Diagnosis Code
8 Timely Filing 41 Wrong Diagnosis Code
9 Timely Filing 42 Wrong Diagnosis Code
10 Timely Filing 43 Duplicate Claim
11 Timely Filing 44 Duplicate Claim
12 Timely Filing 45 Duplicate Claim
13 Timely Filing 46 Duplicate Claim
14 Missing Information 47 Duplicate Claim
15 Missing Information 48 Duplicate Claim
16 Missing Information 49 Duplicate Claim
17 Missing Information 50 Duplicate Claim
18 Missing Information 51 Duplicate Claim
19 Missing Information 52 Duplicate Claim
20 Missing Information 53 Duplicate Claim
21 No Benefits 54 Duplicate Claim
22 No Benefits 55 Duplicate Claim
23 No Benefits 56 Duplicate Claim
24 No Benefits 57 Duplicate Claim
25 No Benefits 58 Timely Filing
26 No Benefits 59 Timely Filing
27 No Benefits 60 Timely Filing
28 No Benefits 61 Timely Filing
29 Lacking Medical Reason 62 Timely Filing
30 Lacking Medical Reason 63 Timely Filing
31 Lacking Medical Reason 64 Timely Filing
32 Lacking Medical Reason 65 Timely Filing
539
 GRAPHS AND QUALITY TOOLS IN MINITAB

66 Timely Filing 100 Duplicate Claim

67 Timely Filing 101 Duplicate Claim
68 Timely Filing 102 Duplicate Claim
69 Timely Filing 103 Duplicate Claim
70 Timely Filing 104 Duplicate Claim
71 Missing Information 105 Duplicate Claim
72 Missing Information 106 Duplicate Claim
73 Missing Information 107 Duplicate Claim
74 Missing Information 108 Duplicate Claim
75 Missing Information 109 Duplicate Claim
76 Missing Information 110 Duplicate Claim
77 Missing Information 111 Duplicate Claim
78 No Benefits 112 Duplicate Claim
79 No Benefits 113 Duplicate Claim
80 No Benefits 114 Duplicate Claim
81 No Benefits 115 Duplicate Claim
82 No Benefits 116 Duplicate Claim
83 No Benefits 117 Duplicate Claim
84 No Benefits 118 Duplicate Claim
85 No Benefits 119 Duplicate Claim
86 Lacking Medical Reason 120 Duplicate Claim
87 Lacking Medical Reason 121 Duplicate Claim
88 Lacking Medical Reason 122 Duplicate Claim
89 Lacking Medical Reason 123 Duplicate Claim
90 Lacking Medical Reason 124 Duplicate Claim
91 Lacking Medical Reason 125 Duplicate Claim
92 Lacking Medical Reason 126 Duplicate Claim
93 Lacking Medical Reason 127 Duplicate Claim
94 Lacking Medical Reason 128 Duplicate Claim
95 Lacking Medical Reason 129 Duplicate Claim
96 Lacking Medical Reason 130 Duplicate Claim
97 Wrong Diagnosis Code 131 Duplicate Claim
98 Wrong Diagnosis Code 132 Duplicate Claim
99 Wrong Diagnosis Code 133 Duplicate Claim

540
 SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

134 Duplicate Claim 143 Timely Filing

135 Duplicate Claim 144 Timely Filing
136 Duplicate Claim 145 Timely Filing
137 Duplicate Claim 146 Timely Filing
138 Duplicate Claim 147 Timely Filing
139 Duplicate Claim 148 Timely Filing
140 Duplicate Claim 149 Timely Filing
141 Duplicate Claim 150 Timely Filing
142 Timely Filing

Select Stat > Quality Tools > Pareto Chart.

The defects or attributes being graphed are the claim denial reasons. There are no
frequencies or variables by which this data is being graphed you simply want to see how

few times in Minitab that you can graph non-numerical data.

d all be counted as separate attributes by the

program. The Six Sigma expert has to ensure data is prepared appropriately and all
categorical descriptions are the same before running a Pareto chart. One way to do this is to
prepare data in Excel, where you can use the find and replace option to ensure all entries
are corrected with regard to spacing or spelling.

541
 GRAPHS AND QUALITY TOOLS IN MINITAB

reasons in the data. If you have ten or more unique attributes, you might select the option
to combine remaining defects after 95 or 90 percent. The defects that account for the final

Click OK.

542
 SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

Minitab displays a Pareto chart with the bar chart, running percentage, and detailed count.
From this chart, you can quickly see that approximately 80 percent of the denied claims are
caused by the first three denial reasons.

Another convenient tool in Minitab is the ability to generate Gage R & R worksheets and
interpret the results of a Gage R & R study. Gage R & R studies were previously covered in
chapter 13 on the Measure phase.

Create a Gage R & R worksheet by selecting Stat > Quality Tools > Gage Study > Create Gage
R & R Study Worksheet.

543
 GRAPHS AND QUALITY TOOLS IN MINITAB

Minitab lets you input a variety of information to set up a randomly-ordered Gage R & R
Study. In the example above, Mike and Ann are quality auditors reviewing dollies that come
off a manufacturing line. In this study, each person is going to review five parts, and they
will do so twice.

Minitab lets you select the number of parts to be reviewed, the number of reviewers or
operators, and the number of replicates, or times the review or measurement is to be
repeated. You can also click in the white space next to each part and operator number to
name those items for clarity later.

Once you set the dialogue box up as above, click OK.

544
 SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

Minitab generates a worksheet for the Gage R & R Study. It randomizes how the parts are
presented to each person for each replicate. Remember from chapter 13, this is a critical
step in a measurement analysis, because it helps reduce the chance that the operator will
remember the measurement they previously recorded.

Mike and Ann might then perform measurements on each of the parts. In this case, they are
measuring the parts in inches to ensure they conform to a required range for quality. The
Six Sigma expert running the Gage R & R is not concerned with whether the parts fit that
range he or she is concerned with whether the measurement system works. Can Ann and

545
 GRAPHS AND QUALITY TOOLS IN MINITAB

Mike repeat the measurements consistently both within their own sets of data and when
compared to each other?

The measurements are taken and entered into Minitab.

Since there is more than one operator and each is measuring all the parts, then the Gage R
& R Study (Crossed) option is chosen to analyze the measurement system. You can see
below, Minitab offers several options for Gage R & R studies, as well as information about
when to use each.

546
 SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

n for operators and the

else as shown and click OK.

547
GRAPHS AND QUALITY TOOLS IN MINITAB

548
 SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

Minitab generates a series of graphical analysis to help determine whether the

measurement system works. The two charts at the bottom show that each operator
duplicates the results of the other operator fairly closely. The Xbar chart of each operator
looks

the information for each operator is very similar the mean appears to be the same, as is
the spread of data.

The first graph details where variation shows up in the measurements. For both Mike and
Ann, the variation came when comparing the measurement of one part in the first run to
the measurement of the same part in the second run. Even with that slight variation
between runs, however, the overall variation is very small and this Gage R & R indicates that
the measurement system is working.

549
Chapter 21 introduced hypothesis testing and included information on how to select the
right test and how to interpret the results of a test based on p-value. Chapter 19 covered
correlation and regression and how to interpret the results of tests run in Excel. In this

Because interpreting test results and setting up hypothesis testing has been covered in
detail in other chapters, this chapter simply offers some guidance for using Minitab to run
the tests, including where to find the tests and how to enter information for various tests.

tests is very similar.

Many of the tools used to run hypothesis tests are found under Basic Statistics in the Stat
menu.

Sometimes, you need to know the mean, standard deviation, or some other descriptive
statistic about your data sample. When running many hypothesis tests, you need such data.

one time, and you can store them in a worksheet for reference as you work on analysis.

Time Temperature A
1 30
2 35
3 37
4 40
5 41
6 48
7 57
8 55

550
 SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

9 59
10 62
11 63
12 65

Copy the data above into Minitab.

Select Stat > Basic Statistics > Display Descriptive Statistics.

Select the temperature column as the variable.

Click Statistics.

Check the box next to each statistic you would like to display.

551
 THE STAT MENU IN MINITAB

Click OK. Click OK again.

Minitab displays the statistics in the session window.

Complete the exact same series of steps, but choose Store Descriptive Statistics instead of
Display Descriptive Statistics. Minitab will place the statistics in the workbook.

552
 SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

Use the data of temperature over time you copied above to run a regression analysis in
Minitab. The question you want to answer is: Is there a correlation between the
temperature and the time?

Select Stat > Regression > Fitted Line Plot.

Choose the temperature column as the response variable and the time column as the
predictor variable.

Click OK.

553
 THE STAT MENU IN MINITAB

Minitab returns the fitted line plot along with R-squared values, which you can interpret
following the information from chapter 19.

Most of the relevant hypothesis tests in Minitab are found either under Stat > Basic
Statistics or Stat > Nonparametics.

The most relevant tests found under Stat > Basic Statistics are:

- 1 Sample Z
- 1 Sample t
- 2 Sample t
- Paired t
- 1 Proportion
- 2 Proportion
- 1 Variance
- 2 Variance
- Normality Test

554
 SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

You might notice that the tests found under Basic Statistics tend to be the tests used for
discrete or normally distributed continuous data. The test for nonnormally distributed
continuous data can be found under Stat > Nonparametrics.

Example 1: 1 Proportion Test

problems discussed in the hypothesis testing chapter.

555
 THE STAT MENU IN MINITAB

Problem: A school administrator believes that the students in a certain classroom are being
impacted by some environmental issue with that room. The administrator wants to see if
the students in that room perform statistically lower on tests than the students throughout
the school; she has already noted that different classes in that room are taught by other
teachers.

The administrator gathers data. The rate of failure for classes taught in any other classroom
is 20 percent. Out of 142 students who have a class in the classroom in question, 38 are
failing that particular class.

The question is, is the rate of failure in the classroom statistically greater than the rate in
other classrooms? You can look back to the 1 Proportion section of chapter 21 to see how
the null and alternative hypotheses are created.

To run the test in Minitab, select Stat > Basic Statistics > 1 Proportion.

The number of events is the number of outcomes that you are measuring. Here, the
administrator was measuring failure rates, and there were 38 failures.

556
 SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

The number of trials is the total number of data points. In this case, the school
administrator looked at scores for 142 students.

Check the box to perform a hypothesis test and enter the hypothesized proportion. In this
case, you are comparing the actual failure rate of the classroom group to the overall failure
rate of 20 percent. The hypothesized portion is written as a decimal between 0 and 1.

Click Options.

Minitab usually defaults the confidence level to 95, which would set alpha at 0.05. Unless
you have a specific need for greater confidence, you can leave this as is.

Select the appropriate alternative hypothesis from the drop down menu. In this case, we
are testing whether the proportion of our data is greater than the hypothesized proportion.

Click OK. Click OK again.

Minitab displays the results, including a p-value you can use to decide to reject or not reject
the null hypothesis.

Example 2: 1 Sample T Test

557
 THE STAT MENU IN MINITAB

Problem: An agriculture company growing corn produced an average of 168 bushels of corn
per acre each year for five years. The company made a change to its seeding process in an
attempt to increase the yield of its fields. The harvest following the seed change resulted in
an average of 175 bushels of corn per acre each year over 500 acres of fields. The standard
deviation is 2.2 bushels. Did the seeding change result in a statistically different outcome in
yield per acre?

Select Stat > Basic Statistics > 1 Sample T.

Choose summarized data.

Enter the sample size, sample mean, and sample standard deviation from the problem.
Check the box to perform a hypothesis test and enter the hypothesized mean. In this case,
you want to know whether the new mean is greater than the old mean, so the old mean is
the hypothesized mean.

Click Options.

Choose the appropriate hypothesis test from the drop down menu.

558
 SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

Click OK. Click OK again.

Minitab displays the results, including a p-value you can use to decide to reject or not reject
the null hypothesis.

Example 3 2 Sample T Test

Copy the data from the table below into Minitab. If you recall from chapter 21, a law firm
believed there was a statistical difference between the time employees on two different
teams were spending on tasks.

Minutes Spent on Quarter-Hour Tasks

Attorney A's Team Attorney B's Team
10 8
8 7
15 10
16 12
17 18
5 6
8 9
16 12

559
 THE STAT MENU IN MINITAB

12 11
11 11
8 9
9 8
15 14
17 15
22 16
25 12
30 19
2 8
9 12
4 6
5 8
7 11
8 12
12 13
13 15
15 17
16 13
18 15
19 15
20 12
22 21
25 32
15 18
14 11

Select Stat > Basic Statistics > 2 Sample T Test.

From the drop down, select that each sample is in its own column. You can also choose to
analyze samples when the data is all in one column or you can enter summarized data
about each sample.

Select the appropriate columns into the Samples 1 and 2 boxes. In this case, you are
selecting the columns that hold the data for each of the teams.

560
 SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

Click Options.

In this case, we only want to know if

leave the alternative hypothesis as simply not equal to the hypothesized difference. In this

0.

Click OK. Click OK again.

561
 THE STAT MENU IN MINITAB

Minitab displays the results, including a p-value you can use to decide to reject or not reject
the null hypothesis.

All of the hypothesis tests in Minitab are run using basically the same steps as in the
examples above. For most test types, you can choose to test the raw data or you can enter
summarized data. Remember, each type of hypothesis test requires different summarized
data. Proportion testing requires rates, means testing usually requires standard deviations
and means of samples, and variance testing requires sample variance. Some of the tests you
.

If you have the right descriptive statistics or have your data in columns in Minitab, it is a

interpreting, the tests. It is usually in determining which test is the right option for your

your data for normality before running hypothesis tests. Remember that you can easily test
for normality by running the Graphical Summary under Basic Statistics.

Some other things to keep in mind when dealing with hypothesis tests in Minitab are that
you should always click the Options button to verify that your confidence interval is set
appropriately and that the appropriate hypothesis test is selected. Minitab defaults to a
confidence interval of 95% and an alternative hypothesis of not equal. However, if you are
working during the same session
562
 SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

you ran that specific hypothesis test, the program will remember what settings you entered
and apply them automatically to the new test you are running. This can be a problem if you
want to run a test on different data or run a test on the same data with a different
confidence interval or alternative hypothesis and you forgot to make the change.

Equivalence Test.

If you are working in Minitab and have questions about these tests, you can access the Help
menu for information about choosing, running, and interpreting each test.

Select Help > StatGuide.

The topics are organized just like the Minitab menus. Simply select the test you want help
with and Minitab offers a dozen or so pages of help content for each.

563
THE STAT MENU IN MINITAB

564
Analysis of variance, often referred to as ANOVA, is a hypothesis test that deals with more
than two populations or factors of X. If you recall from chapter 21, we discussed a variety of
hypothesis tests for dealing with single factors of x or two factors of x. For example, you
learned to conduct a hypothesis test to determine if the mean of a sample was statistically
equal to a target mean. You also learned to compare the means of two groups of data with
each other for the same reason. ANOVA lets you perform this test with more than two
groups of data.

Analysis of variance is often used when you have a combination of both discrete and
continuous variables if the independent variance is discrete a list of employees, for
example and the response variable is continuous a list of errors or length of time worked
ANOVA can be a valuable tool for analysis.

- When testing for one mean usually against a target you would use the Z test or T
test. Remember, the Z test is o

you gather some information so you can run more accurate analysis with other tests.
- When testing for two means which involves comparing two samples against each
other you would use the 2 sample T test or the paired T test.

565
 ANALYSIS OF VARIANCE (1-WAY ANOVA)

- When testing for three or more means, you can use a specific ANOVA test called the 1-
way Anova.

To better understand these designations, consider the real-world applications below.

A Six Sigma team is working to reduce the process time in a manufacturing process. The
team was provided with historic baseline metrics for the process. Specifically, in the last
quarter, the process averaged 35.8 minutes per output. After working through Define,
Measure, and Analyze phases, the team has put an improvement in place they believe will
reduce the average time per output. After piloting the change in Improve, the team takes
new measurements. To compare the new sample to the historical baseline, the team would
use the 1-sample T test.

In a different scenario, a Six Sigma team is working to increase the number of calls that can
be handled by a call center team in a given day. Call center management has repeatedly
asked for additional employees, but executive leadership wants to find out whether
efficiency improvements can increase production without adding additional employees.
After the first half of the DMAIC process, a Six Sigma team is ready with possible solutions.
To verify these solutions, the team decides to implement the changes in half the call center.
Over the course of one month, the team measures performance for both halves of the call
center. At the end of the month, the team has two sets of data and wants to answer the
question: Is the average production for the altered group greater than the average for the
unaltered group? In this case, the team would use the 2-sample T test.

The two examples above are in line with many of the examples used in the chapter on
hypothesis testing. Consider the scenario below, which is slightly different.

A restaurant chain wants to know if its branding and customer-facing activities are working
equally across all locations. Specifically, the company decides to look at the customer
satisfaction scores for each restaurant location. The chain includes five locations.
Satisfaction scores are collected via phone, web, and written surveys at each location. The
company averages the scores each week and reports those numbers. After several months,
the company has five sets of averages. If the company wants to know whether the scores for
any location differs statistically from the scores from other locations, then the correct test is
the 1-way ANOVA.

The way a Six Sigma expert approaches a 1-Way ANOVA is the very similar to the way he or
she would approach any other hypothesis test. First, you begin with a real-world problem
566
 SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

a practical or business problem. In the example above, that problem is whether there is
variation in customer satisfaction among the various restaurant locations.

As with any hypothesis test, a null and alternative hypothesis is required. The null
hypothesis is that there is no difference in the means of the samples.

The alternative hypothesis is that at

least one of the means is not
Between versus Within Sample
statistically the same. Variance
After stating the null and alternative ANOVA calculations compare the between
hypothesis, a Six Sigma expert should sample variance with the within sample
verify that any assumptions within the variance. Between sample variance is the
model are appropriate. This includes variance that occurs across all of the samples
assumptions about errors as well as being analyzed. Within sample variance is the
some basic assumptions about the variance that occurs within a single sample or
ANOVA model. group.

First, the samples used for the ANOVA The reason this distinction is important is that
must be randomly selected. the within sample variance obviously impacts
Remember, this is always a necessary the between sample variance. If you are
assumption for inferential statistics. dealing with five separate samples and each
sample has a big variance, then the variance
After ensuring samples are randomly across the samples is likely to be large as well.
selected, Six Sigma experts must Part of the ANOVA calculation compares
validate six other assumptions about between sample variance to determine if it is
data before running a 1-way ANOVA large enough relative to within sample variance
test. to denote a statistical difference.
1. The dependent variable, or
outcome, must be continuous in nature. This means that it is a ratio or an interval.
In previous examples, the dependent variable included the customer satisfaction
score (numerical), the number of calls handled per day, and the cycle time for a
specific output. All of these are examples of continuous data. Anything that can be

567
 ANALYSIS OF VARIANCE (1-WAY ANOVA)

measured in time, temperature, feet and inches (or centimeters and meters),
money, or ratios is typically continuous in nature.
2. The independent variable list contains two or more unrelated groups or categories.
Teams A, B, and C are three independent variables. If you are measuring
performance for five different workers, then you might have five independent
variables. In the example about restaurants, the five different restaurants are
independent variables. Categorical variables can include people, times, shifts,
teams, departments, locations, various demographic groups (age, gender,
ethnicity), or professions.
3. Observations are independent. First, you must ensure there are no dependencies
within groups. If you are measuring the performance of three different teams in a
department but several employees work on multiple teams, then the results are not
completely independent. Second, you must ensure that observations within groups
are not dependent.
4. -

explainable outlier. As a Six Sigma expert, you do have to review your data,
investigate outliers, and discard them appropriately. For example, if a Six Sigma
expert is reviewing the response times for customer emails, he or she might review
data samples from nine different employees. The data for one employee includes
three outliers where emails were responded to after a much longer amount of time
than all other samples seem to indicate. The Six Sigma expert might investigate this
and note that the employee in question was on a short medical leave, which
skewed results. The Six Sigma expert could remove those three data points from his
or her calculations, but it would be important to note these outliers. The fact that
this particular problem can happen is still important to the overall process
improvement. The Six Sigma team might recommend instituting a process change
that addresses this issue, but they might analyze data minus these outliers to draw
other conclusions.
5. The dependent variable (as described in number 1 above) should be normally
distributed or approximate the normal curve within each group. In the example
above, the Six Sigma team gathers data for nine employees. Within each of these
nine data sets, the data should be normally distributed.
6. The variance within each set has to be statistically equal to the variances of other
sets also called homogeneity of variances. The test for equal variance can be run
in Minitab and will be covered later in this chapter.

568
 SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

After validating assumptions and setting up the hypotheses, a Six Sigma expert can run the
1-
Minitab to run the tests. When using statistical analysis software, you typically spend more
time on validating some of the assumptions than you do running the test itself.

To understand how to validate all the assumptions, run the 1-Way ANOVA, and interpret
the results, consider a real-world example.

A mail-order book company wants to improve customer satisfaction with deliveries. A Six
Sigma team working on the process has identified packing materials as a possible factor in
negative customer satisfaction scores. The mail-order company currently ships books via
cardboard boxes with no other protection, which the team believes is contributing to an
increased amount of damage to shipments before they reach the customer or when the
customer opens the package.

To test this theory and pilot two proposed solutions, the team implements additional
packing options. Some books will be shipped in the regular packaging the box without any
additional materials. Others will be shipped in boxes with packing peanuts and still others in
padded envelopes. The team implements the different packing methods at three different
shipping stations and records an overall customer satisfaction score for each shipment
based on a scale from 1 to 10 with 10 being the most satisfied. That data is presented
below.

569
 ANALYSIS OF VARIANCE (1-WAY ANOVA)

Original Padded Packing

Packaging Envelope Peanuts
5 5 6
5 5 8
4 3 8
6 6 6
4 6 5
6 8 5
6 7 8
6 6 7
5 7 8
4 5 6
6 6 6
6 5 10
5 6 7
4 4 9
5 6 6

To follow along with the 1-Way ANOVA for this example, copy the above data into Minitab.

Is the dependent variable continuous?

Yes, it is measured in interval form.

Is the independent variable made up of two or more unrelated groups?

Yes, the independent variable is how the book is packaged.

Is there independence of observation?

and are not

for example.

It is worth noting here that the team has implemented the various packing at three
different stations. It would be a good idea for the team to ensure that the packing stations

570
 SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

were otherwise treated identically to keep any other factors from influencing the outcome
of the pilot.

Are there any significant outliers in the data?

Graphical analysis can be an easy way to check for outliers. The box plot (also called the box
and whisper plot) is a good way to check quickly for outliers. Boxplotting in Minitab was
covered in Unit 6, but you can run one on the data above now.

1. Click Graph > Boxplot

2. Select multiple Ys
3. Click OK
4. Select all three data columns into the Graph Variables box
5. Click OK.

Minitab returns the graph above, which shows no outliers for any of the three sets of data.
What if someone recording the customer satisfaction scores made a mistake and entered
the number 20 for one shipment? The boxplot analysis changes.

571
 ANALYSIS OF VARIANCE (1-WAY ANOVA)

In the graph above, you can see an outlier dot high above the box for packing peanuts. A Six
Sigma expert reviewing this graph would see that outlier and realize that it was a
measurement error remember, the scale for customer satisfaction was only supposed to
go up to 10. Because there is an explanation for the outlier, it can be removed.

Is the data in each group normal?

The quickest way to test normality for each set of data is probably to run a graphical
analysis in Minitab. This was also covered in Unit 6.

1. Select Stat > Basic Statistics > Graphical Summary

2. Select the column for the first set of data into the Variables box.
3. Make sure the confidence level is set to 95.0.
4. Click OK.

572
 SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

Minitab returns the following chart and data.

Summary Report for Original Packaging

Anderson-Darling Normality Test
A-Squared 0.36
P-Value 0.405
Mean 5.1244
StDev 0.7449
Variance 0.5549
Skewness -0.18142
Kurtosis -1.03701
N 15
Minimum 3.7832
1st Quartile 4.3648
Median 5.1199
3rd Quartile 5.7692
Maximum 6.3274
95% Confidence Interval for Mean
4.0 4.5 5.0 5.5 6.0 6.5 4.7119 5.5369
95% Confidence Interval for Median
4.4348 5.7557
95% Confidence Interval for StDev
0.5454 1.1748

95% Confidence Intervals

Mean

Median

4.50 4.75 5.00 5.25 5.50 5.75

The p-value for the Anderson-Darling normality test is 0.405. Remember, when testing for
normalcy, the null hypothesis is that there is no difference between the data and the
normal curve. Since the p-value is above our alpha level (0.05), then we fail to reject the null
hypothesis and accept that the data is normal.

The graphs generated for the other two sets of data are included below. You can see that
the p-values for each also allow us to accept the null hypothesis and validate this
assumption. Yes, the data for each group approximates the normal distribution.

573
ANALYSIS OF VARIANCE (1-WAY ANOVA)

574
 SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

Is there a homogeneity of variances? Are the variances between groups relatively equal?

Minitab includes an option for testing for equal variances that provides two different p-
values. One is the p-

is normal. This test was not previously covered in Unit 6.

1. Select Stat > ANOVA > Test for Equal Variances

575
ANALYSIS OF VARIANCE (1-WAY ANOVA)

576
 SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

2. Select all the data columns into the Responses box.

3. Click OK.

577
 ANALYSIS OF VARIANCE (1-WAY ANOVA)

Minitab returns an interval graph of all the columns of data along with two statistical values.

For the test for equal variance, you are usually concerned with the p-

the null hypothesis, which is that there is no difference in the variance among the data. In
this case, even a look at the interval graph helps you make this determination. The ranges
are certainly different, but the intervals are not extremely different in length.

578
 SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

To consider a contrasting example, copy the following data table into Minitab.

A B
4.01 10.00
4.91 10.06
6.38 10.04
7.98 10.00
6.00 10.06
10.00 9.97
6.50 9.96
4.14 10.01
9.00 10.05
6.45 10.08

Run the test for equal variance under the ANOVA menu using just the two columns of data
from the above table.

579
 ANALYSIS OF VARIANCE (1-WAY ANOVA)

You can quickly see that these two sets of data do not have equal variance. First, the
intervals make it fairly obvious. The variance in set B is much smaller. The p-value for
est is much smaller than the alpha value, which means you reject the null
hypothesis and accept the alternative hypothesis that there is a difference in the variance.

You can change the alpha value for the tests by selecting Options on the Test for Equal
Variances dialogue box.

580
 SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

You can change the type of graph displayed by selecting Graphs from the Test for Equal
Variances dialogue box.

581
 ANALYSIS OF VARIANCE (1-WAY ANOVA)

Here are the original three sets of data graphed via box plot.

Now that all the assumptions are verified, you can run the 1-Way ANOVA test.

Select Stat > ANOVA > One-Way.

Select the drop down option that response data are in separate columns.

Select all three of the data columns (the data for the original packing, padded envelopes,
and packing peanuts) into the Responses box.

582
 SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

You can click Options to set the confidence interval or ensure it is still set at 95%, if desired.
Then click OK.

Click OK on the main dialogue box. Minitab will perform the 1-Way ANOVA calculations.
Depending on how much data is in columns when you run this test, it can take a few
seconds.

Minitab defaults to returning two things: a graphical analysis of the data and the ANOVA
results in the session window.

583
 ANALYSIS OF VARIANCE (1-WAY ANOVA)

Interval Plot of Original Pac, Padded Envel, ...

95% CI for the Mean
8.0

7.5

7.0

6.5

6.0

5.5

5.0

4.5

Original Packaging Padded Envelope Packing Peanuts

The pooled standard deviation is used to calculate the intervals.

584
 SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

One-way ANOVA: Original Packaging, Padded Envelope, Packing

Peanuts

The data from the session window is included above. Highlighted are the hypothesis and the
p-value. The null hypothesis is that all the means are equal. The alternative is that at least
one of the means is statistically different.

585
 ANALYSIS OF VARIANCE (1-WAY ANOVA)

The p-value is less than the alpha value of 0.05, which means you reject the null hypothesis
and accept the alternative hypothesis that at least one of the means is different.

be an issue with one or more processes. If the team was comparing mean production times
between various teams on the same process, they might assume that the teams would
complete work in roughly the same time. If the 1-
Six Sigma team can then ask itself: What is it that makes the outcome statistically different?
Is there a team doing it better than everyone else? If so, can the Six Sigma team implement
solutions from that group across the other groups? Is there a team performing worse than
everyone else? If so, what is going on with that team and how can an improvement be
made?

you can still run a 1-Way ANOVA test in Minitab. Minitab uses a different statistical
calcu -value in such a case.

Use the data we previously found to have unequal variances to run this test in Minitab.

A B
4.01 10.00
4.91 10.06
6.38 10.04
7.98 10.00
6.00 10.06
10.00 9.97
6.50 9.96
4.14 10.01
9.00 10.05
6.45 10.08

Select Stat > ANOVA > One-Way.

Select the drop down option that response data are in separate columns.

586
 SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

Select the data columns for A and B into the Responses box.

Select Options.

Uncheck the box for assume equal variances.

Click OK. Click OK again.

587
 ANALYSIS OF VARIANCE (1-WAY ANOVA)

As in the previous example, Minitab generates both a graphical analysis and the test data in
the session window.

One-way ANOVA: A, B

588
 SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

You can see from the Minitab results, equal variance was not assumed for this data. You
also have a p-value that is effectively 0, which means you reject the null hypothesis and
accept the alternative that one of the means is statistically different: something we already
did just by looking at the graphical interpretation of this data before. As stated previously,

statistical analysis.

At the end of the last unit, we noted that you could use the help menu in Minitab to look up
information about all of the functions and read in-depth instructions and summaries on
various types of tests. You can also use the Minitab Assistant to help you choose a
hypothesis test including a 1-Way ANOVA and run that test.

Select Assistant > Hypothesis Tests.

589
 ANALYSIS OF VARIANCE (1-WAY ANOVA)

You can have Minitab help you choose a test to compare one sample with a target, two
samples with each other, or more than two samples. Minitab will apply the same
information included in this chapter and the previous chapter on hypothesis testing to help

covered in detail in this book.

Minitab presents you with a second diagram.

590
 SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

If you are comparing the means for more than two samples of continuous data, you would
use the 1-Way ANOVA described in this chapter. If you are comparing standard deviations,
however, you use a different test. If you are comparing attributed data, then you would use
the chi-square tests. You can hover over any of the tests and click and Minitab will open the
dialogue box for performing that test.

591
 ANALYSIS OF VARIANCE (1-WAY ANOVA)

Minitab offers several assistant wizards. For example, below is a screenshot of the graphical
analysis assistant menu.

Here, Minitab helps you choose a graph or visual analysis that best matches your data and
purposes. You can use the assistant tool for help with regression analysis, measurement
system analysis, control charts, and design of experiments, which will be covered in the next
chapter.

592
hypothesis tests, and draw conclusions from those tests. But simply analyzing data no
matter how accurate you can be
process. It can be easy to get lost in the analysis and forget about the need to make
improvements, but continuous improvement is the very heart of the Six Sigma and DMAIC
methodology.

Experiments are one way that Six Sigma experts and teams bring analysis and
improvements together. Douglas C. Montgomery, author of Design and Analysis of
Experiments provides a definition for the type of experiment we talk about in relation to a
Six Sigma project.

An experiment is a test or series of tests in which purposeful changes are

made to input variables of a process or system so that changes in the
output responses can be observed and identified.

The fundamental concept of design of experiments, or DOE, is that all processes can be
ultimately distilled to the Y = f(x) equation. There are some factors, x, that work together to
create some output, Y. The job of the Six Sigma team, as has been stated throughout this
book, is to find out what factors (x) actually have an impact on Y and how changes in those
factors will change the outcome.

One way teams can achieve this is to run a designed experiment. This involves making
purposeful changes in one or more of the input variables Xs and then measuring and
analyzing what occurs with the output, Y. Designed experiments have to be controlled and
well-documented or improper conclusions can be drawn. The process for planning and

between DOE and the scientific method that most learn in primary school.

593
 DESIGN OF EXPERIMENTS

Sometimes, enough data exists or enough change exists within the current process that a
Six Sigma team can make decisions about factors without running a designed experiment. In
these cases, the data is already available or can be captured in a way to isolate various
inputs so that a Six Sigma team can run analysis and hypothesis tests without recreating
physical scenarios. For example, in an agricultural setting, a company might already be using
various types of planting processes to grow the same crop. Teams could gather data about
each type of planting process, ensure no other variables exist between those processes, and
then draw valid conclusions about how the planting processes impact the harvest
outcomes.

There are benefits and disadvantages to analyzing existing data rather than running a
designed experiment. The first benefit is that basing primary analysis and initial solution
selection on existing data typically reduces the timeline for the entire DMAIC process. A
team might still have to pilot a solution, measuring and analyzing the results to ensure they

experiment to make those first predictions. Second, basing decisions on existing data or
processes can reduce the overall cost of a DMAIC project. This is especially true in processes
that involve machines or parts that would have to be retooled to create a change in some

gather information upon which to draw such conclusions, and the Six Sigma expert must use
other tools to analyze the process so the team can draw conclusions and suggest
improvements for a more formal pilot or test.

can use the existing

data in a process when a process includes multiple changes to a factor. You would even run
the same analysis that you might run if you made those changes yourself as part of an

process. For example, a Six Sigma team might be working on how document printing is
managed in a large law firm. The team wants to know if the copy-print station to which
documents are sent impacts the speed with which the jobs are processed. The copy and
print machines have internal logs that record when a job is received and when it finished
printing, so the team can compare the speed of the process from all the stations using
ANOVA or DOE analysis without actually designing and running an experiment. The team
could simply access the historical data for the process from last week or any time period.

594
 SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

W -print process are a variety of external factors including

how many staff were working at a given time, whether machines experience maintenance
needs or paper jams, and whether staff were interrupting printing jobs to handle copying,
scanning, or faxing jobs at the same machines. While the historical data might provide some
valuable information about the difference in productivity for each machine, the team might
not be able to rule out some of these other factors as causation without a designed
experiment that captures all the possibly relevant data.

A well-designed experiment helps you understand what variables provide the most
influence on the process output. This helps a Six Sigma team understand where time,
money, and other resources should be spent. In the example above, the experiment might
help the team rule out external factors and understand if the type or location of the copy-
print station is a major factor in productivity.

Another reason to run an experiment is that it can help you understand how to set inputs to
create the best possible outcome. In one of the earliest chapters of this book, we used the
example of a manufacturing process that makes chocolate bars. Each of the inputs the
ingredient amounts, the mixing time, and the mixing temperature has to be set at the
right level to create an optimal product. A Six Sigma team working on that process might
conduct experiments to test various levels of inputs to find the right combination for the
desired output.

Sometimes, Six Sigma teams want to understand how to control or set inputs that are

chocolate bar example, if the team has found optimal settings for the ingredients, time, and
mixing temperature, and they still see variation in the outcome of the process, they might
then consider influential inputs. Does the temperature in the room influence the process of
mixing and setting the chocolate? If the team decides it might, they could conduct an
experiment to test various room temperatures to find the right level for the best process
outcome.

control the temperature in a climate-

temperature outside, for example. Experiments can help teams understand how to set all
the inputs they can control to reduce any impact on variation and outcome from the inputs
that cannot be controlled.

595
 DESIGN OF EXPERIMENTS

A designed experiment is just that it is designed statistically to test a variety of factors.

Although a designed experiment does have, at its very basis, the idea of trial-and-error,
teams that use design of experiment are not simply taking a best guess approach to the
matter. A best-guess trial-and-error approach is actually not efficient in most cases,
although some people might think so.

To better understand the different between a designed experiment and a best-guess

approach, consider a real-word application. A team is working to improve a process for a
regional furniture warehouse that delivers furniture to retail stores based on weekly orders
from the stores. The team wants to reduce the average delivery turnaround by 24 hours
and does have the baseline measurement the average turnaround time for orders is
currently 72 hours.

Using a best-guess approach, the team would identify what it thinks are the combination of
inputs that are most likely to impact delivery times. The team might decide that factors
include the delivery driver, the dispatcher who provides instructions and directions,
whether the driver has a GPS in the truck, and how many pieces of furniture are being
delivered. To find out which factor most contributes to the delivery time or discover which
factors can be changed to decrease average times, the team would then have to run a series
of trial-and-error tests.

The team sets the factors and runs a scenario. If the delivery time is greater than 48 hours,
the result is not as desired. The team changes one of the factors and runs the test again.
The team has to repeat the process until it finds a magic combination. Not only is this
inefficient, but it uses an approach called the OFAT Approach, or One-Factor-At-a-Time
approach. Only one input setting is changed for each trial and a measurement is taken. The
measurements are compared to make decisions about each factor. In the furniture
example, the team might note that delivery times are faster when a GPS is on board the
truck. The might also note that deliver times are faster when a more experienced driver is
on board the truck, regardless of the inclusion of a GPS.

One of the major problems with testing one factor at a time other than inefficiency--is

Yes, the presence of a GPS increases efficiency. Yes, the presence of an experienced driver
increases efficiency. But what happens when both a GPS and an experienced driver are on
596
 SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

board a truck? Does efficiency increase at some exponential level, or does it simply increase
to the maximum amount of either the experience-related increase or the GPS-related
increase? Likewise, what happens when a negative factor is coupled with a positive factor?
If delivery times increase with each item added to the order, what happens to the impact of
the experience of the driver on efficiency? One-factor-at-a-
answer these questions.

Why Do People Think This is a Good Approach?

Based on the scenario above, you can see that the best-guess approach is cumbersome and
time consuming. The problem is, many people within an organization often favor this
approach because it has worked well before. If you are very familiar with a process, you
might be able to get a good result within the first few guesses. In our furniture moving

having a GPS on board each deliv

to take longer to deliver more furniture.

Because of a tendency to favor a best-guess approach, Six Sigma experts might have to
work to explain the importance of using statistical analysis to
important for teams to understand that a best guess approach does usually lead to more
trials and tests than are necessary, which creates waste, or muda. Yes, teams might get
lucky, especially if someone on the team or a subject-matter-expert being utilized by the
team is extremely familiar with the process and, thus, understands how the inputs impact
the outcomes at a granular level. However, that familiarity with the process can create the
opposite impact. The person might be so close to the process and so ingrained with the
traditional way of doing things that he or she cannot make valuable best guesses.

Even if the first guess regarding changes to the inputs yields a result that is within the
parameters the team was looking for, it might not be the best possible guess. The team in
the delivery example wanted to reduce deliver times by an average of 24 hours. Based on
historical data, any combination of inputs that averages 48 hours or less would achieve this
goal. But what if there is a combination of inputs that avera
that be something valuable for the team to know? If the combination of inputs is viable to
achieve, then the team could not only deliver on the original goal, but deliver well above the
should always set out to do, because setting the bar
too high can have negative results on a project. But knowing that excellent results are
possible is valuable because it lets the team make a more educated decision about the
solution.

597
 DESIGN OF EXPERIMENTS

The solution to all the problems of inefficiency and lackluster data described above is design
of experiment, or factorial experimentation. Design of experiment is also called statistically
designed experiments, and it involves running a series of experimental runs that vary all the

planned runs, which means you have all the data for all the changes to inputs at that time.

For example, consider the factors from the furniture delivery example again: number of
pieces being delivered, experience of the driver, whether a GPS is on board, and the
method of providing instruction or direction from the dispatcher. For this example, assume
that each of these factors could be set with a low value and a high value as follows.

Experience of driver: 0 years, 1 year

GPS: No, Yes
Dispatcher: No policy, follows a policy
Number of pieces of furniture: 1, 5

In a grid, you would create a list of all possible runs. Each run creates a unique grouping of
the possible factors, which means the number of runs you have is determined by the
number of factors and possible settings that you have. The grid for the delivery factors is
shown below.

598
 SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

Run Number Driver GPS Dispatcher Pieces of

Experience Furniture

1 0 years Yes Follows 1 piece

Procedure

2 0 years Yes Follows 5 pieces

Procedure

3 0 years Yes No Procedure 1 piece

4 0 years Yes No Procedure 5 pieces

5 0 years No Follows 1 piece

Procedure

6 0 years No Follows 5 pieces

Procedure

7 0 years No No Procedure 1 piece

8 0 years No No Procedure 5 pieces

9 1 year Yes Follows 1 piece

Procedure

10 1 year Yes Follows 5 pieces

Procedure

11 1 year Yes No Procedure 1 piece

12 1 year Yes No Procedure 5 pieces

13 1 year No Follows 1 piece

Procedure

14 1 year No Follows 5 pieces

Procedure

15 1 year No No Procedure 1 piece

599
 DESIGN OF EXPERIMENTS

16 1 year No No Procedure 5 pieces

This information is then used to generate a worksheet for running the experiment; once
data is collected, a Six Sigma expert analyzes the outcome. Minitab actually has tools that

the end of this chapter and in the next chapter.

Note that the above experiment is an example of a 2k factorial designed experiment. In 2k

factorial design, you limit the factor levels to two: a high and low factor. The driver has less
than a year (0 years) experience or 1 or more years of experience. The GPS is there or it

Experiments that include more than two levels more than a low and a high for one or
more factors are full factorial experiments. The examples in this chapter all deal with 2k
factorials; examples of full factorials can be seen in chapter 30.

As with hypothesis testing, most experiments follow a road map of steps required to define
the elements, set up data collection, and analyze the data collected. These steps are

requirements are slightly different when dealing with full factorials.

Step 1: What, in business or practical terms, is the problem?

As with anything in Six Sigma, you begin with a practical business consideration or issue.

larger problem. For

example, a healthcare facility might want to address the rate of staph infections for
admitted patients. This might be the overall project for the team: reduce the current rate of
staph infections, which is 17 percent.

During the project, the team discovers that clinical staff are not always compliant with
sanitation protocols when treating patients. For this particular situation, the practical
problem might be stated as:

600
 SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

In a 10-point evaluation for sanitation compliance, clinical healthcare workers are averaging
a score of 7.1. The hospital requires an average of at least 9.0.

The business problem or concern that the experiment is addressing should be related to the
sanitation compliance
could be linked to the fact that patients are getting staph infections.

Step 2: What is the objective of running a designed experiment?

Unlike some of the other analysis discussed in this book, designed experiments do often
take the time and resources of someone besides the Six Sigma expert, which means there is
a larger cost associated with DOE than there might be with graphical analysis or even
ed
objective a detailed understanding of why the experiment is necessary and what the team
hopes to achieve via the analysis.

The objective is directly tied to the practical problem. In the healthcare example, the
problem was:

In a 10-point evaluation for sanitation compliance, clinical healthcare workers are averaging
a score of 7.1. The hospital requires an average of at least 9.0.

An objective might be written as:

Increase sanitation compliance scores from 7.1 to at least 9.0.

That increase would meet the requirements from the leadership team and, since the Six
Sigma team believes sanitation is related to staph infections, would hopefully improve the
overall problem. Understanding the amount of change the team wants to create in the
process will be important in determining the sample size required for the DOE. The
objective should be specific enough that the team can use it later to measure success and
involve the scale of the change for sample size calculations.

tive of any given experiment within a Six Sigma project

is not going to be the same as the overall project initiative. Instead, each experiment
objective should be related to or supportive of the project objective. In this case, the
team hopes that increasing sanitation compliance will help decrease staph infection rates.

Step 3: What is the appropriate Y for your experiment?

601
 DESIGN OF EXPERIMENTS

For a designed experiment, you almost always select the output response, or Y, first. The
response needs to be related to the goals of the overall project; because there might be
multiple output responses related to the project and
team should ask some questions about the response and possible associated inputs.

First, is the measurement system capable enough to support the experiment? You
can test the measurement system with a gage analysis. The less capable a
measurement system is, the bigger the sample size and more experimental runs you
or, then the time or
expense for an experiment can be prohibitive.
Second, what is important for the team to address? Does the team want to address
variation within a process or address the center of the process? Is the purpose to
bring variation down or to move the mean of the process up or down? In the case
of the healthcare scenario provided above, the purpose is to increase the rate of
compliance.
The team must also understand how the response might relate to other responses
in the process or whether changes to one response will cause changes to another
particularly if an experiment is being conducted within the live environment of a
process. This helps the team manage any downstream changes that could impact
other processes or the customer.

The rate of compliance with the sanitation protocol is the Y for the Six Sigma healthcare

Step 4: What inputs will be studied?

In the DMAIC chapters, we covered some brainstorming tools for identifying inputs within a
process, including the fishbone diagram. More brainstorming tools are covered in chapter
31. Teams can use such tools, along with process maps, to identify the inputs associated
with the output decided upon in the previous step. Once you have a list of inputs, you can
select the inputs that are going to be studied in the experiment. Note that not all the inputs
to a process will be studied, and teams should choose inputs that are:

Believed to be most influential on the response (consider the Pareto, or 80/20, rule)
Are quantitative or qu
to sample size requirements, and involving too many people in the measurement
system can also add variation, so keep these considerations in mind)
Can be set and controlled appropriately for the experiment

602
 SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

For the healthcare scenario, the team

believes that several factors might Response, Factor, and Level
contribute to a lack of compliance:
education of the staff, the length of time a When talking about design of experiments,
particular staff member has been on shift, Six Sigma experts might use slightly
and the number of patients on the floor at different verbiage to discuss process
a given time. components. The response is the outcome
or output of the process. Factors are the
Step 5: What are the factor levels for the inputs usually those that have influence
experiment? on the response and are being studied
during the DOE. Levels are the values at
What levels will the input variables be set
which the factors are set for the purpose of
at for the purpose of the experiment? If
experiment.
you are running an experiment that
involves baking cookies, then oven
temperature is an input, or factor. Levels for the factor might be 325 degrees F, 350 degrees
F, and 375 degrees F. Remember that in the case of a 2k factorial, each factor only has two
levels.

In the healthcare example, the team decides to set each factor at a low and a high:

Staff education: Received annual compliance update training in the last 6 months
versus has not received the training in the last six months
Time on shift: Less than or equal to 4 hours versus 5 or more hours
Number of patients on the floor: Below 80 percent of maximum capacity versus
above 80 percent of maximum capacity

The levels set for each factor depend on a variety of considerations. If the team is
performing a screening experiment which means they only want to test input impact on
outputs or better understand the process they would set levels at wide intervals. For
optimization, the team likely already has an understanding of the input impact on the
process; the team might already have conducted analysis or experiments, and that
information helps them set the levels appropriately at smaller intervals.

The process itself will dictate how some levels are set. If a Six Sigma team is working with a

most ovens also have a maximum temperature range. Teams also probably know the
temperature around which baked goods will still be raw, so they are unlikely to set
temperatures at something like 100 degrees F. Safety should always be a concern when

603
 DESIGN OF EXPERIMENTS

setting levels. An oven might heat to 550 degrees F maximum, but is it safe to attempt to
bake cookies at that level?

Finally, the levels should correlate with the type of information the team is looking for and
the size of change the team wants to see. If a team is working with a manufacturing process
that cuts lumber, it likely wants to see changes at small increments. Setting tool levels

with the cutting process for 2-by-4 lumber piece that are to be cut into 6-inch pieces.

Step 6: What is the appropriate sample size?

In the case of the desig

you collect but the number of times you replicate the experiment. The more replications,
the greater the accuracy of the statistics, but each run in the DOE takes time and possibly
other resources. Six Sigma experts have to balance the use of resources with the need for
strong experiment conclusions. When you create a design of experiment in Minitab, the
software program generates some predictions about the experiment even before you run
the experiment, which lets you decide on size and replication before you begin. Specifically,
Minitab tells you the size of effect, as measured in standard deviation, that you are 80
percent likely to capture with the experiment as you have designed it.

To understand more about sample size and replication and how to run a 2k factorial
experiment in Minitab, begin with the Assistant menu. You can also access design of
experiments under Stat > DOE, but the assistant wizard is helpful because it reminds you of
all the steps previously discussed in this chapter, helps you decide on the proper type of
experiment, and helps you create the data collection form and analysis.

For the purpose of setting up this experiment, use the healthcare example discussed above.

Problem: In a 10-point evaluation for sanitation compliance, clinical healthcare workers are
averaging a score of 7.1. The hospital requires an average of at least 9.0.

Objective: Increase compliance scores to at least 9.0.

604
 SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

Response: Rate of compliance with sanitation protocols

Factors: Staff education, time on shift, number of patients on floor

Levels:

Factor Low (-1) High (1)

Staff education Did receive compliance

compliance training in training in last 6 months
last 6 months

Time on shift Less than or equal to 4 5 or more hours

hours

Number of patients on Below 80 percent At or above 80 percent

floor maximum capacity maximum capacity

Select Assistant > DOE > Plan and Create.

process.

605
 DESIGN OF EXPERIMENTS

Minitab provides some instructions on identifying factors, setting levels, and choosing
sample size. This information is all valuable and worth reviewing, particularly if you are
conducting a DOE when it has been a long time since you last did so. It is also all information
that has been included in this chapter.

606
 SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

One good note from the above Minitab me

from any stakeholders, project owners, or sponsors associated with the project or process.
Key stakeholders should understand what will happen during the experiment and why it is
necessary to the project. If resources are required, sponsors or other leaders should sign off
as required by organization guidelines.

In the particular healthcare scenario being used as the example, running a standalone
certainly g to give patients a staph
infection just to find out about inputs. In this particular case, the team would need to use
existing data that matched the experiment parameters. This is something that might be an
option or a requirement for teams in any number of scenarios.

607
 DESIGN OF EXPERIMENTS

Minitab displays a Modeling Design dialogue box, requesting all the information needed to
create an experiment. Enter the information for the healthcare scenario.

1. Name the response variable. The team is measuring compliance with the sanitation
protocols,
can name the variable anything that will make sense to yourself or the Six Sigma
team.
2. Choose an option for the response goal. Minitab lets you select minimize the
response, maximize the response, or achieve a target. In this case, the team wants
to improve compliance to a score of 9, but going above 9 would also be a positive

3. Enter the number of factors. Remember, factors are inputs, and the healthcare
scenario identifies 3 inputs.

608
 SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

4. Enter names and information for the factors. Even though some of the factors are

into categories.
5. Select a number of times to replicate the experiment (remember, this is relevant to
sample size). For this purpose, select 2.
6. Ensure your dialogue box matches the one below and select OK.

609
 DESIGN OF EXPERIMENTS

Minitab offers you the option to print the data form to make it easier to collect data. For
the example, select no.

Minitab generates three items. The first is a modeling design report card that provides next
steps and tips for completing the experiment.

610
 SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

The second item is a summary report. The report summarizes all the information you
entered for the experiment and provides an estimated detection ability, which helps you
choose a sample size.

The final item is a worksheet to help you gather results from the experiment. Minitab
randomizes the order of the runs, which would be required if you were completing the
experiment in real-time.

611
 DESIGN OF EXPERIMENTS

For the healthcare scenario we entered, the detection ability based on 2 replicas will detect
standard deviations of 1.36 or more.

If the team wanted to detect smaller changes, it would need to add more replicates. Four
replications would detect 0.91 standard deviations, and 5 would detect as little as 0.81
standard deviations.

612
 SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

To make the best possible choice about sample size, it does help to know what the current
standard deviation is to understand what measurement is going to be meaningful.
Remember, two replicates have an 80 percent chance of detecting 1.36 standard deviations,
not a 1.36 shift in the actual compliance score.

Before moving forward with the experiment, the team might consider some sample data.
Here are the sanitation compliance scores from all of the clinical audits for one week.

7.8
7.5
7.1
7.5
9.1
5.4
7.2
7.2
5.8
8.7
5.8
7.1
6.7
7.8
5.1
7.7
6.5
8.2
7.4
7.1

Copy the data into Excel or Minitab to calculate the standard deviation of that data set.

The standard deviation is 1.028.

According to the first summary, based on the setting for 2 replicates, the detection ability of
the experiment is only an 80 percent chance at detecting changes of 1.36 standard

613
 DESIGN OF EXPERIMENTS

replicates, the difference is right on the line, and in any other situation, a Six Sigma expert
would probably want t
stay with 2 replicates for the rest of the example.

Minitab provides some important tips for completing a designed experiment in the
checklist.

First, anyone involved in the experiment should be properly trained to remove variation
from operator error. Individuals involved must know how to measure the output, or
response, how to set factor levels if that is required during the experiment, and how to
make changes to equipment or processes between runs to affect the experiment. In the
healthcare example we are using, the only training that might be required is ensuring the
audit which applies the compliance score is handled correctly. In another situation,
however, a team might be running a production process in a manufacturing plant. The team
might want to try different materials, different tools, and different operators as part of the

614
 SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

experiment, and they would have to ensure everyone knew how to use the materials and
tools.

Measurement systems should be validated for both the response and the levels for every
factor. In our healthcare example, the Six Sigma team should have already validated the
measurement system used to rate compliance. They might have used a Gage R & R to
determine if separate auditors with separate replications would rate the same situation the
same way more than once, for example. Teams should also validate the measurements
used to set levels. If the speed of a wheel in a manufacturing process is to be set at 1 rpm,
1.2 rpm, and 1.5 rpm as levels of a factor, then calibrations for those settings should be
checked.

When running actual experiments, Six Sigma teams should review all possible combinations
in the experiment to ensure they are all both safe and possible. Perfectly safe or possible
levels for one factor might become unsafe or impossible when combined with another, for
example. The team might want to perform a few trial runs to work out troubleshooting
issues with the experiment and finalize validations before moving on to the actual
experiment.

Finally, the team runs the designed experiment according to the random worksheet created
by Minitab. In the healthcare scenario, the team reviewed the data for the last month in
random order and recorded the first set of data that fit each experiment scenario. For
example, the first run in the list below required that the staff member being audited had
received compliance training in the last six months, was over the 4th hour into a shift when
the interaction being audited occurred, and was working at a time when the particular floor
was at 80 percent or more of capacity. The first score that met those parameters was 7.9.
The results of the data gathering are included below, and you can copy this table into
Minitab if you want to follow along with the analysis.

615
 DESIGN OF EXPERIMENTS

StdOrde RunOrde CenterP Block StaffE TimeShif Complianc

r r t s d t NoPts e
80% or
8 1 1 1 Yes Over 4 more 8.2
4 or
2 2 1 1 Yes under Below 80% 9.5
80% or
7 3 1 1 No Over 4 more 7.2
4 or
1 4 1 1 No under Below 80% 7.8
4 or 80% or
5 5 1 1 No under more 5.2
3 6 1 1 No Over 4 Below 80% 5.4
4 7 1 1 Yes Over 4 Below 80% 8.5
4 or 80% or
6 8 1 1 Yes under more 9.7
11 9 1 2 No Over 4 Below 80% 7.2
80% or
16 10 1 2 Yes Over 4 more 8.8
4 or
10 11 1 2 Yes under Below 80% 9.2
4 or
9 12 1 2 No under Below 80% 6.3
4 or 80% or
14 13 1 2 Yes under more 9.7
4 or 80% or
13 14 1 2 No under more 8.1
80% or
15 15 1 2 No Over 4 more 5.6
12 16 1 2 Yes Over 4 Below 80% 8.9

To analyze the results in Minitab, select Assistant > DOE > Analyze and Interpret.

616
 SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

experiment.)

Click OK.

Minitab generates five items:

617
 DESIGN OF EXPERIMENTS

A report card
A prediction and optimization report
A diagnostic report
An effects report
A summary report

The report card provides overall information about the analysis and some tips for
understanding the results. In this case, the report card notes that no unusual data points, or
possible outliers, are noted and that there is only one significant factor in the experiment
according to statistical analysis.

The prediction and optimization report compares the means associated with significant
factors and provides an optimized level setting for those factors. In this case, only one factor
whether or not the staff had received compliance training in the past six months was
statistically significant.

618
 SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

The effects report provides more information about all the factors. In the lower part of the
diagram below, you can see the main effects plots for each factor. The main effect is the
difference between the mean response at each factor setting. You can see that only a slight
difference occurs between the means of the responses for different amounts of time spent
on the shift and almost no difference occurs between the means of the two settings for
amount of patients on the floor. The largest difference is seen with regard to the
compliance education status. The same is true in the top part of the graph, where different
factors are graphed together.

619
 DESIGN OF EXPERIMENTS

The summary report provides a Pareto breakdown of the influence on the response. In this
case, the biggest influence on the compliance number is whether the staff received
compliance training in the last six months. The next factor is how long the staff member had
been working that shift. The third most influential factor is a combination of A (staff
training) and B (time worked on the shift). Number of patients on th
appearance until the fourth influence.

The summary notes that setting the staff education factor at Yes the staff member
received training within the last 6 months results in a predicted Y (response) that is above
9. The summary also notes that this model explains 68.64 percent of the variation.

620
 SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

Based on these results, the Six Sigma team might run further tests and analysis, but it would
also likely make a recommendation about supplying staff with compliance training. If you
remember, the training that was mentioned originally was an annual training. However, this
experiment seems to indicate that the training has more impact if staff members have
received it within the past 6 months. The team might recommend that staff receive the
training biannually going forward.

Remember that an experiment is not usually what solves the large problem that the Six
Sigma team is working on. It simply helps approach one part of a solution to that problem.
In the case of the healthcare Six Sigma team, recommending the additional training might
increase compliance rates with the sanitation protocol, which could in turn decrease the

found and not the only solution it will recommend.

621
 DESIGN OF EXPERIMENTS

After drawing conclusions from experiments, Six Sigma teams must take steps to validate
results, implement solutions or work the results into the overall project, and create
controls. These steps will be covered in more detail at the end of the next chapter.

622
In the previous two chapters, you saw how designed experiments could lead to the use of
an ANOVA or a 2k (two-level) factorial. This chapter will cover more information on the use
of factorials, including multi-level factorials, and provide information for setting up designed
experiments.

Chapter 29 briefly touched on the concept of interactions. Remember, one reason design of

analyze how certain factors and levels impact results. Design of experiments also lets you
understand how the combination of certain factors and levels impact results. Because
designed experiments run each factor and level against all other factors and levels, they can
unearth interactions that other types of analysis, which consider only a single factor at a
time, can miss. Understanding interactions helps Six Sigma teams come up with the
strongest possible solution while also mitigating potential issues such as waste or error.

In the sample healthcare experiment run in the previous chapter, there wa

of interaction between the individual factors, but the interaction plots did show a slight
interaction between the education and shift factors. If a staff member had received the
compliance education within the last 6 months, they were more likely to be compliant with
the sanitation procedures. If the staff member, who had received the training in the past 6
months, was also within the first four hours of the shift, the increase in the likelihood of
compliance associated with the training was slightly higher than the same increase in
someone who was over four hours into a shift.

623
 INTERACTIONS, MULTI-LEVEL FACTORIALS, AND CREATING EXPERIMENTS

Interactions are indicated in the interaction plots by lines that are not parallel. The less
parallel the lines are, the greater the interaction. As you can see, most of these lines are
close to parallel, which means that any interaction that does exist here is fairly weak.
Consider the bottom left graph. The blue line indicates measurements taken when the
department was below 80 percent capacity for patients. The red line indicates
measurements taken when the department was above 80 percent capacity for patients.
Regardless of the capacity status, the behavior of the measurements is roughly the same
depending on whether the staff member had received the compliance training within the
past six months or not. This means there is not much of an interaction the setting of one
factor does not seem to impact the way the settings for the other factor changes the result.

To better understand how interactions are measured, you first have to understand the main
effect. Consider a real-world example, again with a 2-level factorial. In this example, a Six
Sigma team within a chain of movie theaters is trying to positively impact the performance
of concession stands. Specifically, the team has been tasked with increasing profits in the
concession stand.

One area the team is concentrating on is the popcorn process, and they have defined waste
as a root cause of the problem in this part of the concession stand. The team believes the
current popcorn-making process is generating waste because so many of the kernels
introduced into the popcorn machine never pop. Those kernels are swept out of the

624
 SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

machine regularly and thrown away, and the team believes the amount of kernels thrown
away each day amounts to at least 10 large popcorn orders.

While that might seem like a small amount of waste, the team calculates that the
concession stand earns a profit of $2.65 for each large popcorn order, which means the
stand is losing approximately $9,600 each year in lost popcorn kernels. If the team can use a
designed experiment to identify a quick, inexpensive way to reduce kernel waste, it might
be worth the effort.

After finding out more about the kernel process, the team decides that the temperature of
the machine at the time the popcorn is introduced is a factor. They also decide that the
number of kernels placed in the machine and the brand of popcorn are factors as well.

The team designs the following experiment:

Row Number Brand Temperature Kernels

1 A 400 degrees F 1 scoop

2 A 400 degrees F 2 scoops

3 A 600 degrees F 1 scoop

4 A 600 degrees F 2 scoops

5 B 400 degrees F 1 scoop

6 B 400 degrees F 2 scoops

7 B 600 degrees F 1 scoop

8 B 600 degrees F 2 scoops

Following good experiment protocol, the team decides on a sample size. Remember, the
sample size for a designed experiment is what tells you how many replications of the overall
experiment to run. In this case, the Six Sigma team decides to run 4 replications. Given the
above design, that means the team is going to record 32 runs. Also following good

information in the same order each time through the experiment.

625
 INTERACTIONS, MULTI-LEVEL FACTORIALS, AND CREATING EXPERIMENTS

The Six Sigma expert on the team also knew that simply recording the amount of kernels

three factors the team was considering. This is because one of the factors the team was
considering was the volume of kernels entering the process. Naturally, if more kernels enter
the process, there are likely to be more waste kernels
experiment to test if that is true. The team really wants to find out if a larger ratio of
unpopped kernels is left when popping one or two scoops of certain popcorn brands at
certain temperatures. The Six Sigma expert ensures the team records both total number of
kernels and the percent, or ratio, of those kernels to the number of kernels that entered the
process.

The information recorded by the team is shown in the table below. Note that the data has
been sorted by brand, then temperature, and then number of scoops for the purpose of
analysis. This would not be the order in which the team ran the experiment since it would
have been randomized.

Row Est. No Kernels

Brand Temperature Kernels Percent Left
Number Kernals Left
1 A 400 degrees F 1 scoop 2300 125 5.43%
9 A 400 degrees F 1 scoop 2300 115 5.00%
17 A 400 degrees F 1 scoop 2300 118 5.13%
25 A 400 degrees F 1 scoop 2300 127 5.52%
2 A 400 degrees F 2 scoops 4600 298 6.48%
10 A 400 degrees F 2 scoops 4600 275 5.98%
18 A 400 degrees F 2 scoops 4600 298 6.48%
26 A 400 degrees F 2 scoops 4600 275 5.98%
3 A 600 degrees F 1 scoop 4600 98 2.13%
11 A 600 degrees F 1 scoop 4600 101 2.20%
19 A 600 degrees F 1 scoop 4600 99 2.15%
27 A 600 degrees F 1 scoop 4600 105 2.28%
4 A 600 degrees F 2 scoops 4600 187 4.07%
12 A 600 degrees F 2 scoops 4600 192 4.17%
20 A 600 degrees F 2 scoops 4600 175 3.80%
28 A 600 degrees F 2 scoops 4600 201 4.37%
5 B 400 degrees F 1 scoop 4600 145 3.15%
13 B 400 degrees F 1 scoop 4600 129 2.80%

626
 SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

21 B 400 degrees F 1 scoop 4600 138 3.00%

29 B 400 degrees F 1 scoop 4600 142 3.09%
6 B 400 degrees F 2 scoops 4600 324 7.04%
14 B 400 degrees F 2 scoops 4600 298 6.48%
22 B 400 degrees F 2 scoops 4600 307 6.67%
30 B 400 degrees F 2 scoops 4600 399 8.67%
7 B 600 degrees F 1 scoop 4600 109 2.37%
15 B 600 degrees F 1 scoop 4600 138 3.00%
23 B 600 degrees F 1 scoop 4600 109 2.37%
31 B 600 degrees F 1 scoop 4600 124 2.70%
8 B 600 degrees F 2 scoops 4600 202 4.39%
16 B 600 degrees F 2 scoops 4600 225 4.89%
24 B 600 degrees F 2 scoops 4600 198 4.30%
32 B 600 degrees F 2 scoops 4600 225 4.89%

The effect in a designed experiment is the measurement of the change in the result based
on changing the levels of factors. The main effect describes the measurement of change in
the result based on changes in a single factor. The main effect of each individual factor can
be roughly calculated by comparing the average of the results at both the low and high end
of the factor.

For example, the average percent left for Brand A is 4.45. The average left for Brand B is
4.36. The difference is 0.09. Immediately, it appears that there is not much of an effect for
kernel-popping performance between the two brands of popcorn. Remember, though, that

difference is worth considering depends on other factors, such as the range of difference
the team is hoping to record and the main effect of other factors in the experiment.

The same calculation can be completed for each of the factors in the experiment. The
average percent of kernels left for runs with a temperature of 400 degrees F is
approximately 5.43; the average percent left for runs with a temperature of 600 degrees F is
3.38. The main effect for this factor is larger than that for brand; the difference between the
averages for 400 degrees and 600 degrees is 2.05.

627
 INTERACTIONS, MULTI-LEVEL FACTORIALS, AND CREATING EXPERIMENTS

The average percent of kernels left when the team started with one scoop was
approximately 3.27. The percent left when starting with two scoops was approximately
5.54; the difference is 2.27.

number of kernels in each scoop before running the popcorn process. Instead, they made an
assumption that a scoop of popcorn kernels held 2,300 kernals. The fact that not every scoop is going
to have exactly 2,300 kernels could impact the outcome of the experiment. Why would a Six Sigma
team make such an assumption if it might impact the experiment?

In chapter 29, the possible expense of designed experiments was mentioned. The time and resources
required to run an experiment play a role in whether a team or Six Sigma expert chooses to use this
tool for analysis. But time and resources can also play a role in how a team designs an experiment
and what assumptions and steps they take in setting factors. In this example, running 64 separate
batches of popcorn would be fairly time-consuming, but it would not likely be too time-consuming to
be prohibitive. It would also not be too expensive with regard to use of resources, especially if the
team handles the runs so that popcorn generated by the runs could be used in concession services.

However, if the team counted every single popcorn kernel that entered the process, two things might
happen. First, the team could spend an inordinate amount of time counting what would ultimately be
over 220,000 kernels of popcorn. Remember, the savings that might occur from this experiment is

investment for the team to spend that much time counting kernels. Second, the team could actually
impact the quality of the kernels with excessive handling that could occur during counting. That could
reduce the efficacy with which the kernels popped, impacting the authenticity of the experiment. Six
Sigma teams and experts should always try to avoid impacting the results of an experiment in ways

Sometimes, you might have to make assumptions when conducting experiments. When doing so, you
should always try to handle measurements and assumptions in keeping with all of the information
provided throughout this book. For example, a Six Sigma expert would never simply decide a scoop
equals 2,300 kernels. Some ways you might arrive at this assumption include:

Counting how many kernels are in several scoops and arriving at an average (remember
that sample size and confidence levels would play a role in how confident you might be in
this average)
Weighing 50 separate kernels to get a mean average weight of one kernel of popcorn, and
then weighing multiple scoops to see how many kernels, based on weight, might be
included (again, sample size and confidence levels would play a role in how you might treat
these assumptions.)

628
 SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

When plotted on a line graph the main effects from the above experiment look something
like the image below. These images are generated in Excel, but Minitab and other statistical
analysis software generate such graphs as part of the overall analysis when running
designed experiments.

but there are major changes based on temperature and scoop. Since the team wants to
reduce the percentage of unpopped kernels, it would seem that they would want to set
temperature at a higher level but use only one scoop of popcorn at a time.

But the main effects g

generate interaction plots, or create your own in Excel. To create interaction plots in Excel,
follow the steps below.

1. Import your data, with headers, into Excel. For this example, you can use the data
grid provided for the popcorn scenario above.
2. Calculate the average of the result for each combination of two factors. In the
popcorn scenario, you would calculate the average of the percent results for three
sets of graph data.
a. Graph 1 (Scoops * Temperature)

629
 INTERACTIONS, MULTI-LEVEL FACTORIALS, AND CREATING EXPERIMENTS

i. All rows where scoops equals 1 and temperature equals 400

ii. All rows where scoops equals 1 and temperature equals 600
iii. All rows where scoops equals 2 and temperature equals 400
iv. All rows where scoops equals 2 and temperature equals 600
b. Graph 2 (Scoops * Brand)
i. All rows where scoops equals 1 and brand equals A
ii. All rows where scoops equals 1 and brand equals B
iii. All rows where scoops equals 2 and brand equals A
iv. All rows where scoops equals 2 and brand equals B
c. Graph 3 (Brand * Temperature)
i. All rows where brand equals A and temperature equals 400
ii. All rows where brand equals A and temperature equals 600
iii. All rows where brand equals B and temperature equals 400
iv. All rows where brand equals B and temperature equals 600
3. Arrange the results into three small data tables that resemble the data table for the
temperature and scoop set below.

Temperature Scoops Average

400 1 4.14
600 1 2.4
400 2 6.72
600 2 4.36

4. Use each of the small data tables to create a line graph.

a. Select Insert > Charts > Line Charts > Multi line graph with markers

b.

630
 SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

c.

d. Select the averages for 1

e. Click OK.
f. Click Add again, and repeat step d to create a series for 2 scoops.
g. Click OK.
h.

631
 INTERACTIONS, MULTI-LEVEL FACTORIALS, AND CREATING EXPERIMENTS

i. Select two cells that hold the names for the other factor settings (in this
case, it is the temperature settings) into the Axis labels box.

j. Click OK.
k. Click OK again.
l. Use Excel chart options to add a title and legends as desired so that your
chart looks something like the image below.

Scoop * Temperature
8
6
4 1 scoop

2 2 scoops

0
400 600

5. Repeat all of the steps in 4 above for each set of paired data elements. The results
of all three are show below.
632
 SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

These are the same style of interaction plots that are displayed by Minitab if you run DOE
analysis in that program, and they show some interaction that was not apparent when
simply considering the raw data or even the main effects plots.

Analysis of the first interaction plot shows what the Six Sigma team might already have
gleaned from the main effects. The lowest leftover kernel percentage seems to come when

an interaction the lines are almost parallel. Regardless of the number of scoops you start
with, the highest temperature yields the least waste. Regardless of the temperature setting,
the lowest scoop start yields the least waste.

The second graph, to the right of the first, shows the interaction between brand and
temperature. Remember, the main effects analysis seemed to indicate that the choice of

633
 INTERACTIONS, MULTI-LEVEL FACTORIALS, AND CREATING EXPERIMENTS

brand of popcorn was not important. When brand is analyzed with another factor, though,
does make a difference
in how the change in temperature setting works. At 600 degrees F, brand A performs better
than brand B. At 400 degrees F, brand B performs better. Even though all of this is true,
both brands generate less waste at the higher temperature setting when compared to their
own performance at the lower temperature setting.

The third graph also shows some interaction between brand and number of scoops. Brand B
performs better at the lower scoop setting than brand A, and the reverse is true at the
higher scoop setting. Again, both brands generate less waste at the lower scoop setting
than when compared to their own performance at the higher scoop setting.

Based on all of the information above, the Six Sigma team is likely to choose to start the
popcorn process with 1 scoop and use the 600 degree F temperature setting. Choosing a
brand of popcorn, however, is a little tougher. Opting for brand A saves approximately 0.5
percent of waste over brand B when considering the interaction with temperature.
However, brand B saves almost 1 percent of waste over brand A when considering the
interaction with number of scoops.

Given all of the information and analysis in this experiment, a Six Sigma team might either
say brand is not a big enough factor (so the concession stand could choose brand based on
other factors, such as cost) or recommend that brand B be purchased when it is equal to or
less than brand A in price because brand B is associated with the highest percentage of
waste savings. If brand B is more expensive than brand A, the small waste savings
associated with brand B would likely be mitigated, though, making brand A the better
choice. This is a good illustration of how other data should play a role in decision making a

This was a basic analysis to help you understand how interactions work. Minitab and other
statistical analysis software runs all of this analysis automatically, reducing the workload on
the Six Sigma expert. Most such software will also consider all possible interactions and
present the factors with the most impact in the analysis. If you have a 2k factorial
experiment with three factors, each with two settings, and those factors are A, B, and C, you
could have the following possible effects on the result:

A (the impact of changing only A)

B (the impact of changing only B)
C (the impact of changing only C)
AB (the interaction of A and B)
634
 SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

AC (the interaction of A and C)

BC (the interaction of B and C)
ABC (the interaction of all three factors)

Adding additional factors makes this analysis increasingly complicated; adding additional
levels at each factor does the same thing.

So far in this unit, all the DOE scenarios used have been two-level (or 2k) factorial
experiments. In each experiment, Six Sigma teams chose to set the levels of each factor at
no more than two typically at a high and a low.

factor levels provide more information for the team? In the popcorn example, why would
the team choose the high and low temperature settings? What happens when the
temperature is set at 500 degrees F? What about 425, 450, 475, etc.? How does the team
know that 400 and 600 are the two most optimal settings?

The truth is that the team know those are the most optimal settings, but adding
multiple levels for one or more factors adds to the number of runs that have to be
completed. One replication of the popcorn experiment previously described required eight
runs. Just adding one more factor for temperature 500 degrees F adds four runs to the
experiment.

635
 INTERACTIONS, MULTI-LEVEL FACTORIALS, AND CREATING EXPERIMENTS

Row Number Brand Temperature Kernels

1 A 400 degrees F 1 scoop

2 A 400 degrees F 2 scoops

3 A 500 degrees F 1 scoop

4 A 500 degrees F 2 scoops

5 A 600 degrees F 1 scoop

6 A 600 degrees F 2 scoops

7 B 400 degrees F 1 scoop

8 B 400 degrees F 2 scoops

9 B 500 degrees F 1 scoop

10 B 500 degrees F 2 scoops

11 B 600 degrees F 1 scoop

12 B 600 degrees F 2 scoops

Now, instead of 32 runs for a designed experiment that includes four replications, the team

and the company are being used. If the team wanted to also test temperature settings of
425, 450, 475, 525, 550, and 575, that would entail 24 additional runs for every replication,
or 96 additional runs for four replications. If the team also decided to add additional levels
to other factors, such as a third brand or an attempt at 1.5 scoops of kernels, the team
could be looking at an experiment with hundreds of runs required to achieve the
appropriate sample size.

For these reasons, 2k factorial experiments are typically the most commonly used among
Six Sigma experts. Minitab lets you run these experiments with up to 15 different factors,
providing a powerful ability to analyze processes. Minitab also lets you run experiments
with more than two factors.

636
 SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

Multi-factor experiments are handled in much the same way 2k factorial experiments are
handled. Obviously, if you are handling analysis manua
factors as much as possible. Minitab removes the need to limit data because of analysis
requirements as long as the team can define a need for including factors and levels in an
experiment and has been given permission and access to use the resources needed to run
an appropriate number of replications for accurate statistical conclusions, then Six Sigma
teams can certainly run multi-level factor experiments.

As with almost any other tool in the Six Sigma arena, ANOVA and design of experiment are
only as strong as the foundation you create each time you run an experiment and
associated analysis. In addition to the instructions and information in this unit, this section
provides some tips for successful experiments and analysis

In chapter 29, defining the Y, or response, of the experiment is one of the first steps in the
procedure for conducting a DOE. Remember, a good response is related to the overall
DMAIC process or the overall need associated with the experiment. The team must know
what it wants to address and whether the response it is considering is a good measurement.

Often, training materials or tips for DOE in Six Sigma provide two pieces of common advice:

Ensure that the Y is continuous for the purpose of later analysis

Take the Y from brainstorming or other work done in previous phases specifically
from Define or Analyze tools such as fishbone diagrams.

outside of the box when it comes to the response for an experiment. Ultimately, the
response should be something that will help you truly understand what you set out to
understand at the beginning of the experiment. In the popcorn experiment used as an
example in this chapter, the team could have chosen any number of responses. Possible
responses from the popcorn popping process might include how long the overall process
took, how many kernels were popped, how many orders were generated by each process,
or customer satisfaction with the taste of the popcorn that was popped. But the team

637
 INTERACTIONS, MULTI-LEVEL FACTORIALS, AND CREATING EXPERIMENTS

performance with this particular experiment. The team was asking questions about waste of
raw materials, and the amount of unpopped kernels let the team know exactly how much
waste was created in the process.

When considering an appropriate Y for your experimen

current metrics a process or department uses to measure success. The concession stand
probably never considered unpopped popcorn before this experiment it likely used
number of sales or sales dollars as a metric for succ
Y that you think will help answer the questions the team is asking, consider conducting a
brand new brainstorming exercise focused solely on coming up with an appropriate result
type.

complex quickly. First, you have to validate all assumptions, select the right method, and
choose an appropriate sample size. Then, you have to run the experiments and record all
the data. Finally, you have to conduct the analysis. At any point in the process, you might
introduce an error or create confusion. Planning and documentation are critical to keeping
this from happening, and they can be equally important when you present the conclusions
of your experiment to others. Some specific areas where documentation is especially
important are highlighted below.

Take notes when working through either the ANOVA steps in chapter 28 or the DOE
steps from chapter 29. You will forget or confuse some number or element of the
analysis at some point and need to refer back to the notes. Consider recording:
o Your reasons for choosing a specific method or test.
o Lists of responses, inputs, factors, and levels with a reason for choosing
each. You might need this information to justify the number of runs in your
experiment.
o The process for choosing sample size and any data or assumptions used in
that process.
Record any outside assumptions made regarding the experiment, the components
of the experiment, or the result. You might need to review these assumptions or
data linked to the assumptions to rule out outliers or understand what might seem
like otherwise unexplainable experiment behavior.
Document any instructions or information provided to any operator within the
experiment. Your sponsor or other decision-makers will likely want to understand

638
 SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

exactly how the experiment was performed before they make decisions based off of
the results.
The outcome of any measurement system analysis, particularly as it relates to the
confidence in experiment outcomes.

In any type of statistical analysis, testing, or DOE, the Six Sigma expert is always mindful of
confidence levels. There is always some chance of an incorrect conclusion the alpha and
beta errors covered in the chapters on hypothesis testing tell us this. More accurate data,
larger sample sizes, and choosing the right test or experiment help reduce the chance of
these errors, but decision-makers do need to know how big of a possibility such an error
might be.

When reporting the outcome of a designed experiment, make sure you take all confidence
levels into account. The confidence interval for the actual experiment is not always the only
relevant consideration. Did you make assumptions based on previous analysis to help you
set up the experiment? What were the confidence intervals associated with those analyses?
Consider the popcorn example again: how sure is the team that each scoop holds an
average of 2,300 kernels? What is the variation of that measurement? The answers to both
of these questions contribute to the accuracy of the experiment analysis.

If enterprise leadership or a Six Sigma team will be basing resource-heavy or resource-

contingent decisions on DOE analysis, then those decision-makers must know how
confident the Six Sigma expert is in what he or she is presenting. You should always do
everything you can statistically to increase confidence such as validating data and
assumptions and choosing proper sample sizes but you should never falsely inflate
confidence in a conclusion to encourage leadership to support a certain decision.

You should not and often cannot do everything you want to do in a single experiment. If
there are 25 possible factors, do you really want them all in a single, unwieldy experiment
that could become costly to run and impossible to control? Often, when you are considering
running an experiment with so many factors, you are actually trying to integrate multiple
experiments into one. Stop and consider the possibility that you are comparing apples and
oranges in an experiment when you really should run separate experiments for various
product types or processes.

639
 INTERACTIONS, MULTI-LEVEL FACTORIALS, AND CREATING EXPERIMENTS

If you are actually working on a single product type or process, consider that you might
need to run experiments serially. You might need to find out a certain piece of information
or narrow down interactions and factors to better set up other experiments with new

reduce waste. There are other ways to incr

the waste-related experiment, the team might conduct a second experiment to find out
which brand customers like more and whether increased customer satisfaction drives
increased sales.

Designed experiments are one of the more challenging aspects of Six Sigma to learn

of ways, which increases the flexibility of the tool but can also make it difficult to know
exactly how to proceed. This is an area where experience pays off, so the more experiments

good idea to work with a more experienced Six Sigma expert or a Master Black Belt, who
can act as a sounding board for your ideas and guide you in creating the best possible
experiment.

640
Throughout this book, , including
brainstorming tools
the in-and-out-of-the-box method, and the 5 Whys brainstorming model. This chapter
provides an appendix of additional tools and methods that can be integrated into meetings
for DMAIC projects or simply to help a team or department gain insight into a problem or
process.

As a Six Sigma expert or process leader, having access to multiple process improvement and
brainstorming tools is helpful. While all of the tools previously covered in this book are
extremely powerful when used correctly, if you use the same approach all the time, you risk
coming up with the same solutions and hitting the same obstacles. Sometimes, switching
the brainstorming method or tool can help break a team out of a rut, encouraging a
different perspective or sparking different discussions or debates.

An activity network diagram helps a Six Sigma team chart either the activity involved in a
project or the activity involved in a process. The point of the diagram is to understand
sequential and time relationships within a project or process. Which items come first?
Which activities happen at the same time?

When applied to an overall project, the activity network diagram helps teams and project
leaders understand what resources will be required at each point during the project. It can
also provide project managers with the ability to better estimate project timelines.

641
 BRAINSTORMING AND PROCESS IMPROVEMENT TOOLS

When applied to a process, the diagram helps teams identify all the activities in a process. It
also helps teams understand which activities are up- and downstream and what activities
happen together. Once teams know these things, they can begin to discuss whether the
current sequence is the most effective sequence and understand how changes to one
activity might impact the outcome of another.

An activity network diagram is created by following these steps:

1. List all of the activities involved in the process. Teams can draw from their own
knowledge or from previous tools used by the group such as SIPOC or process maps.
2. Write each activity on a separate card or sticky note.
3. Place activities in order from left to right. Activities that are simultaneous are placed
in a vertical line in one horizontal space on the diagram.
4. Write an estimated time for each activity.
5. Draw a critical path line through each step. Where vertical (simultaneous steps) are
noted, the line goes through the step that takes the longest.
6. Add up the times listed along the critical path to get an idea of the overall estimate
of time for the process. In the example below, the total time for the critical path in
minutes is: 5 + 2 + 12 + 9 + 10 + 5, or 43.

An affinity diagram is a tool that lets the team list and group a large number of ideas quickly
and systematically. Affinity diagrams work well when you are first approaching a process or
idea because they help you sort through all of the associated ideas to identify some of the
642
 SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

main points the team will need to consider throughout the DMAIC project. Affinity diagrams
are also a good tool to use if you have a larger-than-normal brainstorming group, which
means they can be used in department meetings or in DMAIC meetings that include subject
nity diagram is to help
the group understand the relationships between inputs, outputs, products, services,
workers, customers, processes, and ideas or any subgroup of those things.

flat table, wall, or

floor. You can also use a white board, and if you are working with a group across multiple
locations, you can conduct the exercise via web conferencing software, though, as with
most brainstorming tools, this would not be optimal.

An affinity diagram session is conducted by following the steps outlined below.

1. Provide everyone in the session with small notecards, sticky notes, or pieces of
paper.
2. Begin brainstorming ideas about the process or problem. Ideas at this stage are very
loose encourage the team to write down anything that comes to mind. The ideas
should be short one or two words or very simple sentences, otherwise the team
might be recording two ideas on one note card. Encourage participants to record
inputs, outputs, challenges, obstacles, customers, people, processes, actions, etc. If
you are attempting to address a specific question, that question should drive the
type of things recorded on the notes.
3. Brainstorming can occur individually for a few minutes or as a group. Individual

other input, which can be valuable.

4. After brainstorming, have everyone spread the cards on a floor or table, or stick the
notes to a wall or white board.
5. Work as a team to identify ideas that are related. Place those ideas next to each
other. Consider working as a team without talking during this step to shift ideas
around the diagram for a few minutes until most ideas are lumped into larger
groups.
6. If one idea could belong to more than one group, make multiple cards with that
idea on it so it can be put with multiple groups.
7.
8. Once all the notes are grouped, hold a discussion time with the team to answer
questions:
a. Are there patterns in the ideas?
643
 BRAINSTORMING AND PROCESS IMPROVEMENT TOOLS

b. Does the diagram spark any new ideas or clarify ideas or relationships?
c. Do ideas fall into natural sequences for processes or timelines?
d. Do ideas fall into natural groups of causes?
9. The questions asked and answered will depend on what process or overall question
the team is addressing with the affinity diagram.
10. Title each group of ideas by pulling out what seems to be the main idea of that
grouping.
11. Record the diagram for use later. You can have someone copy it onto a piece of
paper, convert it to a computer diagram, or simply snap a digital picture for storing
with other team documents.

An interrelationship diagram is similar to a fishbone diagram in purpose. The goal of an

interrelationship brainstorm is to understand what causes and effects are primary with
regard to a problem or process. The outcome of the brainstorming session is a visual
depiction that makes it easy to see which items are likely causing most of the effects and
which effects are tied to which causes within a process.

relat
factors might be primary within the process, or are just beginning to understand the process
and need more overall knowledge before moving into a fishbone diagram process or
measuring processes.

Conduct an interrelationship diagram session by following these steps:

1. Define the problem on a white board, piece of paper, or sticky note.

2. On other sticky notes, write brainstormed ideas about the problem. Specifically,
what might have caused the problem? What might be causing things that cause the
problems? What other issues might the problem cause?
3. Place the notes all around the problem on the white board.
4. As a team, consider each idea on each note. Decide whether that idea is somehow
related to any other idea on the board.
5. If the idea is related in such a way to a second idea that it could cause, or be an
input or factor, for another idea, draw an arrow going from the first idea to the one
it might be an input, factor, or cause for.

644
 SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

6. If the idea is related in such a way to a second idea that it might be an effect or
output from the second idea, draw an arrow going from the second idea to the first
idea.
7. If necessary, draw multiple arrows to and from each idea.
8. At the end of the exercise, count the number of arrows coming in to the idea. Write
those numbers in one color on each sticky note.
9. Count the number of arrows going out from each idea. Write those numbers in
another color on each sticky note.
10. The notes with the highest total count when you add those two numbers are
likely the key factors in the process or problem. You might also be able to identify
key inputs (causes) and outputs (effects) based on the number of arrows coming in
or out of various notes.

work if Six Sigma leaders are too caught up in the numbers and data to realize that people
play an important role in every process, every project, and every improvement initiative.
One way Six Sigma teams can understand how people and other factors might support or
challenge a process, project, or change is to conduct a force field analysis.

A force field analysis lets teams understand how factors surrounding a process might
support (or drive) change or challenge (or restrain) change. In addition to people, factors
might include culture, materials and resources requirements, customers, needs, machines,
rules, and regulations. Sometimes, Six Sigma teams can work to change some of these
factors to create additional support for an improvement. In other cases, the team cannot
change restraining factors, but must be aware of those factors as they move forward with
improvement initiatives.

Create a force field analysis by following these steps:

1. Start with the definition of the problem. This can be either the overall problem
statement or a more detailed and smaller problem within the scope of the entire
project.
2. Write the problem on the top of a white board or large piece of paper.
3. Draw a bold line down the center of the page. This line represents the status quo
the current state of the process.

645
 BRAINSTORMING AND PROCESS IMPROVEMENT TOOLS

4. Define the solution or desired state. If the team is just starting out, the desired state
might be the overall project objective statement. If the team is working to solve a
specific problem within the project scope, it might have a solution in mind.
5. Write the solution or objective statement at the bottom of the white board or page.
6. Brainstorm driving factors
7. Write driving factors on the left side of the status quo line.
8. Brainstorm restraining factors factors that would keep the status quo as is.
9. Write restraining factors on the right side of the status quo line.
10. Draw an arrow under each driving factor toward the status quo line; the length of
each arrow represents the perceived strength of the driving factor. For example,
the sponsor who holds financial decisions is on board with the solution. This would
be a strong driving factor and might get a long arrow.
11. Draw an arrow under each restraining factor back toward the status quo line. The
length of each arrow represents the perceived strength of the challenge. If the
sponsor on board with the solution yet, this would be a fairly strong
restraining factor.

Once the diagram is completed, the team can see quickly:

Whether there are more driving factors or restraining factors

Whether driving factors
number of longer arrows on either side)
Which factors are strongest

The team can then discuss whether any of the restraining factors can be weakened or
converted to driving factors, which can be helpful in paving the way for smoother solution
implementation.

A responsibility chart helps Six Sigma teams identify which people, teams, or departments
have various types of connections or responsibilities to a process. The reason for
understanding such relationships is for Six Sigma teams to understand who actually acts on
or within the process, who needs to be informed about the process, who is ultimately
accountable for the process, and who might offer valuable information about the process.
Often, these various people can be identified with a charting process known as RACI, which
stands for:

646
 SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

Responsible the person or team who actually does the work (an employee who is
responsible for collecting money at the cash register)
Accountable the person in charge of the work (the supervisor of that shift)
Consult any person involved in steps prior to the activity that should be kept in
the loop (the sales person on the floor, the bookkeeper who creates starting
drawers each morning)
Inform any person involved in steps after the activity that should be kept in the
loop (the bookkeeper, who also balances the drawers at the end of shifts)

Create a RACI matrix for your project or process by following these steps:

1. Create a list of all activities and decisions that are part of the process. Teams can get
this information from previous brainstorming sessions or tools such as process
maps.
2. List those items down the right side of a page or white board.
3. Create a list of all functional roles associated with a process. List those as the row
headers across the top of the page or white board.
a. Activities and decisions might include items such as:
i. Measure lumber.
ii. Cut lumber
iii. Review lumber.
iv. Decide if lumber is the right length.
b. Functional roles might include:
i. Lumber measurer
ii. Lumber cutter
iii. Auditor
iv. Lumber packer
4. For each functional role and activity, decide if the role is responsible or accountable
for each activity listed or if the role should be consulted or informed. In the example
below, you can see that the cutter is responsible for measuring and cutting, the
auditor is responsible for reviewing and deciding whether the lumber is correct, and
the supervisor is accountable for all of it. The packer needs to be informed about
whether the lumber is correct so he or she knows whether to pack it.

647
 BRAINSTORMING AND PROCESS IMPROVEMENT TOOLS

Cutter Auditor Packer Supervisor

Measure lumber R A
Cut lumber R A
Review lumber R A
Decide if lumber is
correct R I A

A Six Sigma leader has to develop a good understanding of group dynamics. Not every
project team will work the same, and sometimes Six Sigma leaders face interpersonal
challenges within the group environment. Some of these challenges can derail
brainstorming efforts because of personality clashes. If you have one or more quiet or shy
individuals, for example, you might not be getting all of their perspective. Even if everyone
on the team is fairly vocal, if there is a natural leader, bully, or more outspoken member,
they might be getting more ideas in than others. On way you can reduce the impact of
personality on brainstorming sessions is to use the nominal group technique to encourage
input from everyone.

1. Provide everyone with pen and paper.

2. Being with a question, topic, or concern that is relevant to the project or the phase
you are in.
3. Take 5 to 10 minutes and invite everyone to brainstorm their own ideas. Make sure
you note that at this stage, there are no bad or wrong ideas.
4. Once everyone is finished writing, go around the table, having each person state an
idea from their list. Each person should only state one idea at a time before letting
others have a turn.
5. uss the ideas at
this time.
6. As ideas are shared, individuals can pass if every idea on their list has been shared
by themselves or someone else.
7. Once all ideas are listed, lead a discussion based on the ideas or use the ideas to
fuel a secondary activity, such as an affinity diagram or fishbone diagram.

648
 SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

Nominal group technique is also a good tool to use if the team is stuck or is dealing with a
particularly difficult concept. The technique helps break thoughts out of ruts in two ways.
First, it removes some of the fear individuals can have at sharing half-formed verbal ideas.
Team members who are writing ideas down can think through those ideas a bit more and
might feel more comfortable listing them later. Second, the act of writing itself is proven to
break down creative and thought barriers, and can help the team members brainstorm
more ideas.

Check sheets
mentioning them again in a more formal way. Check sheets are simple tools that can be
extremely powerful for managing teams, people, processes, and data. The reason check
sheets are so powerful is that they require an action, and when you require a physical and
recordable action, personal responsibility for the action involved is usually higher.

A check sheet is any templated form used to record information. The information you
record can be a specific piece of data, but it might also be a notation that a routine check,
analysis, or certain other task was performed. The purpose of a check sheet, beyond
formalizing and mandating an action, is to streamline the process of reporting the action.

Consider the check sheets often seen in public restroom facilities in restaurants or stores.
These check sheets are usually on the door or wall near the door. Each hour, a staff member
initials and dates the check sheet to indicate that the restroom was checked. By requiring
the staff member to initial and date the check sheet, companies make it more likely staff
will check the restroom at appropriate times. The initial and date indicates that staff
ensured the facilities were stocked, clean, and in working order.

This premise can be used in a variety of processes to ensure data collection or certain tasks
are performed at the right time by the right person. Check sheets can also be used to
manage tasks within a project environment a Six Sigma leader might provide check sheets
to each subject matter expert during a project meeting so that each person knows what
needs to be done before the next meeting and has a physical requirement for recording
process.

As a S
need, and the checklist is often that solution.

649
 BRAINSTORMING AND PROCESS IMPROVEMENT TOOLS

SWOT analysis is a common business brainstorming technique that helps teams understand
high-level strengths, weaknesses, opportunities, and threats (challenges) for a project,
improvement, or process. SWOT analyses can be conducted anytime during the DMAIC
phases, but are especially helpful during define phases or anytime storming has occurred
and the team needs something fast and easy to help them get back on track.

Some people break up SWOT analysis into a visual depiction, drawing a large grid of four
squares, one each to represent Strengths, Weaknesses, Opportunities, and Threats. This can
be helpful if you have a project group that is especially visual, but is not necessary. The
point of a SWOT analysis is really to answer some fundamental questions through
brainstorming.

1. Strengths: What about the process works? What is done better, faster, cheaper,
more unique?
2.
errors, or customer dissatisfaction points?
3. Opportunities: What can be fixed? What are some easy-to-fix issues? Where are
improvements needed?
4. Threats: What obstacles exist to improvement? What risks will the project team
face? What resource challenges might there be?

Starburst brainstorming is a method for generating a series of questions about a problem,

process, or idea. Those questions can later be used in defining processes, problems,
products, or other ideas. On the surface, the starburst tool is similar to the 5 Whys, but
instead of answering each question and then asking another question, the team simply
comes up with as many various questions as possible.

Conduct a starburst brainstorming session by following these steps:

1. Draw a six-pointed star on a white board or large piece of paper.

2. On each of the points, write one of the following: Who, What, Where, When, Why,
How.
3. At the center of the star, write one or a few words describing the central idea.
4. For each point of the star, brainstorm questions about the center that start with the
word at that point. For example, a team might be working to create more customer

650
 SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

satisfaction regarding a food-service product. For the purpose of this example,

imagine a team working to solve a problem of customer satisfaction with an
individually-wrapped snack cake that is typically sold in convenience stores and
newsstands.
a.
b.
i. Who is the primary customer for the snack cake?
ii. Who is most likely to eat the snack cake?
iii. Who will buy the snack cake? (not always the same person who
eats it)
iv. Who will sell the snack cake?
v. Who will market the snack cake?
vi. Who will make the snack cake?
c.
i. What flavor or flavors should the snack cake be?
ii. What packaging is best for the snack cake?
iii. What ingredients are important to the snack cake?
d. Go through all the points of the star, asking questions that match the word
at that point. Generate as many questions as you can within a set period of

point of the star up to five minutes is usually good to keep questions

from become too detailed or out of scope for the problem, process, or idea
you are dealing with.

The purpose of the starburst session is not to answer the questions you come up with, and
that can be a challenge for Six Sigma leaders. Team members will naturally want to move
immediately from one question to an answer, which can derail the process and result in a
discussion that is not productive or ranges out of scope. Direct team members back to
asking questions for the duration of the brainstorming session. Those questions can later be
compiled and used when conducting other exercises, such as SIPOCs or process maps. The

One of the biggest challenges with brainstorming as a Six Sigma team is that, after a few
meetings or as the team gets deeper into the DMAIC process, the team can become stuck

651
 BRAINSTORMING AND PROCESS IMPROVEMENT TOOLS

from them. Teams might find themselves circling back around to the same ideas repeatedly.
Sometimes, this means the team is on the right track and the iteration of the same concepts
throughout the DMAIC process and brainstorming is confirmation that those are the right
things to be thinking about. Other times, however, it can be a sign that the team is stuck.

It can be difficult at times to tell the difference, and Six Sigma experts do have to rely on
analysis and data to help make a decision about whether the team is going in the right
direction or not. If a Six Sigma expert believes that the team is not going in the right
direction, he or she can try figuring brainstorming to try to get team members to think more
creatively or from a different perspective about the issue.

Figuring brainstorming, or role-play, can be combined with many of the other tools in this
chapter to generate new thoughts. Choose one of the simpler brainstorming exercises or
simply ask a question and brainstorm responses. However, ask participants to brainstorm as
someone other than themselves. They might brainstorm as the CEO of the company. They
might brainstorm as a person working on the product line. The might brainstorm as Johnny
Depp, the President of the United States, or Cookie Monster.

Here is where Six Sigma leadership is important. Is the group likely to respond better to a

respond to a serious role-playing situation? This helps you decide whether to go with Cookie
Monster or CEO.

One challenge with this type of team activity is balancing the stress release and creativity of
the silly activity with a true need to be productive. If things are devolving into pure fun
without productivity, then the Six Sigma team leader should wrap up the activity and move
on with another. The time is usually not wasted, though breaking away from the
seriousness of the project for a few minutes is often itself enough to help teams think more
creatively.

Brainwriting is somewhat like the nominal group technique. In fact, nominal group
technique is actually a form of brainwriting. Reasons you might incorporate this type of
brainstorming in your meetings include increasing input from quieter or more reserved
team members, reducing how only one or two vocal members impact the outcome of a
meeting, and generating ideas that other brainstorming methods are not generating.
Remember, the physical act of writing itself contributes to thought processes and can help

652
 SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

the brain make connections between ideas that it is less likely to make during verbal
brainstorming.

There are numerous types of brainwriting exercises that you can use in a team
environment. Nominal group technique is one, and this section provides some other
options.

The pool method can help generate different ideas because participants build off of each

gets a piece of paper and a pen. You can print the pages with a question or idea at the top
or state the prompt for the session once everyone is ready. Each person writes two to three
ideas on their paper and then places the paper back in the center of the table. Each person
takes another paper from the pool (one that already has ideas written on it from someone
else) and adds two to three more ideas. Papers are written on, replaced, and repulled
each participant moves at his or her own pace and
pulls from the center pool as pages become available.

At the end of the exercise, which lasts a predetermined time, the team records all ideas on
a single page or white board and categorizes and sorts them appropriately. The length of
time the session might last depends on the number of people involved and how complex
the problem is. A good time limit for this exercise is typically between 10 and 30 minutes,
and the exercise works best with four to eight participants.

Often, the goal of brainwriting ich is accomplished by

returning and pulling pages from the pool in the above method. The card method uses a
similar approach but records each idea on a separate card for more flexibility during the
analysis or categorization portion of the exercise.

The card method works well with groups of four to eight people. Begin by providing each
participant with a stack of index cards and a pen. Pose a question or problem it sometimes
helps to write this question or problem on a whiteboard or large piece of paper so
participants can reference it throughout the exercise.

Participants start by brainstorming their own ideas by writing each idea on one of the cards.
After an idea is written down, the card containing that idea is passed to the left. When a

653
 BRAINSTORMING AND PROCESS IMPROVEMENT TOOLS

participant has written as many ideas as he or she can think of quickly on cards, they begin
picking up and reading the cards passed from the person on their right.

If a participant can add anything to the idea on any card they read, they write the new or
extended idea on a new card and attach it to the existing card. Those cards are then passed

left.

When a participant receives his or her own cards back, then they keep them. The
brainwriting session ends when all participants have received all of their original cards back,
hopefully now with attachments of other ideas. The moderator then picks up the cards and
the cards can be used in further brainstorming and idea organization. You could use the
cards to create an affinity diagram, for example.

The card brainwriting method is also a good way to start a group effort to create a process
map. Instead of writing ideas for solutions to a problem, the team writes down individual
parts of a process including activities, inputs, and outputs. Those pieces can then form the
bare bones of a process map, which can be added on to by the team.

This chapter, and most of the information throughout this book, concentrates on activities
and methods to be used in a group environment. This is because facing and solving
problems as a team is a fundamental part of the Six Sigma methodology. Brainstorming as a
team is critical because it presents multiple perspectives and lets individual team members

even a good idea to brainstorm about every single issue as a group.

Sometimes, the Six Sigma leader must consider an issue or idea before presenting it to the
group. Other times, the issue about which you want to brainstorm isn't something that is
appropriate to share with the entire team. Remember, if you are in a Black Belt or project
leadership role, you might have access to information that is sensitive or confidential. While
Six Sigma projects and process improvement work best in transparent environments, it's
simply not always possible for every member of a Six Sigma team to know all of the
information about a process. As a Six Sigma leader, you should work to create a team that
you feel is capable of dealing appropriately with sensitive information, but you should also
be prepared to facilitate a team in unique situations when some information sharing is
limited. This can be especially true in industries such as healthcare or finance, where the
data itself is sensitive.

654
 SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

Other reasons exist for brainstorming on your own as a project leader or Six Sigma expert.
Your team might be floundering due to storming or scope creep, two challenges covered in
earlier units. You might want to come into a team meeting prepared with some ideas or
understanding of the problem -- although you should always be careful not to lead others
too closely with your own ideas because you might reduce the efficacy of the team. As a Six
Sigma expert, you might also face your own problems or information needs outside of a
team environment.

Whatever the reason, you can use abbreviated forms of many of the brainstorming tools in
this chapter to brainstorm alone. Tools such as starburst brainstorming, SWOT analysis, and
force field analysis can be conducted as an individual. For example, if your team seems to
be stuck in a certain DMAIC phase, you can use the force field analysis to help you identify
what problems or people are holding you in place. As a Six Sigma leader, you can then use
that information to address the issues so the team can move forward with a project.

Brainstorming is a powerful tool whether you use it as a team or as an individual. It takes

practice to be able to best implement brainstorming methods within various group
dynamics, but when you follow the basic guidelines in this chapter, you almost always have
at least some positive result from a brainstorming session.

655
Process maps are a vital part of many Six Sigma projects. Process maps are valuable tools in
all stages of a DMAIC project. While good process maps can take a long time to create, they
can be reused throughout the project and after the project for a variety of purposes.

Process maps are not just simple flow charts, although teams might choose to create a flow
chart as the process map for a specific project or phase. Process maps are visual depictions
of a process and usually include all of the activities and decisions within the process. They
can show various levels of process, and we'll look at the difference between high-level
organizational process maps and detailed process maps in this chapter.

Teams create process maps for a variety of reasons. First, a process map is a good place to
start when teams are trying to understand the overall process. If a visual depiction of the
process doesn't already exist, the team can work with subject matter experts to create such
a depiction. The act of creating the map is actually valuable in itself because it lets the team
ask many questions about the process and explore the process deeply in order to convert
knowledge gained into a working diagram.

During the Measure phase, process maps can help teams understand what inputs or
outputs should be measured, and when and where during a process data can or should be
acquired. Process maps are also helpful during the Analyze phase, because visual depictions
of the flow of a process often make it easier to see where rework is occurring or where
bottlenecks are possible.

Process maps also let the team visualize a future state for a process. A Six Sigma team might
begin with a current-state process map during the Define phase of a project and create a
desired-state map in the Analyze or Improve phase. Side-by-side comparisons of the
current- and future-state maps let teams see exactly what changes need to be made,
ensure that all current-state functionality is addressed in the updated process, and double
check that all challenges are solved by the new process. Process maps can also help teams
see whether they might be creating new challenges, such as rework loops or bottlenecks
within an updated process.

656
 SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

Finally, process maps in their various forms are valuable documentation for a Six Sigma
team and for an organization as a whole. Six Sigma teams can use all or part of process
maps they create in tollgate and other presentations to ensure decision makers, sponsors,
and auxiliary staff understand the process before moving forward with any aspect of the
project or with making decisions about the project or process. Organizations can use
process maps in training documents or as a guide for non-Six Sigma staff who need to
understand new processes or changes to existing processes.

This chapter will cover two basic types of process maps. The first is the flow chart, which
presents a view of the process that typically flows down the page. Flow charts usually give
the overview of a process or detail a specific part of the process. When multiple flow charts
are used for detailed pieces of a process, they can reference each other to provide an entire
description of a process.

The second type of process map covered in this chapter is the swimlane process map, which
usually provides a left-to-right and top-to-bottom visual of the process. A swimlane process
map breaks up the process into different sectors, often by who is responsible for the work,
when work occurs, or what type of work is being performed. At a glance, teams can not only
see the flow of the work through the process, but can see who or what is responsible for
each step.

Many people attempt to create process maps using various shapes, connecting them with
arrows to show the flow of work. That is the basic idea of a flow chart or process map, but
each type of shape actually says something specific about the information presented inside
or near it. Anyone who is familiar with process map symbology should be able to look at a
process map and discern numerous details just from the shapes and the way they are
presented.
and what they are used for.

657
 PROCESS MAPS

Process Shape

The process shape is the most common shape in most process maps or flow charts. It is a
rectangle. Depending on how you plan
design specifications are, you can use default colors of your Word, Excel, or Visio program,
update shape colors to match branding or publication requirements for your enterprise, or
create a simple black-and-white rectangle. This is true for all of the shapes in this chapter.

A process shape is used to denote an action step within the process. Something is
happening, and usually it is a person or people doing the action. The description of the
action is usually typed as text within the shape, which is also true for most of the shapes
presented in this section.

Alternate Process Shape

The alternate process shape is used to indicate that an action is an alternate to something
that is normal within the process. It is a rectangle with curved corners.

Predefined Process Shape

A predefined process shape is used when there is a process within a process and that
subprocess is already defined in another document. The shape lets readers know that more
than one step is occurring at this point in the overall process, and that those steps are

658
 SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

formerly documented somewhere else in the organization. The documentation might be via
standard operating procedures, but could also be in the format of another process map.

If a subprocess is known and a Six Sigma team decides not to fully document those steps
then
this shape should not be used.

Terminator Shape

When viewing a map for a complex process, it can be difficult to tell where the process
begins and ends. This is especially true when a printed process map spans multiple pages
and might have more than one possible end point. The terminator shape is used to show
the beginning and endings of a process. At the beginning, the terminator shape usually
includes an action that triggers the rest of the process.

Preparation Shape

The preparation shape, which is a hexagon, indicates that a step in the process is in
preparation for something else. An example might be cleaning a piece of equipment before
using it or setting the oven temperature before preparing a recipe.

Manual Operation Shape

659
 PROCESS MAPS

The manual operation shape indicates a part of the process that is performed by a person
and is not automated. In processes that are mostly human-
have to differentiate between manual operation steps and regular process steps. The
manual operation step shape is actually more important to use when depicting mostly
automated or computerized processes because you would want the sudden need for
human interaction within the process to stand out.

Delay Shape

The delay shape lets you indicate that a waiting time occurs in the process. Delay shapes
can be valuable to Six Sigma teams because they often indicate a possible point where
muda is occurring. When teams see delay shapes on process maps, they should ask
questions about why the delay is occurring is it, for example, because one step in the
process is batching outputs, so the next step is awaiting a batch? Is it because a bottleneck
is occurring upstream in the process? These are valuable insights for a Six Sigma team.

Decision Shape

The decision shape indicates that a decision is being made in the process. Typically, the
decision indicated by the diamond shape is a yes/no or go/no go variety. One example of
this type of decision branch in a process is seen when further approval is needed for certain
types of actions. For example, a team that processes customer refunds might have some
leeway to process lower dollar refunds without approvals. If the refund is under $50, the

660
 SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

staff request is processed without further action; if the refund is over $50, a supervisor
might need to be involved.

Connector Shape

A connector shape usually indicates a jump from one part of a process to another.
Sometimes, this shape is used to indicate that an audit or inspection occurs at a certain part
of the process.

Document Shape

The document shape indicates that a step in a process creates a document. Using the
refund example above, at some point in that process, a check is printed. A check is a
document, and that step in the process could use this shape.

661
 PROCESS MAPS

Multi-Document Shape

The multi-document shape indicates a step in the process that creates multiple documents.
ing department batches the
checks, printing all refunds at one time each day or week. That step in the process could use
this shape.

Display Shape

The display shape is used to show that a step in a process involves the display of some
information to a person. For example, a person using a company communication portal to
send email might have to log into the portal first. During that process, the software might
display a user agreement; if the process for using the software were mapped, this shape
could be used to indicate that step.

Data Shape

662
 SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

The data shape is used to show that inputs are going into a process or that outputs are
coming from a process. It is also called and In/Out (or I/O) shape.

Manual Input Shape

The manual input shape is used to show a step that involves manual entry of information by
a person, usually using a computer.

Off-Page Connector Shape

The off-page connector shape is used to show that part of the process is depicted on
another page. You can also use off-page connectors to link multiple processes that are
documented on various pages.

Or Shape

663
 PROCESS MAPS

The or shape is different from the basic diamond decision shape depicted earlier in this list.
The or shape also indicates a branch of the process, usually in more than two possible
directions. Typically, the process map provides some data for how to decide which branch
the flow should take at the or shape. For example, consider the refund example used above.

In the previous scenario, there were two options for each refund request: either the request
was under $50 and could be processed or over $50 and had to be worked by a supervisor. In
a new version of the process, the work can take one of three paths. First, if the refund
request is $20 or under, it is approved automatically when entered by a staff member and is
routed to check generation. If the refund request is $20 to $50, it must be reviewed by a
second staff member, who must agree with the refund decision before it is routed to the
check process. If the refund is above $50, it must be approved by a supervisor before a
check is cut.

On a process map, the three-prong decision would come out of the or shape. The

Merge Shape

If processes can branch off based on decision shapes and or shapes, they might also come
back together. When multiple branches of work flow within a process come back together,
you can indicate this with a merge shape.

Extract Shape

If processes split into two or more parallel flows without a decision or criteria, then you can

on process maps, and sometimes the triangle indicates something else. When the triangle is

664
 SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

presented with an M inside of it, it can show that a measurement must occur at a certain
point in the workflow.

Collate Data Shape

The collate data shape is used when data or material in a process has to be organized using
a standard format. Cutting fabric into certain lengths before sewing or ordering the
admissions forms before considering them are both examples of when this symbol might be
used.

Sort Data Shape

The sort data shape indicates that information or materials must be arranged in a certain
order. Sorting materials by color before moving to the next step in a manufacturing process
might warrant this symbol, for example.

All of the shapes within your flow chart or process map are connected with lines and
arrows. The arrows are important, because they guide the reader through the map in the
proper sequence.

All of the shapes listed in this chapter are available in Microsoft programs, including Word,
Excel, and Visio. For those without specific process-mapping software, Excel is probably one
of the easiest programs to use for creating process maps and flow charts. Simply turn off

665
 PROCESS MAPS

the view gridlines function to give yourself a clean slate with which to work.

You can insert all of the flow chart shape into Excel using the Insert > Shapes option.

While someone familiar with process mapping tools will recognize most of the above
symbols quickly, not everyone in a business environment will. Most people in business
environments are familiar with the basic process shape, decision and connector shapes, and
terminator shape. Because of this, you can create the bulk of most process maps using

map key to your diagram to ensure anyone who looks at it can follow it.

cause not everyone uses all of the above

symbols for exactly the same reasons. The connector shape can mean several different
things, as can several of the other shapes.

666
 SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

You probably noted as you were reading through the list of common shapes that some
activities might be appropriate for two or even three process shapes. Outside of some basic

goal for creating a process map should be clarity. If adding a different type of shape is going
to make the process easier to understand on your visual, then do it. If you can simply use
the basic process shape along with text to spell out the process, then remember that in Six
Sigma, less really is more.

A basic flow chart begins with a trigger activity, includes inputs and processes, decisions,
and a final action. Flow charts usually begin at the top of the page and flow down the page.
Secondarily, you would read a flow chart from left to right as you move down the page, if a
secondary direction is required. In the above flow chart, you can see that a trigger activity
occurs, which involves some type of data input. Using that input, a person makes a decision.
That decision directs the flow of work either to the process on the right or the process on
the left. After that action is done, the overall process, as recorded on this flow chart, ends.

chart or process map rarely ends the work itself. The only time the final terminator signals
the complete end of the work is when you are mapping the very last process before the end
667
 PROCESS MAPS

customer or you are mapping the entire enterprise process at a very high level. In most
cases, however, the final terminator on your process map is likely to indicate the end of this
particular process and the beginning of the next. For example, in a beverage bottling plant,
a team might be mapping the process that prepares bottles to receive ingredients. The
process map begins when a fresh, new bottle is introduced to the process and it ends when
the bottle is cleaned, has a label, and is put on the line to receive the liquid it will house.
After the process that was mapped, the bottle goes through other process to fill it, cap it,
inspect it, and pack it.

One of the most widely used process map formats in the Six Sigma world is the swimlane
diagram. The swimlane diagram is a flowchart, but it is one that is constrained somewhat by
rows and columns, known as swimlanes. You can design your swimlane as you want to best
depict the process you are working with. Two common options are:

1. Using columns to depict time or activity phases in the process and using rows to
depict the person, department, or group responsible.
2. Using columns to depict the person, department, or group responsible and using
rows to depict the time or activity phases in the process.

668
 SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

No matter which method you choose, someone should be able to read your swimlane
diagram by starting at the top left and reading to the right and down.

Start your process map session by identifying the swimlanes associated with the process

then you can use the information from that activity to create your swimlanes. If not, you
can use some of the brainstorming activities from the previous chapter to help generate
ideas for your process map. Responsibility diagrams, for example, can help you understand
how the activities within the process might be broken into organic categories.

ask yourself or your team:

w
even getting all the detailed steps brainstormed. You can add to your map as you create it.
ing too much about the order of steps or
about how they are connected, use notecards to write steps on for this stage. Once the

669
 PROCESS MAPS

team thinks all major steps are written down, you can sort the steps into groups to get a
better idea of possible swimlanes.

To understand how swimlanes might look for a real-world process, consider the process a
brand uses to generate ideas for blog posts, write blog posts, and post blogs to a website.
You can see, just from the previous sentence, the process is already somewhat broken
down into major categories:

Ideas are created

Blog posts are written
Blog posts are posted

Those three sections might become the row designations for the swimlane process map.

Now, consider that different people or groups handle each of the subcategories within the
process. The marketing team generates the ideas for content, a team of contracted writers
writes the blog posts, and an internal quality team reviews and then posts the blogs.

The foundation of the swimlane diagram might look something like the figure below.

670
 SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

Next, create a simple step-by-step list for all the activities within the process. If you are
familiar with a process and are making a process map on your own, you can start putting
the steps into shapes on the grid you can always move them around as needed later. If
you are not familiar with the process or are working with a group to create the process
map, it helps to gather data first.

For the purpose of the blog generation example, here are some steps that are included in
the process.

1. At the beginning of each month, the marketing team generates ideas for the next

2. The marketing team generates 15 to 25 ideas, but chooses only 8 to 10 ideas to be

written.
3. Once the 8 to 10 items are chosen, the marketing team chooses a title and
keywords for each blog post.
4. The topic, along with any necessary special instructions, the titles, and the
keywords are provided to the contract writing team.
5. Writers draft blog posts.
6. Writers check that blog posts fit guidelines.
7. Writers submit blogs posts.
8. Quality editors review blog posts.
9. Quality editors check that blog posts fit guidelines and request revisions if they do
not.
10. Quality editors line edit blog posts.
11. Quality editors format blog posts.
12. Quality editors enter posts into blogging platform.
13. Quality editors schedule blog posts for publication in the next month.

You can break the above steps into various swimlanes and rows. Steps 1 through 4 belong
under the marketing team and mostly in the row for idea creation. The very last step begins
to cross the lines into the next phase of the process.

Steps 5 through 7 belong under the writing team and typically fall into the writing phase of
the process. The last steps belong under the quality editor section, but they do involve a
decision step that could send work back to the writers.

Consider how steps 1 through 4 might be placed on the swimlane diagram.

671
 PROCESS MAPS

The terminal shape the activity that starts the entire process is that it is time to
brainstorm the blog ideas for next month. The team brainstorms ideas and then chooses 8
to 10. For the next step, the activity is a predefined process. This means that somewhere,
the company has a published SOP or guidelines for researching and applying keywords to
blog post ideas. Once that is done, someone chooses a title for each idea and the blogs are
assigned to the writers. Note that the final activity assigning the blogs to the writers is in
a rectangle that somewhat straddles the swimlane line. This is a helpful visual cue that the
responsibility for work at this point in the process shifts from the marketing team to the
writers.

672
 SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

Next, consider how steps 5 and 7 might look on the process map.

The three steps are presented under the contracted writers swimlane and in the row
associated with writing blogs. The final step crosses the line slightly, indicating responsibility
is being passed to the quality editors. Consider the remainder of the process map, which
falls under the quality editor swimlane.

673
 PROCESS MAPS

You can see that the quality editor responsibilities actually span writing and posting phases.

back to the writer. The team can see here that there is a possibility for rework in this
process, and it can see exactly how that rework is handled. The entire process map is shown
in the image below.

While this is a fairly basic example as process maps go, it includes all of the basic premises
of creating a swimlane process map. You can apply these basics to create simple to very
complex process maps of any process in Excel or with Word or other any programs that
include process map shapes as an option.

674
 SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

Admittedly, the process maps shown in the above images are a bit rough. That was done on
purpose to show you how the creation of a process map occurs as you are working through
brainstorming and thinking about the process. If the sole purpose of the process map is to
provide a Six Sigma team with visual guidance as it works through the DMAIC process, you
can leave the process map in a rough-draft state.

If you plan to use the process map in presentations or training documents, however,
consider cleaning it up a bit. First, make sure all the color combinations that you used will
translate well to whatever medium you plan to use the map in. The above map features the
white-on-blue text in each shape, which might not translate well to black-and-white
hardcopy documents.

Next, take time to make sure shapes and arrows line up in an aesthetically pleasing and
logical fashion. Review your map to ensure arrows are pointing in the right direction, try to
keep arrows from crossing over each other too much, and ensure the start and end of each
arrow is touching a it.

Finally, use concise and active language when possible to keep text in each shape short and
understandable. Here are some edits that could be made to the example above to make
language more active and concise.

You can see that the shorter versions almost always start with an action verb, which is ideal
in process mapping and flow charting.

675
Value stream mapping is similar, but in some ways different, to process mapping. Value
stream mapping is a tool perfected by lean manufacturing and works best in processes that
are manufacturing, delivery, shipping, or factory oriented. However, value stream mapping
can be useful in a variety of other industries if the Six Sigma expert understands the purpose
of the exercise and how to translate the method into his or her process.

Value stream mapping is a pictorial representation, including the addition of numerical

data, of the value stream of product or ideas. When used properly, value stream mapping
provides Six Sigma teams with the big picture of a process, lets teams understand what the
process looks like today, helps teams create a vision for improving a process, and links the
flow of materials, people, and information in one image. A value stream map also helps
break down communication barriers because the symbology used tends to be similar or at
least intuitive; as such, it makes it easier for teams and support staff to understand where
waste or muda is occurring in a process.

A value stream map is one of the few tools that truly captures all of the flow that occurs in a
ajor seven flows on a
value stream map, though most value stream maps concentrate most on raw material,
parts, and products than on the other four types of flow. The seven types of flow are:

People
Raw material
Parts
Products
Equipment
Information
Engineering

676
 SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

Value stream maps use different symbols than most process maps do, and the symbols tend
to be more specific. For example, if the value stream involves the delivery of goods via
truck, the map includes an actual picture of a truck usually with a text indication of how
many times the shipments occur in a given period.

The image above indicates one truck shipment per week. A number of other common value
stream map symbols are included below. You can download common symbols for Excel
from www.lean.org/ common/display/?o=866 to create value stream maps in Excel. You
can also use a value stream map software to create a map.

Note that value stream mapping , especially if

you are using the concept of value stream mapping in an industry where such symbols are
less useful. You could use the value stream concept to map the workflow in an emergency
room department, for example, where different symbols would make more sense. Consider
these symbols as a starting point.

The manufacturing process box indicates a specific action in the production process. The
name of the action goes in the top part of the box in thi
can put text notes or illustrations as desired. One common use of the bottom section of the
process box is indicating how much time occurs during that particular step.

677
 VALUE STREAM MAPPING

Outside sources are either suppliers or customers. Use this shape to indicate when inputs
come from outside the organization or process or when outputs go to the end customer of
the process.

Data boxes are important because they house the analytical meat of the value stream map.
Common information recorded in data boxes include:

Cycle Time, or C/T. Cycle time is the time it takes for the process as a whole or a
step within the process to complete one iteration.
Changeover Time, C/T. Changeover time is the time it takes for a change in shift,
tools, or other resources in the process or a single step within the process.
Number of operators
Number of shifts
Time worked in a shift
First-pass yield quality
Batch sizes
Takt time, which is the rate at which products must be finished to meet the demand
of the customer. Takt time for an individual step in the process might be seen as the
time in which the step needs to be finished to keep from slowing down the process.

678
 SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

The inventory box indicates that inventory usually stacks up at a certain point in the
process. It is usually coupled with text that says how much inventory is available in amount,
time, or both. You might note that there are 100 pieces of inventory regularly at that point
in the process. You might note that there are two- rth of inventory at that point in
the process; you could note both.

A push arrow indicates that something either product or information is pushed from one
part of the process to another part of the process. The push arrow means the product or
information is generated by that part of the process regardless of the need for it in the
downstream parts of the process.

The opposite of a push arrow would be a pull indicator, which means that downstream
processes pull inventory, parts, or information as needed from upstream locations.

The finished goods arrow indicates that the output of the process is being pushed or
delivered to the customer whether that customer is internal, external, or another process.

679
 VALUE STREAM MAPPING

The FIFO arrow indicates that this part of the process features a first-in, first-out sequence,
which is important in some processes. Consider processes that deal with food ingredients; if
a recipe in a manufacturing plant calls for milk, then the operators should pull the oldest
(still good) milk from storage first to keep from wasting product that goes bad while in
inventory.

The supermarket is an area in the process where small amounts of inventory are stored.
Downstream processes can stock as needed from the supermarket instead of stocking from
the larger inventory supply, which might be more difficult to access. An upstream process
feeds the supermarket, ensuring enough inventory is stocked. An example of a supermarket
might be seen in a large office. One team might have a drawer where supplies, such as
paper, pens, paperclips, and staples, are kept. The team can pull from that drawer instead
of requisitioning from the larger stock room or from the actual office supply store every
time supplies are needed.

The physical pull symbol indicates that product or materials have to be physically pulled
from the previous step, station, or supermarket to the next step.

680
 SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

The move-by-forklift symbol is self-explanatory, though not always necessary in every

process. You might, however, have a symbol for other types of movement, such as move by
hand cart or conveyor belt. The only time you need to include specifications about how
items are moved is when those details might become important to the process analysis.

This is also a self-explanatory symbol and also not necessary in every process. You might,
instead, have train, plane, or even bike shipments listed on your value stream map.

The type of information indicated by the straight manual flow arrow is typically
communicated via memos, face-to-face speaking, phone calls, and reports. The manual
information arrow might be accompanied by notes regarding the type of information and
the frequency with which it is delivered.

681
 VALUE STREAM MAPPING

The crooked electronic info flow arrow indicates information that is electronic in nature,
such as information communication via Internet, portals, EDI, and automated reporting
systems. The type of data and the frequency with which it is delivered can also be included
with this symbol.

The withdrawal Kanban shape indicates that there is a Kanban card or signal at that process
step. The Kanban card instructs the operator at that particular step to withdraw inventory
from a supermarket or other stock location.

The production Kanban shape indicates that a production Kanban card or signal is located at
that part of the process. This card tells the operator to produce a certain number of parts to
feed a process that is downstream.

682
 SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

-in-
in the process. Safety stock is not like a supermarket it is not intended to be a permanent
storage of inventory. Instead, it is a temporary storage of inventory that reduces the chance
that downtime or other issues upstream will keep all the downstream processes from
occurring. For example, in a fast food restaurant, someone stocks the cup and drink
dispenser each morning. These stocks are not meant to be the permanent storage for those
items, but a smaller number of items are kept at hand so employees can quickly supply
drinks to consumers.

have to include the operator shape in every diagram where people are in some cases, the
nature of the value stream map allows readers to assume people are part of the process.
However, if automated processes are included amid people-powered processes, this symbol
can be used for clarity where needed.

The schedule box is simply a text box that lets you note how work, orders, or other parts of
the process are scheduled. You can couple the box along with arrows to show that orders
are sent weekly, for example, or that a production schedule for staff is created each month.

Creating a value stream map is very different from creating a process map. One point of
value stream mapping is that it forces you to really experience the process, which means
that committee-

683
 VALUE STREAM MAPPING

entire team can get involved in value stream mapping, if you as the Six Sigma expert think
that is valuable, but the best results often come when everyone starts with their own
individual value stream map. This is also a good process for the Six Sigma experts in a team,
because it forces each one to view the process individually.

While you can use the icons above and numerous software programs to create a final copy
of your value stream map, use paper and pencil for your first drafts. This lets you record
information as you walk through the process, changing it quickly as you go.

When creating process maps, you often concentrate on the process, the employees, and
the requirements. When creating value stream maps, you should always concentrate most
on the customer and the product.

Physically walk the process whenever possible instead of sitting at a computer or working
with a brainstorming session. Some experts recommend walking the process backwards;
others say start at the beginning try to map
every single part of the process, especially if you are mapping a large-scale production
process. Instead, follow the major raw materials through to the outputs. If you put all the
processes on a single map, it becomes too hard to read and you lose the value of the tool.

While you might eventually create a value stream map of a desired future state, start by
mapping the way things are
Use real-time data for the value stream map whe
historical production and changeover times; when you walk the process, use a stopwatch to
record your own observations.

Begin value stream mapping by selecting the product family or process you are concerned
with. In a manufacturing environment, you might select product type C and only concern
yourself with that product. For example, a toy factory might make dolls, doll houses, and
toy cars, but the Six Sigma team might only be concerned with doll houses. In fact, the Six
Sigma team might only be concerned with the doors that are made for the doll houses you
decide what scope of value stream map makes sense for your project.

Next, collect as much relevant data on the current state of the process as possible. Do this
by walking the process and taking notes, reviewing actual data, and speaking with subject
matter experts.

684
 SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

To create an example of a value stream map while going through the follow
consider a simple real-world example. In our example, a factory makes crayons. The Six
Sigma team is concerned with the process that makes the 8-pack of primary colored
crayons.

The value stream map will need to include information such as:

The suppliers of raw goods (wax, dye, and paper)

Wax melting
Dye addition
Pouring into molds to create crayon shapes
Inspection points
Wrapping crayons with appropriate paper
Packing crayons into boxes of eight
Shipping to final customer -- in this case, retail stores

1. Begin with the major process steps in the value stream. These are indicated by
process boxes and are joined in a left-to-right line by the appropriate arrows often
steps:
a. The wax is melted and dye is added
b. Wax is poured into molds
c. Molds are emptied
d. Inspection occurs
e. Shaped crayons are wrapped with matching colored paper
f. Crayons are packed in boxes
g. Shipping

685
 VALUE STREAM MAPPING

Above, you can see a rough sketch of the process part of the value stream map. Kanban
cards or signals tell employees when to pull wax and dye from the supermarket inventory.
(The signal might be a light that the previous batch has been processed and the machines
have returned to process start parameters). The employees add the ingredients and start
the first piece of the process, which is automated after the initial ingredient measurements.
The wax is melted, the dye is mixed in, the colored wax is poured into molds, and the mold
sets for 10 minutes before crayons are emptied to the inspection station.

show that they are automated and the process simply runs together unless someone
intervenes in that batch. The rest of the processes have push arrows, because inspection,

is flowing through.

2. Add the important aspects from outside of the process, including process controls,
supplies, customers, and shipments.

686
 SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

3. Add data boxes for each part of the process.

4. Add inventory signals and any other information you think would be pertinent for
the value stream map. For example, in reviewing the crayon process used in the
687
 VALUE STREAM MAPPING

example, the Six Sigma expert might note that inventory builds up before the
packing process because it is slower than the inspection and paper-wrapping
process. That could be indicated with an inventory signal and text.

You can also add controls for the process. In this case, the retailer places orders one
month ahead for how many boxes it will need each week in the next month. The
crayon-making company creates a monthly forecast, but orders once a week from
suppliers. Suppliers deliver in multiple batches a week. Production control manages
every part of the process with scheduling to keep up with the demands of the retailers.

688
 SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

Finally, add a timeline at the bottom that notes the time it takes for a single item to process
through the enter process, breaking the time into value added and non-value added.

In the above example, the yellow portions are value added. They represent time that the
product is actually being made or changed. The white portions are the setup, lead times,
and inventory wait times and this particular process is for a single color of crayons and

You can add up the times to see that the total process time is 194 minutes or 3.23 hours.

The value added time is 19 minutes, so the team can assume there is definitely waste that
can be removed from this process.

they are working

pictures of the process to help you and the rest of a Six Sigma team better understand the
process and related challenges. So, you can add as much or as little information as you think
is necessary. You can also come back and add information later if needed. The first draft of a
value stream map, whether using pen and paper or a computer program, is probably going

make a clean, final copy for inclusion in presentations and training documents if desired.

Value stream mapping is a helpful tool for all of the DMAIC phases. During Define, value
stream mapping as an activity forces you to become involved with the process at a physical
and daily level, which can boost your understanding of the process you are working to
improve. It also provides the entire team with another visual that helps when discussing the

689
 VALUE STREAM MAPPING

process and the problem. A good value stream map lets the team visualize exactly where
the process and problem might occur with respect to overall production.

a foundation for the Measure phase. The value stream map also points to measurement
and data collection points that can be further used in Measure, though teams should never
limit themselves to just the first types of information gathered in value stream mapping.

During Analyze, value stream maps are a tool that helps teams brainstorm and identify root
causes or problem points in a process, and you can use the initial value stream map to
those that create muda in the
process. Finally, during Improve, teams can begin creating a future state value stream map,
building off the old map and using the information and solutions they have come up with
through the DMAIC process. By putting those solutions on paper in the form of a value
stream map, teams can better understand where issues might arise and alter solutions
appropriately to reduce the number of issues experienced during implementation and
beyond.

1. Use the current-state value stream map and all other information gathered and
analyzed to identify as many sources of waste in the process as possible.
2. Categorize waste into two types: those that are the result of remote locations,
machinery, and product designs which are all necessary elements to the end
product and those that are not the result of these things.
3. Concentration on the second type of waste and create a future state value stream
map that removes them all.
4. Review the future state map to ensure processes can flow appropriately. Ask
yourself questions such as:
a. What is the Takt time required at each station to meet overall demands,
and is each step in the process capable of meeting that Takt time? If not,
the team might need to address steps that are not capable.
b. How will production be scheduled? Usually, there is a specific part of the
process that controls the production schedule. Ultimately, the best and
most optimal control is by the customer (The customer needs 100 per day,
690
 SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

so the process should make 100 per day). However, there are times when a
specific part of the process is only capable of X units per time period, and
the team is unable to change that capability because of resource or other

inventory is going to build up before it.

c. Is batch processing needed, or can it be removed?
d. What process changes and improvements are required to implement the
future state value stream? Sometimes, machinery changes, staffing

possible. Before moving forward with future-state plans, Six Sigma teams
have to make sure resources will be approved and available.

requirement. In fact, one of the cardinal rules of value stream mapping is not to map just to
map. Yes, a finished and polished map looks nice and can be especially impressive in
presentations or documents when other people in the organization are not fully familiar

mapping in this fashion, then skip this tool. A Six Sigma expert should always be aware that
muda can creep into any process and that includes the process of evaluating and

creating muda of overproduction.

691
For the purposes of this chapter, healthcare is considered to be inclusive of the entire
industry. In addition to providers both small practices and large organizations such as
hospitals and hospital campuses healthcare can include support services, equipment
manufacturers and distributors and insurance companies. Six Sigma experts who work in or
with any organization that is concerned with patient care, federal and state healthcare
compliance, and auxiliary services such as claims billing, should understand how the
methodology integrates into the field.

Some overlap does exist between the industries covered in these last few units. A company
that manufactures medical devices is likely to operate in several spheres, including both
manufacturing and healthcare. Within that company, departments such as human
resources and IT are concerned with both the manufacturing and healthcare needs as well
as the functions and needs specific to their own department. Because of this type of
overlap, the Six Sigma expert should understand how the methodology can be implemented

how Six Sigma can benefit specific industries and departments and how the environment

about Six Sigma cases and success stories in a variety of industries.

692
 SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

averaged operating margins between 2 and 7 percent between 1993 and 2015. Total
margins for the same time period were between 4 and 8 percent.8 With fairly small profit
margins, hospitals and healthcare organizations have to keep tight reins on processes and

strategic improvements are a must within healthcare environments.

quality of care and the cost of service9 -users such

as patients. Hospitals, and all healthcare providers, compete for referrals, industry
partnerships, accreditations and contracts with insurance companies and payers. Because
processes in healthcare organizations have so many customers, it can be difficult to make
improvements that enhance services for one customer without negatively impacting
outcomes for another. A change that makes a service faster for a patient might make it
more expensive; a change that reduces costs could be noncompliant with one or more
regulations. The DMAIC methodology helps teams consider all customers and outcomes to
avoid such issues.

At the heart of most healthcare organizations even those that serve in auxiliary capacities
is service to the patient. This is particularly true in clinical circles, where errors can lead to
negative patient outcomes, including longer treatments, healthcare-related illness and
injuries, and even death. Not only do these outcomes create a negative impact to the
and open possibilities for expensive malpractice suits, but they are also
undesirable because healthcare employees are in the business of saving lives and improving
living, not in making things worse. Six Sigma can be used in clinical settings to improve
processes and make positive outcomes more likely.

Finally, Six Sigma can be used to reduce errors throughout the healthcare service chain.
Errors in administrative tasks, such as patient enrollment or claims billing, lead to delayed
care or payment. Errors in inventory ordering lead to shortages that impact the ability to
serve immediate needs. Six Sigma has been used in healthcare to reduce physician

8
http://www.aha.org/research/reports/tw/chartbook/2015/chart4-2.pdf
9
http://www.beckershospitalreview.com/lists/200-hospital-benchmarks-2015.html
693
 SIX SIGMA IN HEALTHCARE

error, reduce patient wait times, decrease supply chain steps and costs, create stronger
reimbursement flows, and decrease the time it takes labs to return results. Those are just a
few uses of Six Sigma in healthcare.

While Six Sigma can be extremely valuable in a healthcare setting, the methodology does
face some challenges specific to the industry. One of the first, and arguably one of the
largest, unique challenges Six Sigma experts face in a healthcare setting is compliance. Yes,
almost every industry these days has a compliance or regulatory requirement somewhere
st units. However, the compliance element for
healthcare can be constraining in ways that require creative and strategic thinking on the
part of Six Sigma experts.

Six Sigma experts should never try to work around or over compliance requirements. First,
the regulations of the industry are automatically part of the critical-to-quality characteristics
a Six Sigma team should consider. If something is truly a regulatory requirement, then Six
Sigma teams must honor it and treat it as a quality requirement. Working over that
requirement means you are producing a process or product that is not within quality
standards. That being said, this can be frustrating for Six Sigma experts, particularly if they
are used to working in other industries. Healthcare quality requirements, which include
clinical requirements, HIPAA (Health Insurance Portability and Accountability Act)
requirements, and dozens of other requirements from insurances companies and state and
federal agencies, can force muda into a process. Regulations often require duplication,

It's important to note, however, that some muda that healthcare organizations associate
with compliance is unnecessary. One job of the Six Sigma team is to identify unnecessary
compliance work and replace it with processes that ensure quality without waste. Here, you
find a second big challenge Six Sigma experts are likely to find in healthcare settings a
challenge that is found in most industries and which was covered in the beginning of this
book. When people are used to doing something a certain way, they come to believe they
have to do it that way. In healthcare, this often

694
 SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

Sigma teams must work hand-in-hand with compliance to understand when this is true and
when a more efficient quality process can be put in place.

While healthcare environments present numerous challenges for Six Sigma teams, the last

know what outcomes will be affected by decision-making within the DMAIC process and
must set confidence levels and sample sizes appropriately when conducting analysis. When
dealing with processes that involve administrative functions such as claims billing, teams
might default to the 95 percent confidence level (alpha = .05). When dealing with processes
that directly impact patient welfare, such as surgical procedures, teams might consider
setting confidence levels at 99 percent (alpha = .01). Remember, higher confidence levels
require bigger sample sizes or more runs in a designed experiment. In some cases, they are
more expensive overall, so teams need to truly understand the impact of confidence levels
on both the cost of statistical analysis and the accuracy of conclusions before moving
forward in healthcare arenas.

Located in New Jersey, Virtua Health spanned four hospital systems at the time it launched
a Six Sigma project to create solutions to challenges in cardiac medication processes. The
healthcare organization started to look at quality improvements in this area in 2001; one of
the goals at the time was to align the processes of the organization with the goals required
by the Centers for Medicare and Medicaid Services (CMS). The organization collected and
reviewed data, noting that the process for providing cardiac patients with medication and
documenting that provision needed work to come up to standards.

The organization launched a traditional quality program in 2002. It called the program POE,
or Cardiac Program of Excellence. Through the program, Virtua Health attempted to adopt
best practices and advanced technologies and increase connections between providers, but
the results were lackluster. Data from the first quarter following the implementation of the
program showed no measurable improvement. At that time, the organization launched a Six
Sigma project. The team included subject matter experts from multiple disciplines and they
approached the issue via a DMIAC methodology.

Define. Increase quality of patient

care by use/non-use and appropriate documentation of aspirin, beta-blockers, and

695
 SIX SIGMA IN HEALTHCARE

ACE (angiotensin-converting enzyme) inhibitors in CHF or AMI patients to achieve or

10
exceed Virtua be
o Scope: all four hospital campuses
o Goal: Meet Joint Commission Accreditation standards and improve patient
outcomes
Measure. During measure, the team conducted a Gage R&R analysis on the quality
review process and DPMO was calculated regarding the administration of
medication process.
o The team defined a defect as any failure to properly administer medicine,
which meant errors in dosage amounts or the time medication was
provided would count as defects. Defect rates for administering medication
to CHF patients was 10.2 percent; defect rates for administering medication
to AMI patients was 4 percent. (CHF and AMI are two types of cardiac
patients.)
o The team also noted that there was an issue in the quality review process.
The reviewers were not consistently reporting quality in a way that
matched their own review results. To address this problem, the Six Sigma
team, along with subject matter experts from the quality review teams,
created a detailed process map. They also came up with a standard for
reviewing and a plan to communicate that new standard to all quality
reviewers.
Analyze. During the analyze phase, the Virtua Health Six Sigma team met biweekly
and focused on approximately half a dozen issues identified during Define,
Measure, and Analyze. The team noted that concurrent reviews of CHF and AMI
patients required better coordination of staff. They also noted that nursing and
clinical staff needed better access to information about core indicators, especially
when that information evolved or changed. The Six Sigma team identified common

planned to post storyboard templates featuring core indicator data at each location.
Other information and activities that developed in the Analyze phase included:
o A need for better communication between clinical staff and physicians
o Discovery of overproduction in reviewing charts multiple staff were all
reviewing the same charts on a regular basis without a real clinical need to
do so

10
https://www.isixsigma.com/new-to-six-sigma/dmaic/achieving-and-sustaining-
improvement-cardiac-medication/
696
 SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

o Real-time communication and service administration was not always

available
o All of the activities needed to work together, requiring a point person
assigned to supervise and coordinate
Improve. During Analyze, the team noted that the majority of medication errors
involving physicians were related to defects in documentation. Specific factors
included missing documentation, inconsistent patient census, missing diagnostics
, compliance issues with regard to completing discharge instructions, and
inconsistent patient care across multiple disciplines within the hospital setting.
Remember, some cardiac patients might have other diagnoses and be treated in
other areas or by other staff. The team considered each of these factors, developed
root cause analysis, and proposed solutions.
Control. The team created a flow chart for the distribution of key information and
trained appropriate staff members on that process. Standards were put in place for
physician-to-physician communications and a monthly dashboard was built so the
staff and organization could track ongoing process. A standard data collection form
became part of the process to reduce inconsistencies, some forms were shortened
to make consistency more likely, and certain medications were stocked on each
floor to reduce administration times.

At the end of the project, the Virtua Health organization did see statistically significant
results. Some Y benefits from the process improvements included:

An increase in the compliance requirement for administering aspirin to cardiac

patients
Increased compliance in documentation
Increased quality of care for patients
Increased timeliness of patient care
New processes that brought the organization in line with CMS requirements for
cardiac medication provision.

A medical transcription company in India servicing physicians and other providers in the
United States was having difficulty returning transcriptions within a desired time period

697
 SIX SIGMA IN HEALTHCARE

while maintaining appropriate quality.11 The company worked with a Six Sigma consulting
firm to make improvements to the process. After some initial training regarding Six Sigma
and Lean methodologies was provided to key employees at the transcription company, the
consultants and transcription company staff began a journey through an improvement
project.

The first challenge was identifying what the customer wanted. Staff members knew
customers wanted dispatches by 5:30 p.m. each day, but what information did that include?
After using some Define tools, the team noted that all files received between 7:30 a.m. on
day 1 and 7:30 a.m. on day 2 must be transcribed and returned by 5:30 p.m. on day 2. Once
this was defined, the team began to gather data about the process, noting that there was a
wide variation in performance that was causing the target to be missed. The team also
noted that work arrived steadily from the United States but peaked between midnight and
2 a.m., making it difficult to catch up if processing of work batches began at 8:00 a.m. each
day.

Ultimately, the team arrived at the conclusion that at least one night team was required.
The team decided to use two night teams and one day team; work processing would start at
8:00 p.m. instead of 8:00 a.m.

At first, some team members vocalized worries that staff working the night shift would be
less productive. The general feeling of team members was that people would be more tired
at night or that night work would not be managed as well. The Six Sigma team ran a pilot to
test this theory and found that there was no statistical different between efficiency for
night teams and day teams.

The pilot was successful leading to on-time or early delivery of all piloted work. The
transcription business quickly adopted the changes across the organization, resulting in an
average delivery of work that was 134 minutes early rather than 89 minutes late. The
system was also able to handle 22 percent more work on a regular basis.

11
https://www.isixsigma.com/new-to-six-sigma/dmaic/medical-transcription-six-sigma-
case-study/
698
 SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

To be successful with Six Sigma in healthcare, teams and Six Sigma experts must:

Understand all of the customers of the process

Take time to develop a deep understanding of compliance requirements related to
the process at hand
Work hand-in-hand with clinical, administrative, and auxiliary staff to make
appropriate improvements
Work with the appropriate confidence intervals depending on the type of process
involved

699
Every organization in any industry has a finance component. Money is made and
collected, invoices are generated, and bills and employees are paid. This is true even for
nonprofits. And while the information in this chapter might be somewhat relevant to the
accounting, purchasing, receiving, or finance departments in any organization, this chapter
is going to concentrate on organizations in the finance industry. Note, however, that
finance-related departments should often be included in Six Sigma projects in every
organization, if only from the perspective of as-needed subject matter experts that can
provide details or assistance regarding budget, costs, and revenue data.

For the purposes of this chapter, finance organizations are considered to be those that deal
with money, investments, collections, mortgages, debt, and related services and products.
Banks and credit unions, insurance companies, loan companies, and collection agencies all
fit this description. Other types of entities that might fall into this arena include CPA firms,
financial software vendors, money transfer services such as Western Union, and online and
app-based money services such as PayPal.

All of the typical benefits of Six Sigma can be experienced when applying the methodology
to finance processes, including increased efficiencies, reduced defects, and an increase in
profits. As with healthcare, the finance industry as a whole is managed under some very
specific regulatory laws and requirements. Requirements address how investments must be
handled, how data particularly personally identifying data must be secured, accessed,
and transferred, and who can do what with financial information and when they are
allowed to do it. For example, if you work for a company and have insider information
about that
investments before the knowledge might be considered public; doing so is called insider
trading and is a felony. Failure to comply with regulations can mean enormous financial and
brand implications for an organization and on an individual level, can even translate to
criminal charges for those involved.

700
 SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

The ability to maintain compliance in a complex industry with so many moving parts is a big
benefit of Six Sigma. The critical-to-quality approach of Six Sigma helps ensure that
processes are designed and improved with compliance factors in mind, which can help
organizations avoid damaging expenses and negative brand publicity from regulatory issues.

The challenges of implementing Six Sigma in an organization in the financial industry are
often similar to those experienced by Six Sigma experts in healthcare. First, you face the
compliance and
you often have to fight a false perception of what those regulations might be.

Second, while the finance industry has embraced technology innovations such as app-based
money management, it is still steeped in tradition and outdated policies to some degree.
One reason so many financial industry employees and leaders are reluctant to embrace
change is because of the high stakes that are inherent when you deal with money
especiall
hard to create an atmosphere where financial leaders are comfortable relying on the
methodology and the statistical analysis process for decision making.

Third, like healthcare organizations, financial organizations often deal simultaneously with
multiple customers and often, the goals and desires of those customers are somewhat at
odds with each other. For example, a publicly-traded financial institution has both
shareholders and end-user customers. The end-user wants the best possible product at the
lowest possible cost and highest possible return. The shareholder wants maximized profits
for the biggest return on investment. While these things are not always mutually exclusive,
they are also not always aligned, which can be a challenge for Six Sigma teams looking to
improve outcomes for one customer teams have to consider the outcomes for all other
customers too.

Fourth, unlike healthcare industries, which tend to remain fairly siloed within national
borders unless you are working with agencies such as Doctors Without Borders or the
World Health Organization, companies in the financial sector often work across borders.
Depending on the location and scope of a Six Sigma project, teams might deal with global
processes or interactions. Challenges in a project that spans cultural and government
borders are many: language barriers might exist, and even when everyone speaks the same

701
 SIX SIGMA IN FINANCE

different, and teams must understand how varying government entities play a role in a
process and project. Finally, cultures and dynamics are different, so Six Sigma leaders must
be sensitive to different ways people deal with communication, assignments, and each
other.

Finally, like some of the other industries on this list, the financial sector often deals with
services and ideas that are not immediately or never will be tangible. This is even more
true in the digital age. Yes, some services and products have very definitive values. An
insurance policy is worth a certain amount; an investment portfolio has a specific current
ducts that the end-
made from raw goods through a manufacturing process. Because of this, many people who
only have a basic understanding of Six Sigma falsely believe that it is not the right
methodology for improving processes in their organization or department. Because there
tenets of Six Sigma

to develop training, explanations, and successful initiatives that continue to create a culture
of Six Sigma-driven process improvement.

The case study presented for the finance industry illustrates how Six Sigma can become a
long-term culture in an organization, creating constant improvement not just through
months or years, but through decades. In an effort to become a premier financial institution
across the globe, Citibank turned to Six Sigma as early as 1997.12 At that time, the bank
partnered with Motorola, which had developed its University Consulting and Training
Services in the last part of the 20th century. As part of its commitment to quality and process
improvement, Citibank wanted to reduce cycle times, increase customer satisfaction and
loyalty, empower team members to make decisions and help customers, and improve cash
management processes.

In 1997, the Motorola University Consulting and Training Services team started teaching key
Citibank employees about Six Sigma methods how to reduce defects and why that was
important and how to use Lean concepts to positively impact cycle time. Here, in the late

12
http://www.qualitydigest.com/dec99/html/citibank.html
702
 SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

1990s, a financial organization asked itself: could what has worked so well in manufacturing
circles be applied to financial services? It turned out that CTR (cycle time reduction) did
provide benefits outside of a manufacturing or factory environment. Citibank used CTR
strategies to create faster, less costly, and higher quality consumer banking processes; it
also used CTR to maneuver in emerging markets in a manner that kept up with the quickly
evolving economies.

In addition to identifying defects using traditional Six Sigma tools discussed throughout this
book, Motorola worked with Citibank to apply a method developed by Motorola called
cross-functional process mapping, or CFPM. The purpose of CFPM is to reduce cycle times
via a series of steps.

Plan for the project for two to four weeks. Planning includes choosing a process for
improvement, identifying a champion and team members for the project as well as
a steering committee to guide the project, and choosing a Six Sigma team leader.
Over the course of two to five days once a team is created, they map the current
process, asking questions about where waste is, what might create dissatisfaction
for the customer, and what are the value added activities of the process.
Over three to five weeks, teams work to validate their map and assumptions with
other departments, using the time to create a culture that is accepting of the need
for change.
During a second two to five-day session, teams create a process map or flow chart
of what they want the process to look like.
Over five to tenth months, teams work with the organization to design and
implement the process.

Citibank used

improvement journey to reduce waste and defects in its processes.

While Citibank began the Six Sigma journey in the late 1990s, it has continued to develop a
culture of process improvement throughout the decades. Leaders with Citibank have
deployed Six Sigma thinking to create ongoing improvement and to solve real-time issues.
One managing director identified a problem when a function in his purview was relocated
to an office in a different state. Because not all employees wanted to relocate, the function
ended up in the hands of staff who were not experienced with the work. This resulted in a
process that now took much longer than it traditionally had because employees and their

703
 SIX SIGMA IN FINANCE

being increasingly routed to higher levels, adding to the overall cycle time.

The manager implemented a cross-functional performance challenge, modeled in part on

the CFPM method. Through that process, the team learned to handle issues and they were
also given the authority to do so. Empowering employees at appropriate levels to make
decisions is an innate part of Lean and Six Sigma methodologies because it can reduce
waste. In the Citibank case, process times were reduced to 30 minutes from an average of
two hours.

In 2000, dot-com fever in the United States and across the globe was still at a high pitch,
and financial institutions were funding all types of online projects. The industry began to see
a growing emergence of online banking, and some banks were opening via Internet only.
The thought process behind online banks was, and still is, that cutting out physical banking
locations results in substantial savings and those savings can be passed on to end users.

Online banks without branch locations faced a number of challenges in the early part of the
21st century. One of those challenges was how to accept customer deposits; before the
advent and wide adoption of mobile deposits, banks had to rely on mailed deposits. One
bank in particular made some decisions based on basic customer input and founder
intuition regarding deposits.

First, the bank decided that customers would be more comfortable mailing their deposits to
a local or regional address. Second, the bank decided that if a deposit was mailed to a local
address, it would take less time to process than if it were mailed to a central location. Based
on these assumptions, the bank created a process whereby customers mailed individual
deposits to a local or regional address. Every day, the local depositories would overnight all
deposits received to a central location.

The bank was surprised when customers complained that deposits were taking too long to
process, and a Six Sigma Black Belt was brought in. The Black Belt worked through the
DMAIC process to identify root causes for the problem, come up with a solution, and
implement it. Because the Black Belt relied on data and statistical analysis rather than some
basic feedback and intuition, he was able to determine that the local deposit structure was

704
 SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

not faster than a central deposit structure. Some problems he found with the local deposit
structure are included below.

Mailing, even locally, took up to three days on average, not including weekends.

not a daily process that could be engrained in staff. Because of this, when deposits
were
The local deposit process did not operate at a high sigma level.

The Six Sigma team performed data analysis and hypothesis testing. They also compared
the process for the bank to benchmark processes, finding that competitors were using the
central mailing process. Hypothesis testing showed that central mailing processes were
statistically faster than local mailing processes when the measurement was time to deposit
funds once they were mailed by a customer. The team also conducted customer surveys
and found that customers were not, in fact, overly worried about what address payments
would be mailed to.

The result of the DMAIC process was the implementation of a central mailing location and
the savings of $4 million annually for the bank.13

To be successful with Six Sigma in finance, teams and Six Sigma experts must:

Understand all of the customers of the process so that positive changes for one
customer do not result in unacceptable changes for another customer
Take time to develop a deep understanding of regulatory requirements, particularly
when working for publicly traded entities or dealing with processes that handle
personally identifying information
Work to create a culture of process improvement and an understanding of Six
Sigma methods throughout an organization.

13
https://www.isixsigma.com/new-to-six-sigma/dmaic/bank-deposits-black-belt-case-
study/
705
Hospitality is a broad industry that is sometimes considered to include any entity that deals
with lodging, food, beverage, and entertainment. Food and beverage is considered as its
own industry in a later unit. For the purposes of this chapter, the hospitality industry is
considered to cover the services and products that deal with tourism and certain services
related to travel and events. Examples of hospitality companies might be those that deal
with hotels or accommodations, travel arrangements, theme parks and attractions, cruises,
transportation for travel or leisure purposes, tourism, wedding planning, general event
planning, and festivals. Companies that support all of these services and products might
also fall within the hospitality industry but likely have a foot in another industry too. For
example, caterers definitely have hospitality-related concerns, but they also have to follow
guidelines and requirements for the food service industry.

Hospitality organizations typically walk a very fine line between productivity and customer
satisfaction, and where that line is depends on the type and level of service being
performed. Hotels that cater to business travelers, for example, must provide efficient, no-
hassle services with enough amenities to make the business traveler comfortable. Too many
amenities might price the service out of range for most business travelers; not enough
comfort, and the business traveler goes elsewhere. Luxury resorts walk a different line, as
do discount services that entice family vacationers with integrated pricing deals.

To appropriately walk the most profitable line, hospitality organizations must be plugged
into what their target customer needs and wants. Six Sigma helps organizations do that
through tools such as Voice of the Customer and statistical analysis. One reason customer
satisfaction is critical for hospitality organizations is that so many companies in this sector
rely on customer loyalty to bolster revenues. From hotels to theme parks, hospitality
organizations offer loyalty programs and rewards for a reason: they want repeat customers.
Hospitality organizations also rely heavily on social media and online reviews to ensure new
customers are more likely to choose their service or location. Employing Six Sigma and
706
 SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

statistical process control lets hospitality organizations create consistent quality service so
customer satisfaction wins can be repeated.

Like manufacturing organizations, hospitality companies have to keep a close eye on

inventory. Too much inventory increases costs and contributes to waste, neither of which
are good for profits. Too little inventory leaves the customer without, though, which can
result in poor customer satisfaction, loss of repeat business, and poor word of mouth
marketing. Recall the principle of Just in Time manufacturing, or JIT, from the chapter on

overstocked towels taking up all the space in his or her bathroom, but the customer does
want clean towels when they are needed. Using Six Sigma and Lean methods lets hospital
organizations meet customer needs in a timely, cost-effective manner.

Possibly the biggest challenge facing Six Sigma experts in the hospitality industry is the large
role people outside of the organization play and how much control those people seem to be
able to exert. Hospitality is a people-facing service, an
industry without a lot of people coming through your door or making use of your service.

which means there is always an inherent unknown in a hospitality process.

This is true in almost any industry to a certain degree. Certainly in a healthcare

environment, the patient is at least partially responsible for outcomes in some processes. If
arge instructions, then the outcome of

billing can be impacted. In many other industries, including healthcare, the staff within the
industry command a greater authority by weight of the situation, their credentials, or the
response ingrained in culture. In general, people are more likely to listen and respond to
instructions provided by a doctor than they are by a theme-park ticket taker, for example.
When designing or improving processes, Six Sigma experts must keep the fact that people
are so involved in the process in mind.

Another challenge that Six Sigma experts might face in the hospitality industry is that so
many processes are seasonal. While successful Six Sigma projects can actually reduce the
woes associated with stocking, staffing, and running a seasonal venture, the fact that teams

707
 SIX SIGMA IN HOSPITALITY

must think seasonally in scope can be difficult if Six Sigma experts are used to applying the
methodology in more long-running processes.

Turnover rates in hospitality organizations, and the level at which many of these
organizations hire, can present another challenge for Six Sigma experts. First, with high
turnover rates at all levels of the industry, it can be difficult to get enough buy-in to support
Six Sigma initiatives. Once buy-in at leadership levels does occur, a changing of the
executive or middle-management guards can impact that buy-in, leaving Six Sigma experts
within the organization constantly working to educate leaders about the importance and
efficacy of the methodology. Second, constant change in the lower ranks of employees can
mean an ongoing need to retrain. This is actually something Six Sigma and Lean
methodology can address, making it easier to control processes despite the people
involved, but attrition is something Six Sigma experts and teams should always be aware of.

The Starwood Hotels and Resorts chain was reportedly one of the first hotel organizations
to embrace the change-management and continuous process improvement methodology of
Six Sigma, and it was certainly one of the first to do so at enterprise and global levels.
Among other brands, Starwood owned both the Westin and Sheraton hotel names, and the
organization wanted to distinguish those chains from others on the market. Beginning in the
early part of this century, Starwood harnessed Six Sigma tools to drive efficiencies and
improvements
and hospitality elements.14

The need to protect brand creativity was an important factor in how the hotel company
chose to implement Six Sigma and a reason many people in the industry were surprised
that Starwood chose Six Sigma over what were considered more traditional approaches in
hospitality. In the last part of the 20th century, Six Sigma garnered a reputation for being

of manufacturing believed that Six Sigma and Lean methods were about making everything
the same to increase productivity, which they believed would naturally decrease creativity.

14
http://www.bloomberg.com/news/articles/2007-08-30/six-sigma-kick-starts-
starwoodbusinessweek-business-news-stock-market-and-financial-advice
708
 SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

that this is far from the truth. Six Sigma, when properly implemented, actually greases the
wheels for creativity by involving everyone in improvements and fostering out-of-the-box
eative
approaches with data to ensure that final decisions are more likely to work out as desired.

One of the reasons Starwood integrated Six Sigma into its processes was to increase its
ability to drive creativity across the organization and to repeat the results of successful
programs. One example of the Six Sigma approach used by Starwood can be seen when
reviewing a 2006 initiative called Unwind.

Unwind was an initiative Starwood launched to engage travelers within its upscale hotels.
While the project itself was definitely creative in nature, the Starwood team chose it after
reviewing industry trends, customer needs, and other data. Specifically, an industry study
indicated that travelers often feel lonely because they are away from friends and family at
home. Approximately a third of travelers expressed this feeling, and Starwood saw a good
opportunity to align its own desires with a real customer need.

As previously stated, customer loyalty and engagement is critical in the hospitality industry,
and Starwood believed that it could grow customer loyalty for its brand by providing
solutions to these feelings of loneliness. The organization used Six Sigma to come up with
and deploy solutions l.
Teams properly using Six Sigma recognized that every hotel all in disparate locations
across the country serve customers with very different desires and needs. What works for
Chicago, and
it was about
making everything better.

As part of its Six Sigma approach, Starwood trained hundreds of black and green belts.
These Six Sigma experts led teams across the globe, constantly coming up with ideas for
improvements and new programs. The teams had to pitch ideas to company leadership, and
only those ideas that met certain criteria were green lighted, because resources even for a
large, upscale hotel chain are never unlimited.

offering for travelers. The suggestion came after a black belt, some green belts, and subject
matter experts in a Chicago hotel met to brainstorm. The fitness director suggested
massages, and everyone liked the idea. The Six Sigma team launched a project to pilot a
complimentary massage offering. The pilot was so successful, teams had to figure out a way
709
 SIX SIGMA IN HOSPITALITY

to solve a new problem: How could the hotels tactfully communicate that complimentary
massage time was limited? Customers wanted more time in the massage chairs!

Working through Six Sigma processes, the team addressed those issues and later launched
the program across many hotel locations. The project was a large success not only was
customer satisfaction and loyalty increased, but massage-based revenues went up 30
percent in locations where the hotels offered paid spa services. Apparently, customers
agreed with Starwood: a massage was just the thing a traveler needed.

hotel locations, the Starwood team designed, piloted, and launched 120 features to engage
customers and increase loyalty. Features ranged from added amenities, such as massages,
to entertainment such as Chinese water painting classes and fire dancing. Amenities and
entertainments were matched to the look, theme, feel, and purpose of various hotels so
that they best met the needs of the clients staying in those facilities.

reports. In one year alone, the company says it saw an additional $100 million in profits
thanks to new programs designed and implemented using Six Sigma.

To be successful with Six Sigma in the hospitality industry, teams and Six Sigma experts
must:

Be able to account for the uncertainty that people bring to every process and know
how to mitigate that uncertainty with Six Sigma and Lean tools
Understand the seasonal nature of the industry and take trends into account when
developing processes and projects
Be able to expertly convey Six Sigma principles in a manner that is relevant to
hospitality industry professionals.

710
As with accounting and finance, human resource departments or, at least, human
resource functions are present in any company that has more than one employee. Six
Sigma is a valuable tool for improving human resource departments and functions as well as
improving human resource companies. For th

Sigma plays a role in organizations that provide human resource services as a product.
Those companies include recruiters, human resource consultants, HR software vendors and
developers, and payroll companies. You can already see that some of those companies span
multiple industries a software developer might have to concern itself with human
resource issues, but it also falls in the information technology field.

Sigma initiatives both within the department and within the enterprise as a whole. Like
finance departments, human resource departments are important partners for any Six
Sigma initiative. HR staff are typically the ones fulfilling open requests for new hires or
working with knowledge management on training programs, which makes them essential
for projects that involve adding staff or changing workflows and functions.

project themselves. If a project involves creating new training material, the core Six Sigma
. In some cases, the core team
as might be the case with
recruiting and hiring which is why involving auxiliary departments such as human
resources is necessary.

Within a human resource company or department, the common benefits of Six Sigma are
similar to those felt in any industry. Six Sigma can reduce costs and defects and increase
productivity and quality for the human resource organization. Within an enterprise,
improvements in human resource functions typically cause improvements across the
711
 SIX SIGMA IN HUMAN RESOURCES

organization think of human resources as the root of the plant that is the whole company.
t-
well or producing quality fruits. If the human resource functions are flawed or full of

results in the wrong team members on the job, poor onboarding and training, and poor
staff morale. Six Sigma methods can help human resource departments improve functions
to better serve the entire enterprise.

Modern human resource departments are also facing new challenges in recruiting and staff
management. The amount of data available to human resources departments today is
astounding, but without the ability to assimilate the data into something meaningful for
recruiters and hiring managers, HR employees are tied to traditional hiring methods. Six
Sigma can help HR organizations create data-rich processes that integrate information into
real-
sets. The results can be faster times to hire and better accuracy in filling open positions with
qualified applicants.

In a human resource company one that provides auxiliary services and products to other
organizations the Six Sigma DMADV process can be used to design and improve products
and services for end customers and remove waste and defects to make those services more
profitable. Dr. Mikel Harry of iSixSigma provides a number of other ways Six Sigma can
benefit an organization, including reduction in employee turnover, increasing employee
satisfaction, improving the management of benefits and payroll, managing healthcare costs,
gaining useful information from employee exit interviews, increasing compliance with
safety requirements, and deploying policies and training in a more comprehensive and
timely manner.15
can help with some of these areas.

Six Sigma faces the same challenges in human resources that it does elsewhere, including a
need to convince key stakeholders and sponsors of the need to make improvements and
the efficacy of the methodology. One challenge Six Sigma leaders face when working with

15
https://www.isixsigma.com/ask-dr-mikel-harry/ask-six-sigma-methodology/ask-dr-mikel-
harry-how-can-six-sigma-be-applied-human-resources/
712
 SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

ration with other areas and industries

has to do with the nature of the data and processes being addressed. Almost no human
resource function is without some access to confidential or sensitive information. When
dealing with Six Sigma projects in this envi
into contact with such information or that Six Sigma leaders will have to work around
restrictions on what information can be made available.

Often, organizations work around confidentiality issues by choosing team members who
can work with the data by virtue of their level or position in the company. While this
approach can work well if the team is led by a Six Sigma black belt, some precautions are
still necessary. First, Six Sigma teams must make sure everyone on the team understands
the importance of confidentiality. In some cases, organizations might even require team
members to sign confidentiality agreements especially if the Six Sigma initiative involves
consultants or vendors. Data must be safeguarded both during the project and after team
members can never use the information they see during a project for purposes outside of
the project. This is especially true of team members who might be in supervisor roles.

Data used during a Six Sigma project should never be used in a retaliatory manner, and
enterprises should be very careful when using such data as a disciplinary reason for an

Six Sigma human resource project discovers there is an issue, the Six Sigma team and

employee might be under the impression that everything is going well, and disciplining
someone out of the blue because of a Six Sigma project can create a negative culture
regarding Six Sigma in the future. Instead, Six Sigma teams and relevant supervisors should
work on retraining or correcting the process and this is true even when the project is not
in the HR realm.

A final human-resource specific challenge faced by Six Sigma teams is the fact that cutting
costs often means eliminating jobs. This can be true of any Six Sigma initiative and is one
reason why human resources should be involved in projects. Across industries, there is a
mistaken belief -
but that is actually far from the truth. Companies spend a great deal of time and resources
recruiting good employees, and letting good employees go can actually be a form of muda.

removing a position is not the same thing as removing a job. Often, Six Sigma improvements
make better use of the skills of existing staff, and that can mean a change in function or
position once improvements are made to a process. Human resource personnel and the Six
713
 SIX SIGMA IN HUMAN RESOURCES

Sigma experts working with them face the challenge of communicating and implementing
such changes in the most positive way possible.

In cases where process improvements do mean cutting jobs, then human resource experts
are challenged with helping teams conduct such activities in the most positive manner
possible. Some ways to mitigate the negative aspects of such an activity could include
helping employees find positions elsewhere in the company and ensuring employees are
provided with excellent references where applicable.

Payroll processing can often be improved in organizations of any type, and general
estimates are that organizations lose around 5 percent of revenues to payroll mistakes and
inefficient payroll processes. One company that provides document management services
to clients across the globe turned to Six Sigma to save money and time on payroll
processing.

The company, Océ Business Services, operates multiple locations that support thousands of
clients, and at the time of the Six Sigma project, the company paid approximately 5,000
employees based on hourly wages. The Six Sigma team, led by a Master Black Belt, included
subject matter experts from human resources, operations, IT, and payroll, as well as a
Green Belt. The goal for the team was to decrease costs and risks associated with payroll
errors or fraud. The team approached the goal via the DMAIC method.16

Define and Measure

The team spent the Define phase understanding the current payroll process and identifying
o big of a scope when you
are dealing with 5,000 hourly employees, and the team needed a more tangible and
measureable goal. Through the define process, the team learned that a single payroll
ing disparate methods
often manual in nature to calculate the number of hours worked by employees during

payroll application, which was proprietary and called I-Times.

16
https://www.isixsigma.com/implementation/case-studies/case-study-making-accurate-
payments/
714
 SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

are notoriously fraught with errors when compared to automated calculations no matter
how careful the managers are. Consistency was a problem too one manager might be
rounding up and one down while another entered time exactly to the minute or second.
And what about time off? How did each manager handle such situations? The opportunities
for error were many, and the managers were spending time on an inefficient manual
process. The Six Sigma team estimated a loss of approximately $1 million every year
because of the payroll process.

Moving into the measure phase of the project, the Six Sigma team was able to validate
many of their assumptions. The team surveyed 325 of the managers a plentiful sample
size to ensure statistically valid conclusions and found that only 15 percent of the sites
were using the propriety company payroll system for tracking, and less than a third of sites
even featured time clocks for tracking employee hours. Approximately a fourth of sites used
hand- s at
each location were, indeed, treating time off, tardies, and other attendance issues
differently. Managers were giving grace periods for tardiness that ranged from 0 to over 30
his mean the
company was paying for time not being worked, but it also meant that employees were
being treated differently in various locations and departments. Disparate treatment of
employees can lead to a range of human resource issues, including claims of discrimination
or favoritism.

Analyze and Improve

Using the data collected during the measure phase, the Six Sigma team identified two main
root causes for payroll errors. First, manual calculations led to errors in the number of hours
reported. Second, disparate methods for calculating and managing time sheets led to
inconsistencies in the process.

The team wanted to show that improvements to the process for capturing and managing
time worked would result in savings, so it started with a pilot program. Using the pilot as a
sample, the team estimated potential savings that would occur if the improvements went
live successfully across the organization. The data captured during the pilot indicated that
manual time management processes featured a defect rate of 0.5 percent. By reducing
those errors, the team felt it could save up to $1.2 million each year for the company.

715
 SIX SIGMA IN HUMAN RESOURCES

As part of the project, the Six Sigma team and relevant company subject matter experts
evaluated numerous options for capturing employee time worked. The team created a
checklist of capabilities it wanted in such a solution, and it settled on an ATAS, or automated
time and attendance system. The team implemented the ATAS, which included features
such as web and phone-based logins and rulesets to ensure employees were compliant with
schedules.

After implementing the solution, the team presented new data compared to the baseline
data. Prior to the change, the company paid an average of a bit over 80 hours per employee
per pay period. Simply switching to the more accurate system dropped that to
approximately 78.6 hours per pay period. In other words, the old time management
practice was off by a couple of hours each week on average for each employee the new
system also drastically reduced variation in reporting.

Control

Océ Business Services transitioned completely to the new system over a two-month period.
Throughout the transition, they saw increasing savings as the system was integrated across
more locations. The company put a control chart in place to monitor improvements, and
after about four months, the process was still working and estimated annual savings were
as much as $1.9 million.

To be successful with Six Sigma in human resources, teams and Six Sigma experts must:

Understand the confidentiality needs that might be involved with human resource
projects and related data
Be able to tactfully approach improvements that have a real impact on the jobs and
positions of other staff members
Understand overall company needs and how human resources support those needs
Be able to leave certain knowledge learned within the project environment and not
use it for specific supervisory or retaliation purposes

716
As with human resources, the information technology industry can be considered from two
perspectives. First, there are those companies that provide IT products and services. This
includes app development companies, tech startups, web development firms, and
companies that consult in IT fields or deliver software-as-a-service models. Second, the
information technology field also encompasses departments in many other industries that
handle database administration, networking, hardware and software support, and the
design and build of proprietary systems. IT employees are often involved in Six Sigma
s
computer-driven world projects are rarely completed without some need for specialized
data retrieval or changes to existing software or hardware.

While information technology does present some challenges to the Six Sigma expert, it is
also one of the fields where benefits are easiest to apply because information technology
often deals with either automated or people-powered workflow processes or the design
and implementation of a product. In the first case, Lean Six Sigma principles are easy to
apply; in the second case, DMADV approaches are helpful. In addition to cost savings,
defect reduction, and efficiency wins, Six Sigma helps IT organizations implement strong
change management programs and interface better with the business.

One of the biggest challenges that IT organizations of all types face is communication and
integration with the operations side of the business. The purpose of IT is to support the
needs of the business, but it can often feel like business employees and technical
employees are speaking two different languages. A common fail point for many products or
improvements is that the business explains a need, IT works to support what they
understand about that need, but a communication problem causes the mark to be missed,
even if only slightly. Six Sigma is actually a solution to this problem. By enveloping changes
717
 SIX SIGMA IN INFORMATION TECHNOLOGY

and product design in a Six Sigma methodology, organizations create an even playing field.
The DMAIC and DMADV approaches force teams to find common communication ground,
and when you are dealing with data and process maps, it can be easier to visualize what the
other side is talking about. Once organizations start using Six Sigma to approach specific
problems or needs, they often find that staff grow increasingly accustomed to
communicating in such a manner, and general communication between IT and the business
can be improved too.

technical requirements and abilities. When creating a new product or computer process, IT
employees can be tempted to add features as they work because they sound like a good

mud
cause a problem in the business arena. Because Six Sigma always starts with business needs,
it can help IT organizations reduce both muda and defects in its design and build processes.

Resources are often a challenge for IT departments and companies. Surveys of employers
consistently show a lack of skilled technical workers in many industries, which means IT
organizations often overwork the resources they do have. That can lead to burn out, high
attrition rates, or losses in productivity. Implementing a culture of Six Sigma helps teams
quickly implement solutions in a way that meets constantly changing business needs
without damaging IT resources for the future.

implemented Six Sigma before, they are likely to run into employees and departmental
gy is relevant to the industry. Measurement and

IT and software is measurable. In reality, though, IT functions are some of the most
measureable processes in the
ways that specific IT functions are measureable and why those measurements matter.

Six Sigma experts have to understand and communicate why IT measurements are
important both to the business and to the IT organization. For example, if a large automated
report process takes three hours to run and causes slowdown of a particular operating
system while running, this is a problem for both the business and the IT organization.
718
 SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

Measurements how much does the report really slow the system down? are critical and
of interest to everyone. Some questions Six Sigma answers that can benefit both the IT
organization and the business include:

Is the report really necessary to a business process, or can IT discontinue it and use
those resources elsewhere?
Is the slowdown to the system negligible or is it impacting productivity?
Can the report be run at another time, such as at night when no one is working?
If the report is run at night when no one is working, does that decrease the time it
takes for the report to run? This can be important if the IT organization wants to

day.
Why does the report take so long to run? Is the report pulling fields that are
unnecessary for the business use? Can the report be streamlined to create a faster
process?

This is just one example where measurements of all types can help the IT organization.

Another challenge Six Sigma experts might face when working with IT organizations is a
perception that there are never enough resources and a new project or initiative is simply
not possible. This challenge is especially hard to deal with at times because the perception is
not wrong. As previously stated, many IT organizations do struggle with resources, and it is
probably true that there are too many things to get done and never enough time. IT
employees might already put in numerous hours of overtime every week without clearing
the workflow backlog, so when you present them with what sounds like another

others in such an environment IT or otherwise must be adept at explaining Six Sigma

principles so that they do al work but do sound like a smarter way to
approach work, processes, and problems.

A final obstacle for Six Sigma in some IT organizations is the fact that Six Sigma often comes
after numerous other attempts at quality improvement programs and tools. In seeking a
fast, reliable way to implement IT solutions, some organizations have wound their way
through half a dozen or more methodologies, and by the time Six Sigma is broached as a
tool, the IT staff are tired of learning and integrating different programs. They might see Six

and that leads to failures that will make Six Sigma simply another thing organizational
leadership tried. Six Sigma experts dealing with such situations have to work hard to

719
 SIX SIGMA IN INFORMATION TECHNOLOGY

approach technical staff in a way that is unique and highlights the benefits of the
methodology. Starting with how Six Sigma is an ongoing methodology and not a quick-fix
tool is a good idea IT employees often know that there is no quick fix, and they are usually
happy to work within a methodology that takes that to heart.

A cellphone provider featured a top-up service through its online web portal. The service let
customers log on and add minutes to prepaid cellphone plans. The purpose of the service
was to offer a fast, easy way prepaid cellphone holders could manage their minutes and
plans. In reality, however, the service had about a 40 percent failure rate for top-ups and
about a 20 percent failure rate for registration. That means that a large percent of
individuals had to call customer service to seek help with either adding minutes or
registering for the online service.

The defects from the online top-up system created extra work for both the customer and
the customer service call center. That meant increased costs for the organization on top of
potential loss of revenue associated with customers that would simply not top up because
the system
Some customers might even seek another provider, especially if they dealt with the issue
more than once.

The cell phone provider created a team to address the problem. Following a DMAIC
methodology, the team stated a problem, defined processes, and began brainstorming
inputs and root causes for the problems. During the early parts of the DMAIC project, the
team listed nine core inputs it felt were related to the process and the issues. During the
Analyze phase, the team employed hypothesis and regression testing to determine which
inputs were actually driving fail rates for registration and top up.

The team originally thought that certain elements would have a strong correlation to top up

would correlate to future success with top ups. At a common-sense level, this seems
plausible: if the customer was able to top up last time he or she used the system, then the

the case history of successful top ups had no statistical correlation to the problem at
hand.

720
 SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

The team did note that some inputs it believed would not be related at all presented a high
statistical correlation to the failure rate of top ups. One of those was the value of the top up
varying dollar or minute values seemed to correlate to varying failure rates. Since the
system for topping up was t
would perform differently. This, along with several other correlations in the data, led the
team to look at the algorithms and internal rules of the system itself. The team came to
believe that the internally programmed process that rejected or accepted each transaction
was flawed.

During the Improve phase, the team created a new ruleset for the internal system. One
specific area they improved was how the system detected a fraudulent transaction. Prior to
the change, the system was rejecting legitimate transactions because they met encoded
criteria for possibly fraudulent transactions.

After addressing the fraudulent transaction process and other issues with the top-up
luding how IP addresses were blacklisted, the team launched an
updated process. The new process had a registration success rate of 91 percent compared
with the previous success rate of 80 percent. It also had a top-up success rate of 90 percent
compared to only 60 percent previously. Increased annual revenues from the new process
were estimated to be $300,000.

IT is likely to be involved in some part of the process. Most organizations involve IT

representatives in almost every Six Sigma project because all processes are somehow linked
to computers. One area where technical resources are likely to be required is in automated
control phases. While teams can conduct control methods including control charting

additional failure points. If someone has to manually measure, enter, and chart data, more
opportunities for error exist. IT departments often work with Six Sigma experts and process
owners to automate control process to mitigate risks of errors.

ResMed, a medical device company, uses such automation in its control process. One
example is how ResMed uses Minitab macros and other automation to manage control
charting of the process that manufacturers machines and masks. After using some Minitab

721
 SIX SIGMA IN INFORMATION TECHNOLOGY

solutions during improvement processes, the team decided to continue using Minitab for
control. The team also attended a workshop on leveraging Minitab for addition functions
that helped them come up with an automated process.

The production line is continually tested and that data is stored in a database. Every 30
minutes, Minitab automatically queries the database file and creates a control chart. The
control charts are automatically saved following a filename convention that lets the

dashboard for review by leadership and process employees. The SharePoint system itself
updated every 10 minutes, so the team always has access to control data that is no more
than 40 minutes old.

The automation lets the team identify and address problems quickly. If a process is out of
statistical control, the team knows within the same day, if not within a few hours, reducing
losses associated with defects or process problems. If the team relied on a business analyst
to pull and compile data to be presented in a control chart, it is likely that the team would
not gain such insight within the hour or, possibly, even within the day.

To be successful with Six Sigma in the information technology industry, teams and Six Sigma
experts must:

Be able to convey and illustrate the benefits of the program to engage IT resources
effectively.
Be well-versed in or able to quickly become fluent in technical functions and ideas
Work well with both technical and business resources and be able to bring together
teams that are comprised of both

722
Unlike the previous fields covered in this unit, engineering has a more formal definition. For
this chapter, engineering covers any organization or department that is engaged in
engineering work whether that work is civil, industrial, mechanical, or chemical in nature.
While engineering is sometimes part of an organization or project with bigger scope a
metal building manufacturer has an engineering department, for example engineering

work within the engineering space because the project at hand is engineering in nature and
not because they are meeting a compliance need or seeking assistance with a workflow.
That being said, when an engineering question comes up in a project that is not, at its heart,
an engineering project, Six Sigma teams are probably wise to consult an expert in the field.

Successful Six Sigma engineering projects reap all of the benefits that have been highlighted
throughout this book, including reduction in defects and costs. When property deployed, Six
Sigma can also improve the ability of engineers and teams to meet timelines and budgets
on projects, particularly in a civil engineering field.

Engineering projects usually involve many, many moving parts. Civil projects require both
design and build phases. During both phases, teams have to manage costs and resources,
meet compliance requirements from a variety of government and private agencies, meet
the demands of the client, and ensure safety and those are just some of the things that
have to be considered. The DMADV and DMAIC processes force teams to consider all
requirements, understand the project details and timelines, and identify challenges to be
worked through. On the surface, it might seem like DMADV or DMAIC would slow down the
overall engineering process. In reality, applying the same methodology to every project
actually speeds up processes without impacting quality.

Six Sigma also provides benefits with regard to safety. Engineering processes in all industry
subsets come with safety risks. Mistakes in civil engineering can cause safety risks to people

723
 SIX SIGMA IN ENGINEERING

working to build or remodel a structure; even if the structure is already built, if engineering
integrity is faulty, then dangers continue to exist for those using the structure in the future.
Industrial and chemical engineering come with other safety risks, especially to engineers
and employees working closely with machinery or chemical processes. Some fields closely
related to engineering, such as the oil field, feature some of the highest numbers of
workplace accidents in the United States, and statistics for other countries tend to be
similar. Implementing Lean Six Sigma tools helps create consistent quality throughout
engineering processes, and that quality improves safety at the process point and further
down the line.

Another benefit of Six Sigma in the engineering industry is that the methodology fits very
well with the mindset of most engineering teams. Engineers are trained to think by the
numbers, so they are less likely to balk at a data-driven approach than individuals in other
industries are. Engineers also usually appreciate a step-by-step approach that is used across
all projects because they understand the value of consistency. The appreciation for this type
of approach makes it somewhat easier for Six Sigma experts to work with engineering
personnel as a Six Sigma expert, you might have to fight fewer battles in support of your
methodology.

Even though Lean Six Sigma methodologies align well with engineering mindsets, Six Sigma
experts are not without challenges in implementing the method in this field. The challenges
you might face in working with engineering projects depend heavily on the culture of the
organization in which you are working and the mindset of the subject matter experts.

When working with any group that is highly educated, you might run into the problem of
expertise. This is true of almost any specialist industry where individuals have achieved
years of school and special certifications. You might find this problem in healthcare with
physicians, in legal fields with lawyers, or in financial fields with CPAs, for example. The
problem of expertise occurs because someone feels they know something because they

or review statistical analysis. You might

724
 SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

In such cases, the Six Sigma expert must do two things. First, the Six Sigma expert must see
that someone with that much experience and education has a lot of knowledge and insight

appropriate in all situations. If you remember from Unit 1, Six Sigma is not the right
approach when the solution is already known. However, as a Six Sigma expert, you need to
be able to tell when a solution actually is known and when someone only believes they
know the best solution.

Second, the Six Sigma expert must work tactfully to convince everyone involved that there
is value in the process. The typical fix to a problem might work and work well, but with the
collective experience of the group and an outside-the-box mentality, the team might
discover a way that is even better.

In addition to challenges that come from those within an organization, Six Sigma experts
who are not also engineering experts might face internal challenges when working in this
industry. Any process in an engineering field is going to deal heavily in technical ideas

they deal with

field, for example. However, you do have to be able to understand processes and how they
flow, and in the engineering field that could require the ability to quickly learn at least some
basics about the science or math behind the process.

Six Sigma experts working in engineering fields must also understand that all engineering is
not about building structures such as bridges or buildings. Those types of projects are
spec
studies in this unit, is often much more tied to manufacturing processes, and chemical
engineering is yet another very different field.

Often, Six Sigma projects are conducted by subject matter experts within an organization
and Six Sigma experts from a consulting company. In one such case, a consulting firm called
Product Modeling Corporation, or PMC, worked with an OEM automotive parts
manufacturer. The OEM hired the consulting company to assist with improving and
controlling a process in a plant that stamped and assembled some parts. PMC used Six
Sigma and industrial engineering knowledge during the improvement process.

725
 SIX SIGMA IN ENGINEERING

As with all Six Sigma projects, the team first worked to understand the existing process
within the large manufacturing plant. The plant included 2.5 million square feet of space
and 23 press lines. Stamping lines ran linear, usually beginning on one side of the plant floor
and ending on the other a traditional setup for assembly or plant line production. Raw
goods were stored in racks or containers and dollys, forklifts, and people were used to
transport materials throughout the lines.

During the Define, Measure, and Analyze phases, the PMC team noted several opportunities

inadequate, as was the reporting system for how equipment was used and maintained. Too
many forklift operators on some shifts also represented muda.

PMC worked with plant subject matter experts to develop both static and dynamic
simulation models that helped predict the needs of the plant and the flow of material
through the plant. Engineering analysis helped make these simulations as accurate as
possible so that business decisions could be made based on the simulators. The plant
wanted to be able to use such solutions to make ongoing decisions and changes to resource
allocations.

PMC also worked with the plant to improve bar coding processes. Bar codes were essential
for management of both resources and automated industrial processes. Bar codes let
machines identify parts and work appropriately; they also let the plant better manage
inventory.

The simulations and bar code improvements were tested before being rolled out to the
entire plant. The result was a total of $5.6 million in savings annually.

The same consulting firm, PMC, worked with a company that manufactured curtain walls,
doors, and skylights. At the time PMC was brought in to facilitate improvements, the client
company was working with three different processes for each product. While the
engineering processes were somewhat similar, the client was using different software to
design and manage the manufacture of each product. Disparate systems not only opened
the door for more possible defects, but they also increased resource use and expense. PMC
was tasked with reducing muda and improvement the processes in part by assisting with

726
 SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

the implementation of singular CAD and PLM systems for all the engineering and
manufacturing processes. (CAD and PLM are both engineering software tools.)

PMC first conducted audits of all the sites involved in the three processes. In conjunction
with client subject matter experts, PMC identified the requirements of all the processes a
single engineering solution would have to meet all the requirements of all the processes.
Standards for drawing were developed and the team converted AutoCAD libraries so that all
the data and functionality matched. Some of the processes were using 2D libraries. Those
were converted to 3D libraries so that the necessary functionality of all processes could be
handled with a single format.

In addition to the implementation of consistent software tools for all processes, PMC made
recommendations that would help to control quality and consistency going forward. The
team recommended implementing change management controls so that future changes in
one process would be rolled out appropriately across all processes and locations. To
support these changes, the facility followed a six-step assessment and upgrade method that
included upgrading software and engineering workstations and adding version control
components to both drawing and engineering processes.

Ultimately, a single CAD software solution was implemented across all the processes and
locations. The CAD solution draws from the same pool of rulesets and engineering
knowledge regardless of location, which reduces consistency issues and eliminates muda
associated with extra knowledge resources. The CAD solution feeds directly to the PLM
software, mitigating risks of data disparity or mistakes and reducing manual work between
the two processes.

The benefits to the organization included the elimination of rework at design; that helped to
increase bidding efficiency so the organization could make more accurate and timely bids.
Use of the same resources across all processes reduced overall costs, increasing profits, and
version control helped eliminate mistakes and misunderstandings stemming from processes
using outdate versions of drawings or rulesets.

To be successful with Six Sigma in the engineering industry, teams and Six Sigma experts
must:

Develop a deep understanding of the compliance and safety requirements

associated with any project

727
 SIX SIGMA IN ENGINEERING

Be able to encourage and work with highly educated and experienced subject
matter experts who might not immediately buy into the methodology
Be able to work with multiple complex and technical ideas and integrate new
knowledge quickly
Understand how engineering plays a role in various types of projects, particularly
with regard to different engineering fields, and not take a one-sized-fits-all
approach to various initiatives outside of the consistency of Six Sigma

728
Field services are those that rely on a person or team going into the field to perform the job.
In some ways, delivery companies such as UPS or FedEx might be categorized as field
service companies,
companies that perform services such as HVAC repair, plumbing, electrical maintenance, or
cable or phone hookup. Business-to-business services are also included in this sector, such
as businesses that supply and maintain copy machines. For all of these businesses, simply
getting to the right place at the right time is a big priority, so they will share some Six Sigma
benefits and challenges with the shipping industry. However, field service employees also
perform processes once they arrive on site.

Most field service organizations include a few common processes or departments. First,
someone has to take customer calls and dispatch field service technicians appropriately.
This might involve managing work orders, scheduling field calls, and routing drivers. Most
field service organizations also have to manage inventory and parts, ensuring that the right
tools and parts for the job make it to each location. Drivers or teams must then find the
service location, assess the problem or need, and offer recommendations or a quote for
service. Finally, if a quote is accepted, the field service representatives provide the service.
Once service is provided, the organization must capture the data and use it to manage any
invoicing processes.

729
 SIX SIGMA IN FIELD SERVICE

Given the above definition of a field service organization and what you now know about Six
to such
companies. Six Sigma can help field service teams boost relative revenue the amount of
money that is made for the time a team or field service representative spends in the field.
Remember the lessons on Lean process management in every process, there is value
added and non-value added time. In some field service organizations, value added time
might be considered time you can bill for. Non-
bill for.

Consider a plumbing company that charges a set amount for some basic service calls but
charges hourly for major plumbing jobs such as installations or lengthy repairs. The clock for
the billing starts when the plumber starts work and ends when the work is completed, but it
for time the plumber spent taking business calls that
were not related to the job. If the field service rep has to leave the job and return because
he or she forgot a tool or supply through no fault of the client then the company
e that travel in the billable time. Applying Six Sigma to these processes
can help the plumbing company eliminate the non-value added time so that more of the

The same benefit is even more applicable when a company charges a flat rate. A company

staff spending any more time on the job than is necessary for excellent customer service.

Organization is a key factor for success in field service companies, which often deal with
numerous moving parts at the same time. People, supplies, and tools all have to be routed
appropriately. Not all jobs are the same, so the company needs a process by which it can
efficiently and accurately assess job needs for planning and dispatching purposes. If a single
rep is dispatched to handle three jobs in one day, but the first job ends up taking 7 hours,
then the last customer might not be served that day. Likewise, if a level one rep is
dispatched to handle a level three job, there might be a problem because skills and

organizations create processes that minimize such mistakes.

Another area where Six Sigma can benefit field service companies is in the quoting and
assessment processes. Poor quoting processes lead to miscommunications, errors in cost
730
 SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

information provided to the customer, loss for the company, or a dissatisfied client.
Without processes in place that remove error and create consistency, representatives might
quote varying prices to clients, leading to poor word of mouth or the belief that some
customers are being unfairly treated. Someone might under quote a specific job because
they forget to include necessary supplies or labor in the calculation. This is especially true
when field representatives provide same-day quotes while on the job site, because they are
often performing these tasks in a short amount of time and with the potential customer
looking on and interrupting with questions. While field service companies usually make it
clear that a quote is not a binding agreement regarding cost, consistently being too far off
ing templates and
poka yokes, can help reduce the chance of these issues arising.

Field service companies can benefit greatly from Six Sigma, particularly when it comes to
customer service, increasing revenue, and decreasing costs. However, implementing Six
Sigma in such an environment is not without challenges. One of the first challenges depends
on the type of field service company involved. Field service companies are often somewhat
unique because they hire skilled experts in specific niches. Just as individuals with high

technical or vocational staff can balk at the concept that a method involving statistics and
brainstorming will help them repair a plumbing problem or fix a business copy machine
quicker and better.

A particular challenge of the Six Sigma expert or leader in this field is to meet the
communication needs of all the people involved. Administrative and corporate managers in
field services might respond to business-speak, but technicians and many field service
representatives will not. They might have been told too many times how they should do
their jobs by people who do not have the skills they do. In a large HVAC service company,
managers might constantly tell service technicians to complete work faster or provide tips

what the service representative faces on the job ultimately, a good field service technician
knows his or her job and wants the ability to use experience on the site to provide a better
fix or better customer service.

731
 SIX SIGMA IN FIELD SERVICE

In such situations, Six Sigma experts must do a good job at communicating the ultimate
freedoms that the methodology affords service representatives. Remember, a Six Sigma
process reduces defects in part by making it possible for individuals to address specific
problems during production. That concept is usually explained with regard to a
manufacturing line the defect is addressed before it reaches the end of the line and a
quality assurance check point. But the same concept is relevant in field services the field
ce for every single decision, and he or

time, field service represent

that makes decision-making similar across the company. Six Sigma helps teams build that
framework, empowering field service technicians to make decisions on site without
impeding overall company performance or consistency.

Another challenge you might face when implementing Six Sigma in field service companies
especially if you work as a consultant with numerous field service organizations is that
each company has its own unique requirements, processes, and goals. Yes, all field service
organizations ultimately want to provide quality customer service while making a profit. But
what works for an insurance adjustor team is not going to work for a team that paints
houses or fixes roofs. As a Six S
DMAIC methodology or to understand how to apply statistical analysis. You have to be able
to quickly understand the basic tenets of the companies you are working with so you can
provide active ad
mean you have to become an expert in plumbing or foundation repair, but you do have to
know what makes plumbing different from foundation repair or any other field service
niche if you are working with a plumbing process.

Route management is a common trouble spot for many field service companies. Companies
need processes that quickly route drivers and teams to job sites without muda which
-
arriving at job sites without the information that lets them be prepared to perform
appropriate work. Many times, the introduction of automated route-generation software

throughout this book, there is always room for more improvement.

732
 SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

One company learned this lesson and applied a DMAIC project to increase improvements
that were originally experienced from an automated routing solution. The company in
question provides business-to-business services that involve collecting and processing both
non-hazardous and hazardous materials from automotive and industrial clients. The
corporation handles approximately 45,000 service calls every week throughout the United
States, and the work is managed by regional or local service branches. Each service branch
has a manager who dispatches work crews each morning.

The company wanted to improve both user satisfaction and the stability of its work system,
so it implemented a third-party transportation management system (TMS). The TMS was
loaded with data about routes, work requirements, and customers. Each day, work order
information is entered into the TMS and the system generates a dispatch list for each team.
The goal of the process is to create a list that can be handled by each team in a single day
while maximizing value-added time and minimizing drive time. To do this, the TMS
considers best routes to and from customer locations, the amount of work the customer is
requesting, and the capacity of the trucks driven by each team.

While the TMS works to create many efficiencies, the automated system alone cannot
account for certain factors. For example, if a customer requires work on three machines,
then the TMS allocates the time it takes for all three jobs to be completed on site. It

three machines being located far apart from each other. When something like that occurs,
the team takes longer than expected at the job site, making every other service call that day
late. In some cases, there might not be enough time to handle all of the scheduled service
calls.

As a result, the automatic routes were being reviewed by management to determine if they
were actually feasible. They were then being provided to service workers. During the define
and measure phases, the Six Sigma team noted a bottleneck in the process. The service
workers were making suggestions to alter the routes for better performance, but that
required the manager to go back and make changes. This slowed the entire process down,
created additional work for everyone, and created more opportunities for defects.

The Six Sigma team also noted some inconsistencies with TMS estimates of travel time and
distance. Investigation during the measure phase showed that the TMS could not always
parse an address fully, and it applied a certain type of geocode in these situations. The Six
Sigma team noted that the geocode could be used as a measurement to determine whether
the TMS was likely to be accurate about time and distance details.

733
 SIX SIGMA IN FIELD SERVICE

Ultimately, numerous changes were made to the routing process. First, the geocode scoring
was made visible in the TMS application and reporting, which let branch managers see
immediately where human analysis might be necessary in improving a route. The process
was also altered to offer a chance for service representatives to provide feedback on routes.
Service representatives are the direct link to the client, and they are the people who have
most recently been on a client site. They know how big the sites are, where equipment is
likely to be located, and whether any abnormal challenges might be experienced on site. By
providing feedback about route feasibility early in the process, service representatives avoid
rework or scheduling problems later in the day.

The outcome of the Six Sigma project was an increase in both user satisfaction and return
on investment above the increases already experienced by implementing the TMS to begin
with.

To be successful with Six Sigma in field service environments, teams and Six Sigma experts
must:

Be able to communicate and educate about Six Sigma in ways that resonate with
both business and administrative staff and technical and vocational staff.
Take time to understand the unique challenges experienced by the specific
business.
Understand the tools and requirements of the specific business.

734
Like many of the niches covered in these last few units, sales and marketing are both their
own niche and a very necessary part of almost every organization. Without sales and
marketing, almost no brand or business can succeed, though the way in which organizations
go about these critical functions differs broadly. While sales and marketing are often
integrated, they are two different slices of the same pie.

Sales functions are often associated with the act of selling. This is the direct act of selling
goods or services to a customer or client, and it usually involves time spent engaging with
the customer face-to-face, via telephone, or online. The goal of sales functions is typically to
close the deal to convince the customer to buy the product or the service.

Marketing functions, on the other hand, are often more general in nature and are about
generating prospects, or leads, for sales teams. Marketing includes functions and campaigns
to reach potential customers and convince them that the product or service is of interest. At
a car dealership, for example, the marketing work might be what brings a new customer
onto the lot to look at a potential vehicle. The sales person then takes over to work directly
with the new customer to ensure a deal is struck.

technically the niches work closely together while following some basic boundaries as
stated above. What is usually called the sales cycle, however, involves both marketing and
et resources
have also combined to create more integrated sales and marketing approaches in some
organizations.

The general benefits of Six Sigma in marketing and sales organizations are similar to those
experienced in any organizations that deploys the methodology. Increased production and
revenue and decreased defects and cost are all benefits that are applicable to sales

735
 SIX SIGMA IN MARKETING AND SALES

and marketing departments and processes. The customer-centric nature of Six Sigma
actually makes it a great methodology for supporting improvements in these two areas. The
Voice of the Customer tools and the fact that any good Six Sigma project uses those tools
for a foundation -- help provide critical success to functions that rely on customer response
as a success measure.

Remember, the Voice of the Customer tools help Six Sigma teams understand what the
customer needs and wants. Advanced analysis of VoC information can also help teams
understand where customer needs are different than customer wants a very important

then deliver needs, but speaking only to needs without understanding wants might not
result in a good conversion rate on marketing campaigns. By implementing VoC and critical-
to-quality measurements, Six Sigma teams can help marketing organizations better
understand all aspects of the target audience.

On the sales side of an organization, Six Sigma can help teams convert more prospects into
customers, retain more customers, and even win back customers who were previously
in the past, then
you might see sales as a one-and-done proposition. The sales person engages the customer
for a short time, closes the deal, and moves on to the next opportunity. Outside of some
non-specialty retail establishments, modern sales functions
Sales has become a more complex process that often involves forming lasting client
relationships, providing essential client education, and understanding client values and
needs to better make product and service recommendations. For many sales teams, long-

Because of the long-term nature of the modern sales relationship, sales organizations must
present a consistent, quality approach. Clients who are offered varying information from
the same sales organization or who deal with inefficient communication and processes

and the ability to continuously improve processes help sales organizations avoid these
troubles.

Another reason Six Sigma has become a go-to methodology for sales and marketing
organizations is that it offers proven results without impacting flexibility and creativity.
Sales and marketing organizations have to be able to move quickly to deploy new strategies,
particularly in an environment imbued with connectivity and technology. The Internet offers
many options for connecting and engaging with potential clients, but online marketing
736
 SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

trends evolve at a breakneck pace. Six Sigma lets companies develop processes that can
keep up with trend evolution, and DMAIC projects ensure organizations validate all
assumptions before making major changes.

While Six Sigma is being embraced by an increasing number of executives in the sales and
marketing fields, Six Sigma experts do still face some challenges when implementing
projects in these areas. One of the most common obstacles to Six Sigma success in sales and
marketing organizations is related to a misconception about how processes within such
departments can be measured. Obviously, sales can be measured in the number of
transactions that occur, but the impact of a marketing campaign is harder to understand. Six
Sigma experts must work with sales and marketing employees to understand how success
can be measured and assure those involved that measurement is possible.

appropriately measured, but they might have to convince someone that the current

to evolve to customer demands, they are also often known for a strict adherence to metrics

Consider an American electronics and appliance retail chain that measures sales success
based on a variety of factors, including conversion rates. The retailer arrives at an estimated
conversion rate by creating a rat
door and the number of transactions that occur during the day. Sales staff were held
accountable for this ratio and sales manager bonuses were somewhat impacted by this
measurement. Without any other information about the scenario, you can likely apply the
knowledge gained from this book to see some obvious flaws with the door count metric.
What if someone went in and out of the store multiple times? They would be counted more
than once at the door, but would likely only make one sales transaction if they chose to buy.
What if children stood near the automatic door, causing it to open or close multiple times?

are just some of the scenarios that might throw off the door count metric. Six Sigma experts
working with sales organizations might have to work diligently to illustrate why some

737
 SIX SIGMA IN MARKETING AND SALES

Another challenge Six Sigma experts face in sales and marketing is the perception that

might argue that their work is about building relationships, and putting specific processes in
place would limit their ability to do so. Sales staff, in particular, sometimes hold tightly to
the perception that structure inhibits their ability to sell. In reality, though, some of the
most successful sales people are planners they arrive early to structure their day, follow
plans they set for each interaction, and always return calls and messages immediately.
Successful sales people are usually already following a process Six Sigma just helps engage
the entire organization in processes that work.

A and
illustrating how much process already exists in sales and marketing organizations. Writing
and posting blog posts is a process; maintaining social media presence is a process made up
of many smaller processes. Calling a customer is a process. Receiving a lead from marketing
is a process. Signing a contract or closing the deal is a process. Putting in the right
information to invoice the customer is a process. Once sales and marketing professionals
understand that process already exists in their organization
to flood their day with additional processes that make it more difficult to engage with
customers they are more likely to be open to continuous process improvement.

Six Sigma experts applied the DMADV approach to marketing and advertising campaigns
associated with the Corporation of the City of Kawartha Lakes, which is a location in
Ontario, Canada. Remember that DMADV is a phased-approach to Six Sigma projects similar
to DMAIC. DMAIC is used to address existing processes that are in need of improvement
and include five phases: Define, Measure, Analyze, Improve, and Control. DMADV is usually
employed when a new process needs to be created or a process is being overhauled
completely. The five phases of DMADV are: Define, Measure, Analyze, Design, and Verify.

During the define and measure phases of the project, the Six Sigma team worked to
underst
upon. The team came to understand the following information about the location and its
advertising requirements.

The city and surrounding area includes 250 lakes, which leads to numerous
recreational and tourist activities.
The area includes around 73,000 residents who live mostly in rural environments
and around 31,000 additional temporary residents or tourists in summer months.
738
 SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

Tourism and activities play a large role in the area economy.

Advertising events and branding the area is critical to the success of those events
and to the overall economy.
The full-time residents also rely on advertising in the form of brochures, web copy,
newspaper adds, and mailers to understand options for entertainment and events
in the area.

The Six Sigma team discovered two major issues during the first phases of its DMADV
project. First, it discovered that no one had ever taken time to understand how the
residents of the area preferred to receive advertising information about the area and its
events. Second, each city and activity organization handled its own advertising processes,
creating confusion, disparity, and additional costs for the city.

To better understand how each department or facility managed advertisements, the Six
Sigma team worked with each area to track that information. Specifically, the team wanted
to understand what staff were involved in the process and how much time was spent on
these tasks. The team used the information gained to create a process map of the current
process commonly used by the organizations, and it found that the process was imbued
with both financial and time muda. As many as 56 percent of the process steps required to
play an advertisement about events or programs required the action of senior
management. That meant that someone at the city level, or a director or manager at the
department or organization level, had to approve or take action, in almost every campaign
and for many of the steps within that campaign. According to statistical analysis from the
Six Sigma team, the city was spending up to $1,198 per hour worked on a campaign that
could be cut with more efficient processes.

During the analyze portion of the DMADV project, the Six Sigma team illustrated an
important point covered several times in this book: assumptions are not always correct. The
population of the city in question featured 39 percent over the age of 55, and this led the
city departments to believe that print advertisements would be preferred. Analysis of data
gathered via focus groups, surveys, and one-on-one interviews indicated otherwise. In fact,
the analysis showed that 75 percent of individuals wanted to receive information via email,
59 percent wanted to see information on the
internet searches to find information. Only 41 percent wanted the information in print
format.

Because the information generated by the analysis was so different from the long-term
assumptions, it was questioned. There were feelings that the data was not representative of

739
 SIX SIGMA IN MARKETING AND SALES

the city a good illustration of the challenges mentioned above relating to the acceptance
of new data when sales and marketing staff have always relied on certain assumptions or
metrics. The Six Sigma team alleviated these worries by performing a Chi-square goodness-
of-fit test, which showed that the sample was a statistical representation of the population
concerned.

The team also used the focus-group and other VoC data to understand what was important
to residents who received event and location advertising. Specifically, customers wanted to
receive information in a timely and immediate manner, wanted information to be accurate,
wanted access to both specific and general information, and wanted to be presented with
information in an accessible and aesthetically pleasing manner. Using all of this information,
the team entered the design phase.

Leveraging all of the VoC data compiled, the team created a requirements document for the
new advertising process. Following that document, the team drafted a process map for a
more efficient process that better met customer needs and allowed supervisors to make
some advertising decisions, reducing the costs associated with constantly involving senior
management. The define phase involved implementation of standard operating procedures,
templates, creative briefs, and branding guidelines, all of which increased efficiency and
removed disparity between advertisements.

Verification of the changes illustrated they were successful. The immediate savings were
approximately 15 percent, or $90,000 annualized, saved on supplies and other costs. The
city also saw an increase of approximately 20 percent in production and efficiency, leading
to an additional cost savings estimated at $53,000 per year.

To be successful with Six Sigma in marketing and sales environments, teams and Six Sigma
experts must:

Be able to expertly apply Six Sigma tools such as the Voice of the Customer and CTQ
trees.
Be able to educate sales and marketing staff about the fundamentals of process and
apply those lessons to the individual organizations and teams.
Understand sales and marketing metrics.
Be able to apply statistical analysis to validate metrics, create new metrics, and
measure success with a high confidence level.

740
 SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

Links:
https://www.isixsigma.com/operations/marketing-and-sales/guidelines-making-lean-six-
sigma-work-sales/
https://www.isixsigma.com/operations/marketing-and-sales/10-challenges-overcome-
when-deploying-lean-six-sigma-pharmaceutical-sales-and-marketing/
https://www.isixsigma.com/operations/marketing-and-sales/
https://www.isixsigma.com/methodology/design-for-six-sigma-dfss-methodology/case-
study-a-dmadv-approach-to-marketing-and-advertising/
https://www.isixsigma.com/operations/marketing-and-sales/customer-winback-concept-
begs-use-six-sigma/
http://www.sellingpower.com/content/article/?a=1751/how-to-apply-six-sigma-to-sales

741
In most organizations, customer service departments are adjacent to sales and marketing,
and overlaps often occur. Some companies combine sales and customer service within the
same teams sometimes even within the same functions. Investopedia defines customer
service as any process that works to ensure customer satisfaction with a service or product.
Commonly, these processes involve interaction with the customer either face-to-face,
over the phone, via mail or email correspondence, or via the Internet. In fact, the Internet
has drastically changed how organizations provide customer service today, allowing for the
evolution from mostly in-person and phone support to a mix of digital support options.

Every organization that provides a service or product to any other person or organization
must consider at least some elements of customer service. Customer service even occurs
within organizations, as departments serve employees in other areas.

While important in every industry, customer satisfaction as a driver does rank differently in
each niche. In most retail environments, for example, customers have many options for
seeking service and products. Someone shopping for groceries can usually travel just a little

service. Even individuals shopping for specialty items can turn to online stores for products

especially when all other things, such as product quality and price, are equal.

For other organizations, customer choice is not as prevalent. If you are taking advantage of
government services, for example, there is very rarely a competitor to which you can turn if
you are unhappy with those services. While go

true in part because of the type of service being performed. Consider a hospital: many
patients do have some choice about where they receive their medical treatment, even if
those choices are limited by geography and payer plans. That makes customer service a high
priority for medical facilities but good customer service will never supersede priorities
such as patient safety or outcome. Emergency department staff, for example, are more

742
 SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

likely to act quickly and present terse explanations to patients or family because, in a
trauma situation, time may be of huge importance.

the function of customer service with the Six Sigma idea of the

any process and that Six Sigma teams should always be aware of both critical to customer
and critical to quality characteristics within a process. However, meeting critical to quality

emergency room, the CTQ is keeping a patient alive; secondary metrics might include
patient satisfaction and comfort. But a good medical team will generally sacrifice patient
comfort if it is required for a viable treatment. Remember, an important factor in applying
Six Sigma and its customer quality methods to any process is first understanding both who
the customer actually is and how quality is and should be defined for the process.

organizations and processes. For the purpose of limiting scale and providing appropriate
examples, this chapter will concentrate on organizations and processes that would naturally
put a high priority on customer service, including service, retail, and call center companies.

One of the biggest benefits that Six Sigma can bring to any customer service organization is
that it opens the doors for teams to concentrate clearly on customer outcomes. Yes, any
organization that identifies itself as a customer service department should already be
concentrating on the customer, but companies often fail to do so with the right mindset,
perspective, and data.

In the last chapter, you saw how a well-meaning marketing organization sought to provide
the market with a certain product based on what it assumed was a customer preference for
hard copy information. The Canadian city, however, later found out that its assumptions
were incorrect, and thus it was potentially wasting time and money on customer service
endeavors that did not best serve the customer. Six Sigma helps organizations pinpoint true
customer demands and desires and then provide the products or services that best meet
that criteria. The results of such changes usually include additional sales and the ability to
demand more money because the product is more valuable to the end customer.

Six Sigma also helps customer service organizations dig deeper into metrics for better data-
backed decision-making. As with marketing and sales organizations, customer service
organizations often already have metrics in place. Common metrics might include the
743
 SIX SIGMA IN CUSTOMER SERVICE

average time it takes someone to answer or handle a telephone call, the number of
customers served in a given period, or a customer satisfaction score calculated from surveys
and feedback forms. While these are all fine metrics for a customer service organization to
be concerned with, you know by now that they might not be the right metrics by which
specific process success should be measured.

For customer service organizations, continuous improvement related to Six Sigma can result
in increased customer loyalty. Statistically, many customers leave an organization because

what seem like more obvious factors. This is why so many organizations rush to reduce
prices or compete with others on what type of price or services are offered. But they might
be doing so at great expense to the organization without favorably impacting customer
loyalty. Yes, prices and products are important. If your prices are simply outrageously high,
most people cannot make a case for remaining loyal. Prices and products also help you get
customers in the door. But service levels are what keep customers coming back, and Six
Sigma helps you reduce defects and non-value added processes to increase customer
satisfaction.

Finally, Six Sigma provides a framework that makes it possible for all customer service
employees to reach achievement milestones and serve customers well. This increases
employee morale, thus increasing the chance that employees will be confident and happy
when serving customers. This circular improvement impacts the customers, who are
happier, which in turn might also improve employee morale. Serving happy customers is
much easier than serving upset customers.

The challenges of implementing Six Sigma in customer service are similar to those of
implementing it in sales and marketing organizations. First, individuals might argue that
people-
manu
service leaders or employees might argue that a strict framework for processes will remove
.

it can build a stable foundation where creativity flourishes alongside consistency. In fact, in
a customer service organization, Six Sigma can help ensure everyone is heard so customer

744
 SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

service representatives who have excellent ideas for assisting customers can help the
company turn those ideas into overall processes.

Another challenge in implementing Six Sigma in a customer service environment can

actually come from both sides. Six Sigma experts who are not experienced in customer
service can be somewhat gung-ho in how they approach improvement, and that can result
in processes that seem to smother the life of the customer service organization. At the
same time, individuals on the business side might balk at any measurement or quality
program that threatens to change the status-quo, especially when those measures are

t applicable experience. In such a situation, the Six Sigma expert must

take time to understand the needs of the customer service organization and present Six
Sigma as a solution in the context of those needs.

The case study for this chapter comes from a company in India. The company provides
financial services to a range of customers, and one of those services is a telephone and
email helpdesk that is supposed to respond to customer queries for information or
assistance. The organization began a project during which team members would be trained
on both Six Sigma and Lean approaches while applying those methodologies to achieve
increased positive results for the help desk.

At the start of the project, the help desk was already being measured. The response time of
the help desk on each call was a time measurement that began when the call was registered

project, leadership teams brainstormed possible issues and a Voice of the Customer tool
was used. The leadership team decided that customer service was critical to help desk
success, so the project would concentrate on applicable improvements in customer
satisfaction.

Through the define and into the measure phase, the team discovered that the current
measurement -- help desk response time
job of showing whether the processes were successful or not. Specifically, the time stopped

different times. Sometimes, tickets were closed when a response was issued to the

745
 SIX SIGMA IN CUSTOMER SERVICE

customer; other times, the ticket was held open until a resolution was finalized. Other
consistency issues arose because different calls required different standards of service and
the help desk often had to rely on work from outside departments, which it had no control
over.

Because it was learning about the Six Sigma met

address all issues in one project. The leadership team had identified 25 different problems,
but you know at this point that attempting to improve all problems in one DMAIC project is
setting a Six Sigma team up for failure. To bring the project into a workable scope, the team
considered two major factors.

First, it looked at the fact that two major measurements existed. Calls could be measured
via response time the time it took to provide the customer with a first-level response to a
call or email. They could also be measured in resolution time the time it took to fully

an issue. Many times, the issue had to be escalated to another department; at that point,

successfully implement an improvement project over processes that are out of their
control. In this case, the team made a choice to concentrate on response times, because
response times were always within the purview of the help desk staff.

The team also looked at disparate types of tasks, which came with various service level
requirements. Different calls had to be responded to within two days, five days, or seven
days. The team created a Pareto chart illustrating these calls and found that almost all calls
98 percent were of the type that should be responded to within two days. Following the
80-20 rule, which had an obvious choice in this case, the team decided to concentrate on
these types of calls.

The team took a baseline measurement and noted that the calls were being handled in an
average of 12.7 days. The goal was to bring that into compliance with the two-day call time
limit, which meant reducing call responses by 10.7 days.

At this point in the project, the team was beginning the analyze phase. Instead of relying
solely on brainstorming and analytical tools, the team again turned to the customer to find
out more about root causes and customer needs. Going back to the customer is rarely a bad
idea, especially when additional customer feedback can help solidify assumptions for root
cause analysis. When you are dealing with a customer service oriented process, you might
have to return to customer data frequently during a Six Sigma project.

746
 SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

The team in question found that customers who called or emailed the help desk very often
wanted to know when something would occur. Even when the customer wanted to know
what the problem was, he or she also wanted to know when the resolution would happen

questions.

to close? The team looked at various factors and brainstormed answers to this question,
nment,
then a call could be closed in approximately 20 minutes. Clearly, if the current average close
rate was over 12 days, there were some major problems with the processes.

Obviously, the gut reaction of help desk staff was that information was not available to
them in most cases. Because they depended on other departments for the information,
they had no real control over how long it took to respond to a customer. You can probably
pletely true.

As data was gathered about the calls and emails, it seemed that about 2/3 of the time, the
help desk did have the information or could have the information to respond. Even if the
other third of the calls took a few days to handle, if 2/3 of calls were answered within 20
minutes, the average would be much lower than 12.7 days.

To apply Lean process management principles that would let the help desk department
schedule staff according to call volumes and handle resources appropriately, the team first

the process map was confusing and fraught with muda. Processes often involved half a
dozen people and included back and forth movement of the work without any real value
added. The team noted that the manager actually resolved about 15 percent of calls himself
during the measured time period. Ideally, most calls should be answered with only the
involvement of the customer, the help desk staff member, and one other person who might
be involved in doing the work the customer was inquiring about.

The team used what it had learned to create a new process map. The goal was to reduce
wasted effort, which meant providing help desk staff with the resources and training
required to close more calls themselves. The Six Sigma team also put solutions in place that
would keep difficult calls from getting caught in the back log. Specifically, the help desk
team would handle calls in a first-in-first-out (FIFO) method whenever possible. Difficult
calls ones that might require more work or communication than normal would be swept
747
 SIX SIGMA IN CUSTOMER SERVICE

twice each day at preset times, and calls that must go up to another level would follow a
specific path that ensured they reached individuals who were most likely to be able to
resolve the calls.

The results of the improvements which basically included applying Lean methods to the
call processes were enormous. Within four weeks, the help desk had reduced response
times by 66 percent. The goal was achieved within six weeks, and staff were surprised at
how easy it was to maintain those results using Lean process management.

To be successful with Six Sigma in customer service, teams and Six Sigma experts must:

Be able to communicate the benefits of Six Sigma to people-centric processes

Take time to understand the needs of customer service organizations and how to
present Six Sigma tools and analysis in light of those needs
Understand the difference between customer service as a function and the idea of
customer satisfaction within the Six Sigma methodology

Links

http://www.processexcellencenetwork.com/lean-six-sigma-business-
transformation/articles/6-ways-six-sigma-can-benefit-your-company
http://www.unitiv.com/intelligent-help-desk-blog/bid/98278/How-Six-Sigma-Improves-
Customer-Service
http://www.qualitydigest.com/may03/articles/01_article.shtml
https://www.isixsigma.com/implementation/basics/measuring-and-improving-service-
processes-six-sigma/
http://www.investopedia.com/terms/c/customer-service.asp
https://www.isixsigma.com/new-to-six-sigma/dmaic/helping-help-desk-satisfy-customers-
case-study/

748
A call center is an organization or department that is set up to handle high-volume calls
from customers, prospective clients, or other interested parties. Typically, a call center
involves a large office space inhabited by numerous individuals who work in shift
environments to ensure calls are answered within appropriate times. In some cases, call
center employees are tasked with making outbound calls in addition to or instead of taking
inbound calls. Inbound calls occur when a customer or someone else calls into the call
center; outbound calls occur when an employee calls someone outside of the call center.

Usually, call centers handle telephone work associated with assisting customers to place an
order, troubleshooting customer service issues such as missing orders or returns requests,
providing account assistance such as bill payments or service upgrades, offering technical
support on products, or helping customers with any other issues. Call centers can process
hundreds or even thousands of calls in a day or week because of call routing. If there are
one hundred individuals in the call center ready to take calls, then the call center can handle
up to 100 customer calls at any given time. Calls are routed to individual work stations,
usually based on algorithms programmed into the phone system, to help reduce the time it

can answer the phone.

While traditional call centers usually have hundreds of employees working in the same
building often in large open spaces that are divided by cubicles modern connectivity has
loyees in a single
building often, workers answer calls from their own homes using VPN or web portals to
perform work. Calls are routed to off-site employees the same way they would be routed to
individuals in an office environment. One example of this type of setup can be seen in the
way Apple handles technical service calls from customers. Many Apple technical
representatives work from home, taking calls from those locations. This lets Apple maintain
long call center hours and employ thousands of representatives without overhead expenses
associated with maintaining numerous office locations.

749
 SIX SIGMA IN CALL CENTERS

premises by which they are managed are very similar. Chat representatives can work in an
office or in their own homes and they answer customer service needs via Internet chat
rather than phone calls. Many eCommerce and web service companies employ chat reps
that are available at certain times of the day to immediately assist with customer needs.
Verizon Wireless, for example, offers a chat contact option online. Verizon customers can
chat with someone online about bills, service questions, or troubleshooting needs.

When discussing Six Sigma in the context of call centers in this chapter, we will mainly focus
on traditional call center environments. However, Six Sigma can also be applied to virtual
call center and chat call center processes.

Call centers benefit from Six Sigma in many of the same ways customer service
organizations do. More organized, streamlined approaches lead to better service and let call
center employees handle calls more efficiently. Because Six Sigma puts control for many
decisions at the production level in this case, in the hands of the person first answering
the call
just any action by coupling the principles of Lean process management and Six Sigma, call
centers can create consistent action, even across hundreds of representatives. The results
are increased performance, lower costs, and happy customers.

While such benefits are critical to the success of any organization, being able to measure
performance and meet specific metrics can mean the difference between keeping or losing
a contract for some call centers. Call centers often operate at the behest of partner
organizations. Third-party call centers might provide service for consumers who deal with
cell phone providers, computer companies, health insurance companies, or even online
retailers. When the call center signs a contract with the company for which it will provide
service, it typically agrees to meet certain requirements. Often, these are known as SLAs, or
service level agreements. Failure to deliver on SLAs could mean loss of revenue or bonus
payments, payment of a fine, or loss of the entire contract.

Six Sigma can be extremely beneficial to organizations that are SLA heavy. First,
implementing Six Sigma approaches helps the call center improve processes so that it can
make SLAs. Second, Six Sigma actually improves the way organizations measure and control
processes, which makes it easier to document compliance with an SLA. In cases where
old up, organizations might be meeting SLAs but not be able
to prove it or even know it.
750
 SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

One of the biggest challenges of implementing Six Sigma in call centers is a misconception
regarding the methodol
that the primary goal of Six Sigma is continuous improvement that both increases customer
satisfaction and business performance, where overall business performance is often
measured as a revenue or bottom-line goal with other, more specific goals, used to
measure the performance of processes.

While Sigma levels, as discussed in the first chapter of this book, are a relevant
measurement for many processes, reaching a level of 6 sigma is not always easy or even
possible for every process. Often, reaching a defect level of 3.4 defects per million

can control every input and the definitions of what is considered a defect are too
unrealistic. The processes are too reliant on people and some of the people-based inputs
are customers or those calling in. Reaching a level of 6 sigma within the first year of
implementation or even several years can require grueling or impossible process
changes that set the entire call center up for failure if Six Sigma experts try to push for too
much improvement or too much change too quickly. Call center jobs are already
considered stressful because they involve stringent response and resolution performance,
so you can quickly alarm employees if you come in heavy with discussions on improving

One of the lessons about Six Sigma presented throughout this book is that a successful
approach to process improvement involves defining challenging, yet realistic, goals. If the

entire DMAIC process. You can probably achieve those goals with some minor changes

extreme measures that might negatively impact customer service, morale, or the bottom
line in the end. In a call center environment, Six Sigma teams have to be able to understand
the current processes and capabilities as well as the available and likely resources so they
can set appropriate goals.

Not only do Six Sigma experts have to be able to set appropriate goals and motivate call
center leaders and employees to continue making improvements via new projects as they
succeed in implementing and controlling changes, they must also be able to communicate
751
 SIX SIGMA IN CALL CENTERS

about Six Sigma in a proactive manner. Six Sigma leaders working in call center
environments have to be prepared for these types of objectives and they should be able to
present illustrations and explanations that use call centers as examples. Presenting
manufacturing examples to illustrate the benefits of Six Sigma could help cement the idea

The case study for this chapter is a call center that was facing the potential loss of a contract
because it was failing to meet the service level agreement it had with the client. The third-
party call center was answering customer calls for the client, providing information and
resolution to customer questions. The call center had a contract with the client that agreed
to a 75 percent first-call resolution and a 90 percent five-day resolution. That meant that 75
percent of the calls that came into the center should be resolved within that first call the
representative should not have to call the customer back or the customer should not have
to make more than one phone call. Additionally, 90 percent or more of the calls had to be
resolved within a five-day period following the initial call. Statistically, that means that 15
percent of the calls not resolved within the first call had to be resolved within five days.

It was known that the client was considering termination of the contract with the call
center, but no one knew what the real problem was. In fact, a first look at the call center
from a Six Sigma perspective revealed an enormous issue of measurement. The call center
was not collecting data relevant to the SLAs it could not provide information about the
first-call resolution or five-day resolution success. It was, however, tracking data such as
hold times and availability of representatives to answer calls. Additionally, someone with

receiving an average of 15 calls every week, she was also a key decision maker for the client.

since none existed. It found that the company had a 50 percent first-call resolution and a 62
percent five-day resolution both metrics were far below the SLA goal.

Analysis also uncovered a number of issues associated with both metrics and process
capability. Data collected during measure and analyze made the team aware of a Catch-22
type problem in the call center. Performance of representatives in the call center was
measured primarily by a single statistic: how often the person was available when a call
752
 SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

came through. If the representatives were not available eve

because they were currently helping another customer it reflected badly on them. The
speed with which they handled the call not the customer problem was seemingly more
important.

The Six Sigma team discovered that many times, the reason a customer issue could not be
closed on the first call was because it required a level of research that would keep the
representative on the phone and unable to take calls. The representative, feeling limited in
how much time he or she could spend with a single caller, would not take the extra time to
-up on issues that
were not resolved during a call. This resulted in these customers having to call back a few
days later
just that customers were not satisfied and the SLA was not met. All those extra, and
unnecessary, calls created muda in the call center and negatively impacted productivity.

The Six Sigma team continued to gather data by monitoring random call center work,
speaking to subject matter experts, and measuring performance. The team also used tools
such as fish bone diagrams to identify root causes and begin brainstorming possible
solutions. Throughout the DMAIC phases, a new process was born and new metrics created.
The call center team was divided into two half the employees took calls and the other half
performed research. Call center representatives took turns handling each task, and metrics
to measure performance accurately were designed for both groups.

The IT department was also included in work to improve the process. They modified the
system to include fields in the call logs where research and requested information could be

management. Within a few weeks, the call center saw incredible results. Not only were
first-call and five-day resolution times well within SLA requirements, but the decision maker
at the client was receiving less than one customer call each month as compared to dozens.

To be successful with Six Sigma in call centers, teams and Six Sigma experts must:

Understand the people-inherent nature of the call center environment

Realize that automated solutions and templates are valuable in call centers but that

753
 SIX SIGMA IN CALL CENTERS

Understand the level of stress many employees in a call center might feel to meet
performance requirements and position Six Sigma as a tool to help them meet the
right requirements rather than a demand for additional requirements
Be able to communicate regarding Six Sigma in a way that invokes examples and
illustrations that are relevant to the call center and not to manufacturing or similar
industries.

Links:

https://www.isixsigma.com/implementation/case-studies/using-lean-six-sigma-improve-
call-center-operations/
http://www.sixsigmaonline.org/six-sigma-training-certification-information/can-a-call-
center-benefit-from-six-sigma-certification/

http://www.callcenter-iq.com/operations/columns/six-sigma-may-be-dangerous-to-your-
call-center

754
Most people take retail for granted, happy that they can simply pick up the items they need
from a local store. However, a great deal of logistics and effort lie underneath the surface to
make this happen.

Retail operations mainly take place in a storefront, where cashiers and front end personnel
are aided by store management, receiving personnel to handle the receipt of new products,
and back end employees to stock shelves. If a retail chain has multiple stores, it increases
the work force and presents challenges in rolling out changes to all locations. Additionally,
each retailer must work with a long list of vendors to order new products, coordinate
deliveries, and return spoiled merchandise.

While the physical brick-and-mortar location is the foundation of retail, the industry is
becoming increasingly reliant on online sales. Online retailers enjoy the benefit of being
able to sell products that are not sold in stores, increasing the options of their customers
without wasting valuable shelf space. The online model has its fair share of potential issues,
particularly shipping times, shipping costs, and the availability of products listed on the

Another growing concern for retailers is the merging of the physical location and the online
shop. The most popular use of this synergy is items purchased online for in-store pickup.
This requires the store and the website to be in sync in real time so that inventory figures
are accurate and customer wait time is minimized.

The retail industry comprises 42 million jobs and contributes more than $2.5 trillion to the
GDP of the United States. (https://nrf.com/advocacy/retails-impact
every retailer to operate as efficiently as possible and to maximize their profit margins. Six
Sigma can help virtually any retailer to eliminate inefficiencies, improve customer service
and encourage customer loyalty.

755
 SIX SIGMA IN THE RETAIL INDUSTRY

Inventory Management
Marketing and generating interest in products is only one small part of increasing revenues.
Companies must actually have the products desired by consumers in stock. Although the
inventory process has become increasingly automated, Six Sigma can still play a role in
effectively managing inventory resources.

The typical inventory process has multiple potential bottlenecks. From the physical process
of ordering to delivery time to loading the products off the truck, there are many ways in
which the normal flow of products can be disrupted. Identifying these problem areas can
help any company to zero in on the true reasons for delays or delivery of the wrong
products.

Companies that deal with products that have longer lead times stand to benefit the most
from a thorough analysis and process improvement initiative relating to inventory.
However, companies of all sizes can streamline their approach to inventory, creating an
iron-clad process that can be followed in locations across the country.

Scheduling
The bigger a company gets, the more difficult the small tasks become. Scheduling is a major
issue for retail companies, particularly in those operations with multiple locations and
longer hours.

Automated scheduling solutions can help in a variety of ways. With the use of forecasting
based on historical data, modern technology enables the schedule to meet the anticipated
inflow of customers. Like many Six Sigma tools, this allows store managers to see the ideal
solution, rather than simply what they believe the answer to be. Managers can also
maximize in-store personnel while minimizing salary, a major benefit since most retail
employees are hourly employees. Utilizing part-time on the lower spectrum of the salary
scale can help retailers to achieve significant cost savings. Six Sigma can help retail stores to
find the right mix of reliable full-timers and part-time help.

Employee Productivity

difficult to focus on the productivity level of employees. Six Sigma can help managers to
identify the metrics most closely aligned with productivity, then track these areas over time.
For customer-facing employees, these metrics can include items scanned per minute, credit

756
 SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

card applications or extended warranties sold per day, customer survey ratings, and add-on
items sold.

Six Sigma can also help to track the productivity levels of back end employees as well. More
importantly, the potential exists for process improvements that can overhaul internal
operations, increasing efficiency and employee morale. Implementing processes that are
easy to follow and track is vital due to the high turnover rates associated with retail.

Tracking Difficulties
Although a great deal of potential exists for the future of Six Sigma in retail, there are
significant hurdles that stand in the way of progress. Initiatives to track inventory and
productivity go only as far as the individuals using those systems. Inadequate training and
lack of attention to detail can make it very difficult for an organization to fully understand
the issues at hand.

For example, many stores use electronic reordering systems that receive its figures from the
number of items scanned at the point of purchase. As long as the correct number of items is
scanned, the store can successfully reorder the correct number of items. However, if the
- for instance, if a cashier scans one can of cat food 24
times instead of accounting for all of the different varieties -

lines, productivity analysis of the cashiers who fail to follow the typical process is similarly
derailed.

Lack of Interest
Retail operations rely heavily on part-time talent for their scheduling flexibility and for their
lack of financial burden. Not only do part-timers make less money than full-time employees,
t as much money for their benefits packages.
However, these advantages come at a cost.

Part-

inevitable change that comes with Six Sigma. A high school student who works in a retail
store on the weekends has no real motivation to comply with the employee dedication
757
 SIX SIGMA IN THE RETAIL INDUSTRY

implement Six Sigma, and many retail stores struggle to come up with the winning mix of
resources.

Difficulty Implementing in Multiple Locations

Embarking on a Six Sigma initiative is an arduous task for any company. The prospects
become that much more difficult when multiple locations are involved.

On one hand, having many retail locations within one organization can be helpful in terms
of obtaining the most data points and coming up with new ideas. But the reality is that
getting people to buy in and take ownership of the Six Sigma project is hard enough in one
location. Expanding the scope to more than one location comes with serious challenges.
Even if the actual project is limited to one store and subsequently rolled out to the other
locations, convincing entire stores to change so much can be very difficult.

Additionally, each retail store within a larger company may have its own culture and ways of
doing things. What boosts morale at one location may derail another store entirely.
Implementing a consistent set of policies and metrics across all physical locations may be
the biggest challenge faced by the company.

Am -largest retailer in the

world in 2016 was the result of careful planning and consistent dedication to performance
improvement.(http://www.forbes.com/sites/laurengensler/2016/05/27/global-2000-
worlds-largest-retailers/#7b21c79b29a9) While the company has not published information
about any Six Sigma projects to date, a
the principles of define, measure, analyze, improve and control have been in place from Day
One. (https://www.isixsigma.com/community/blogs/six-sigma-amazoncom/)

problem solvers is a testament to the influence Six Sigma has had on Amazon. The fact that
Amazon has received such a benefit from adhering to the principles of Six Sigma is no
surprise. Managing warehouses all over the United States means keeping track of the
inventory levels of these warehouses and knowing which warehouses have to ship to a
given area. This is especially critical given the costs of shipping merchandise and the
responsibility that comes with guaranteed delivery dates.

758
 SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

-
CEOs - explained that the philosophy of Amazon held the Six Sigma principles dear. Amazon
adopted technological solutions to decrease the variability of the warehouse workflow,

failures by large companies to handle online ordering, such as Toys R Us during the 1998
and 1999 holiday seasons (http://articles.sun-sentinel.com/1999-12-
24/business/9912231103_1_toysrus-com-order-status-shoppers) Internally, Wilke and his

everyone could understand and that everyone could get behind. This type of buy-in is the
goal of all Six Sigma operations
number of products and services over the years.

Whereas many businesses compete on one core competency, Amazon aims to achieve
every strength conceivable - the biggest selection, for the lowest price, with the best
customer experience possible. Amazon has famously sacrificed profits in the name of re-
investing into the company, aiming to become even more efficient and prolific. The results
million, while the company
posted a $513 million profit in the first quarter of 2016.
(http://www.cnet.com/news/amazon-says-first-quarter-sales-and-profit-blew-out-
expectations/)

company. Newer services, such as Amazon Fresh (online grocery shopping) and Amazon
Web Services receive the benefit of years of the practice of Six Sigma methodology in other
areas. Because Amazon has gotten the process of order fulfillment and timely delivery down

maintaining a vigilant eye on the internal metrics that translate into success.

As stated earlier, a major challenge for retailers is reconciling their in-store retail operations
with online sales. In the midst of navigating these waters, the company is still responsible
for maintaining its usual workflows and sales targets. Given the uncertain timeframes that
come with technical solutions, this is a difficult problem for many companies to solve.

A popular home shopping retailer noticed inefficiencies within both their in-store and online
sales processes. The most notable issue was their cycle time when it came to issuing new
759
 SIX SIGMA IN THE RETAIL INDUSTRY

products. The length of time between the conception of a new product and its subsequent
launch became prohibitively long, which is a dangerous situation for a clothing company to
be in. It is vital for apparel retailers to be the first to the market and to always be on the
cutting edge in terms of fashion. Featuring outdated fashions in a catalog or on a website
can spell disaster for even the most prominent retailers.

The company decided to work with a consulting agency, which recommended a Six Sigma
http://www.palomaconsulting.com/case-
studies/retail/case-study-2-improving-process-efficiencies-within-the-creative-process-of-
leading-home-shopping-retailer/) While the company wanted to improve its online process,
it was decided that focusing on this area
growth, it was already undergoing a series of IT initiatives, meaning that solutions of a
technical nature would have to wait until those projects had completed. Besides, the real
-sale date. The
consulting group and the company agreed that this cycle time should be the main focus of
the Six Sigma project. Additionally, while the company owned 30 different clothing brands,
it kept the focus to only one brand in order to narrow the scope of the analysis.

Before beginning, company leadership and the consulting agency met with the stakeholders
for the processes at hand. Both groups knew that the only way the project would succeed
was if every member of the organization bought in. To that end, employees were allowed to
ask questions and make suggestions based on their everyday work experiences. In the end,
a series of smaller projects was devised based on the feedback from the staff of the
company.

The Six Sigma initiative focused on two main areas - reducing variability and minimizing
waste. Improving performance in these areas meant that the process to move a creative

group and the company worked together to improve the use of internal resources; they also
simplified the creative briefing process, which meant the catalog writers and online store
personnel could more easily describe and promote new products.

In the end, the company achieved its goal of reducing its cycle time. Furthermore, the time
necessary to produce creative briefs was reduced, as were the number of errors related to
artwork. Most importantly, the improvements made in this Six Sigma process proved to be
repeatable. When the company implemented these changes in their other clothing brands,
they saw similar positive results. The benefits this company received from their Six Sigma
initiative was well worth their investment, and the resulting improvements have stood the
test of time.

760
 SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

Six Sigma methodology has the potential to transform an entire retail operation, turning it
from an unstable environment into the model of efficiency and customer satisfaction.

Developing a plan to convince employees to buy in may be the most critical step in the Six
Sigma journey. It may be difficult to sell a teenage cashier on the merits of a Six Sigma
-in
may be hard to achieve when there are multiple locations involved. Coming up with a
strategy to convince employees to go along with the plan is essential.

A clearly defined project scope is vital for success. If a retail operation has many stores, it
should focus only on one location as it begins to implement and analyze changes. Once the
modifications prove to be successful, then - and only then - should the new process be
rolled out in other locations.

Retail is a highly competitive industry that must always be able to change with the times.
However, Six Sigma requires a strict inward focus. The more a company worries about what
everyone else is doing, the more ill-equipped the company will be to accurately assess its
own flaws and defects. Any strategy can be tweaked to fit into the context of a competitive
landscape after the fact. To really make positive internal changes, the company must keep
its eye on the ball.

Source

http://www.forbes.com/sites/laurengensler/2016/05/27/global-2000-worlds-largest-
retailers/#7b21c79b29a9
https://www.isixsigma.com/community/blogs/six-sigma-amazoncom/, http://articles.sun-
sentinel.com/1999-12-24/business/9912231103_1_toysrus-com-order-status-shoppers,
http://www.cnet.com/news/amazon-says-first-quarter-sales-and-profit-blew-out-
expectations/
http://www.palomaconsulting.com/case-studies/retail/case-study-2-improving-process-
efficiencies-within-the-creative-process-of-leading-home-shopping-retailer/

761
The eCommerce
seen large online retailers like Amazon expand to areas far beyond their original niche,
providing a fantastic, low-cost customer experience that everyone can enjoy. As a result,
many retailers have chosen to move to an online-only model, while others have simply
chosen to close their businesses altogether.

do prefer the instant gratification they derive from making an in-store purchase in some

and the immediate need for an item is diminished .

(http://www.chicagotribune.com/lifestyles/sc-cons-0904-savvy-shopper-20140904-
story.html) To that end, more than 75 percent of consumers aged 15 and over made an
online purchase in the first three months of 2014.
(http://www.business.com/ecommerce/retail-or-e-tail-buying-online-vs-buying-in-person/)

deal. eCommerce also

provides many benefits to retailers. Because they have limitless space to advertise their
products, companies can sell more products online than they can in stores. Additionally,
ing and running a physical
showroom, their overhead expenses are far less than those of brick-and-mortar retailers.

While eCommerce is a term that generally describes any online purchase, there are several
different platforms in which eCommerce takes place. The traditional model is where one
company hosts its own website and inventory storage systems. This company handles all
aspects of the order, from its inception to its ship date. The eCommerce retailer is also
charged with handling after-sale support functions such as refunds, shipping issues and
general product support. Another way companies can embrace eCommerce is through the
use of online storefronts, in which a company takes orders, but the actual items are
processed and shipped by a third party. This practice is also referred to as drop shipping.

User-based online stores give individuals the ability to sell their products online through a
larger site. Companies like eBay and Etsy take a cut of each sale, but this fee entitles the
seller to the protection and security that comes with partnering with a well-known name
brand. These affiliate-based models are highly beneficial for the producers of the materials
762
 SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

sold because it gives them a much more visible platform. The retailer benefits because they
always have a steady stream of products and interested customers that are ready to buy.

Shipping and Shipping-Related Logistics

Shipping is the lifeblood of any eCommerce company. Without a system in place to process
and fulfill orders, the company has no chance of succeeding.

shipping burden is placed on the customer. The customer needs their item within a
reasonable amount of time, and the customer needs to not pay an exorbitant amount of
money to have the item shipped. Excessive shipping costs are the top reason why
consumers abandon items in their online shopping carts.
(https://www.shopify.com/blog/8343330-how-to-choose-a-shipping-strategy-for-your-
online-store
ry sale due to shipping issues.

Six Sigma can help online retailers to notice inefficiencies in their shipping processes -
assuming those processes even exist in the first place. Variability in the process can make it
difficult for companies to guarantee a delivery date for consumers, which then disappoints
potential customers to the point that they look elsewhere. Additionally, poor deployment of
internal resources can help to increase shipping costs and delays. This is especially true for
companies that have multiple warehouses or fulfillment centers. Six Sigma can help
companies to straighten out the logistics behind the process, clarifying the responsibilities
of each fulfillment center so that each unit can produce at optimal levels.

Scope of Products Offered

Many eCommerce retailers take advantage of their ability to offer an increased amount of
products to their customers. A product list that would otherwise be limited by shelf space
can now be displayed in full, giving customers a potentially endless list of products from
which they can choose.

make it difficult for consumers to find what they want; they may also find it hard to choose
between potential options. As a result, those prospects may end up buying nothing, victims

consumers may end up seeking out other retailers with simpler layouts and product pages
that are easier to locate.
763
 SIX SIGMA IN ECOMMERCE

The temptation for all eCommerce retailers is to offer more products as opposed to fewer

This is one area where Six Sigma can be highly beneficial. In addition to using purchase and
website activity data to identify the best performing items on the website, a Six Sigma
project can also dig deep into the unspoken desires of consumers. Companies may not
realize that an expanded slate of products can result in their audience feeling alienated,
believing that the company has abandoned its niche and its core customer base. A Six Sigma
initiative will show a company where its focus should lie, both in terms of product
categories and the varieties of those products that should be carried.

Enhanced Ability to Track How Purchases Originate

The digital marketing landscape provides organizations with a variety of ways to spread the
word about their products. This is vital because the Internet is the best way for a marketer
to a

Because of the many marketing options available - social media, pay-per-click and organic
search engine optimization, just to name a few - it can become very challenging for

their site. How

curious?

Implementing Six Sigma can help any company that wants to get a better handle on their
marketing. Such an initiative can help the company to curtail unnecessary spending, identify
areas where the company is successfully reaching customers and eliminate any marketing

avenues it should pursue from a marketing perspective, and it also enables the company to
calculate its marketing return on investment. Knowing this number will help the business to
anticipate demand in future marketing campaigns so that the proper infrastructure - order
fulfillment, online servers, customer support - can be put into place ahead of time.

The Role of Outside Parties in Shipping

One of the best aspects about Six Sigma is that it encourages total control over the internal
process. This is a major problem in eCommerce because so much of the order fulfillment

764
 SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

, the customer

from the postal snafu can have dire consequences for the business.

Products that arrive on time, but were damaged during shipping, can be just as problematic.
In this instance, the company may be forced to incur the cost of return shipping as well as
the costs to ship a non-defective item. Furthermore, the company now has one less item to
sell. Once again, a problem not caused by the company can significantly impact normal
operations.

all the same, and the organization is forced to deal with them. As these instances pile up,
they can make tracking normal process flow difficult. It may also potentially sour business
owners on implementing Six Sigma initiatives. After all, if they follow the process to the
letter, and the product becomes defective through no fault of their own, why even bother
doing the project in the first place?

Working with Outside Vendors

work flow. The other external players in a process can be just as damaging.

Any force outside of the direct control of an individual can cause a disruption. For example,
an independent retailer who sells their goods on eBay may have seen a dip in business
because of their data breach in 2014.
(http://www.forbes.com/sites/gordonkelly/2014/05/21/ebay-suffers-massive-security-
breach-all-users-must-their-change-passwords/#65fcb3173c15) Just the same, a large
retailer like Sears that employs drop shipping may experience a downturn in users selling on
their site because a competitor, such as Amazon, begins offering better margins to sellers.

Outside vendors that experience difficulty in fulfilling orders not only contribute to
inefficiency, but can also create public relations nightmares for the company. Many
printing-based storefronts actually have their prints fulfilled by larger companies. This helps
them to save on resources while offering their end users a variety of printing options.

storefront, while the printer gets off scot-free. These problems would also lead to the

their ability to gather the data necessary to begin a Six Sigma project.

765
 SIX SIGMA IN ECOMMERCE

Changes in Customer Preferences

Perhaps the biggest deterrent to eCommerce success has nothing to do with outside
vendors or product defects. The main threat, it could be argued, is the fickle nature of
online customers. Slow-loading sites, layout changes and a general shifting of tastes can
dramatically alter traffic and revenue figures.

mobile devices than go online via laptops or their home computers.

(http://www.smartinsights.com/mobile-marketing/mobile-marketing-analytics/mobile-
marketing-statistics/) As a result, online retailers have had to adjust the layout of their sites
to appeal to online cus -friendly site
were heavily penalized by Google in the form of unfavorable search engine rankings.
(http://searchengineland.com/mobilegeddon-beginning-not-ending-220512) In light of this
rapid growth of mobile-based eCommerce, each retailer had to adjust. Those that failed to

their numbers took such a hit.

This ever-changing nature of eCommerce is a major deterrent that might not even show up

be quantified. And even if a company suspects that their site might not be cut out for
modern times, the leadership team might not be prepared to admit that change is

-based bidding ensures that the

site has something for everyone.

At its core, eBay is essentially a middleman that links sellers of goods with people who want
to buy those products. The site allows users to list products and facilitates buyer bidding;
eBay also provides pre- and post-sale support and integrates with PayPal to handle user-to-
user payments without disclosing any of the financial information of either party.

sellers to see how many transactions each individual has been part of, as well as any
complaints that have arisen from previous customers. Leaving negative feedback on a
purchase is a major negative mark that can have long-lasting ramifications, creating

766
 SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

skepticism among the customer base and affecting future business. Before a buyer or seller
leaves negative feedback, he or she is encouraged to exhaust all possibilities for resolving
the issue before taking that last step. To accomplish this, eBay customer service must
sometimes get involved.

A Six Sigma project was undertaken by Pete Abilla, former employee of eBay, to eliminate

customer service team always utilized the principles of DMAIC, additional study was needed
to identify some of the causes for customer dissatisfaction.

One of the more notable areas for concern was the online chat department, which handled
customer concerns and provided service in real time. (http://www.shmula.com/lean-six-
sigma-ecommerce-presentation-case-study/12002/) This eliminated the need for a lengthy
email chain that could take days, if not longer, to resolve. Online chat would resolve issues
virtually immediately, unless there was a bigger issue that required intervention from a
manager. However, the online chat model had some major flaws that interfered with the

After opening the chat window, the user was asked a few questions to verify their identity
before the chat could begin. The user had to provide their User ID and email address so that

necessary support. However, as soon as the customer service representative joined the
chat, he or she began by asking the user for additional information -
address and telephone number, along with the same User ID and email address that the
user had just provided.

Not only was this a significant redundancy, its fallout affected everyone. First, it frustrated

a second time. Furthermore, it added to the time of the chat. This kept the user in the chat
for longer than necessary, and it also reduced the overall number of users that the
customer chat team could help. Additionally, some users may have developed concerns that
they were being phished for their personal information. Although the initial message from
the eBay chat team clearly stated that they would never ask for information such as credit
card numbers and passwords, users had reason to be skeptical, given that they were just
asked for information they already provided.

By using Six Sigma methodology, eBay realized that they were asking for basic information
too often. They also realized that the pre-
the necessary information needed by the chat team. As a result, the pre-chat survey was

767
 SIX SIGMA IN ECOMMERCE

expanded to ask for all necessary information to initiate the support chat. This also helped
the support team to have this information laid out in a predictable and organized way,

In the end, eBay was able to reduce the verification time from 3:42 to 2:57. This difference
of 45 seconds shaved 20 percent off the average verification time, enabling eBay to serve
more customers and giving customers a little more time to do what they most like to do on
eBay - buy and sell items.

Six Sigma can be a valuable tool for identifying the strengths of an eCommerce operation.

itself too thin in the search for revenues.

Keeping a narrow focus on what brings customers to an eCommerce store helps to keep
that store well-positioned within its chosen niche. Understanding the customers that
frequent the store is crucial when it comes to changing with the times and adapting to the
needs of the customer base.

manageable. Adding to the product mix may result in sunk costs and excessive inventory,
not to mention

Understanding that the Internet is a constantly changing medium is a necessary component

of any eCommerce business. Incorporating Six Sigma while knowing the entire industry
could change at any minute is a bit tricky. However, sticking to the pre-established company
principles will keep the company on track and enable the organization to focus on getting
the desired results out of their Six Sigma project.

Source

http://www.chicagotribune.com/lifestyles/sc-cons-0904-savvy-shopper-20140904-
story.html
http://www.business.com/ecommerce/retail-or-e-tail-buying-online-vs-buying-in-person/
https://www.shopify.com/blog/8343330-how-to-choose-a-shipping-strategy-for-your-
online-store

768
 SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

http://www.forbes.com/sites/gordonkelly/2014/05/21/ebay-suffers-massive-security-
breach-all-users-must-their-change-passwords/#65fcb3173c15
http://www.smartinsights.com/mobile-marketing/mobile-marketing-analytics/mobile-
marketing-statistics/
http://searchengineland.com/mobilegeddon-beginning-not-ending-220512
http://www.shmula.com/lean-six-sigma-ecommerce-presentation-case-study/12002/

769
In a sense, the warehouse is the unsung hero of the retail industry. When things are going
well, nobody notices the work the warehouse puts in to keep operations running smoothly.
However, when something goes wrong, people are quick to blame the warehouse for its
inability to handle seemingly simple tasks.

rely on intricate processes to fulfill and ship orders in a timely fashion. Even the smallest
failure can result in serious problems that can derail an entire retail operation. In other
words, without an efficient warehouse, it becomes impossible for the retailer to locate and
sell its products.

If the showroom of a business is the front end, the warehouse is the back end. The typical
image of a warehouse is a back room where employees are tasked with locating and
transporting the items requested by a customer. This can take the form of simply bringing
up an item from the basement and handing it to a salesperson on the sales floor, or it can
be an online transaction in which the order is shipped. Customers do not interact with the
warehousing process; it is their assumption that everything will operate smoothly. And
since this is what the customer expects, it must become the reality of the customer
experience.

The responsibilities of the warehouse become more complex when looking at larger
corporations with multiple warehouses. A company like Amazon, which hosts 90
warehouses around the world, has serious choices to make about each item it stocks.

770
 SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

(http://www.wired.com/2014/06/inside-amazon-warehouse/) The company has to decide

which warehouses it will be stored in, how many units to distribute to each warehouse and
how many different locations that item should be placed in each warehouse. The actual

comes in.

Process Overhaul
Warehouses are constantly in action. Employees of warehouses are so consumed with
fulfilling orders and doing what has to be done that they rarely get the chance to think
about how operations could run more efficiently. Even if a warehouse worker were to come
up with a good suggestion for what could be changed, it would most likely be impossible to

afford.

A Six Sigma project is the perfect opportunity to create these necessary changes that will

project will illuminate the ideal solution, not just what an uninformed observer thinks
should be done. The solution brought about as a result of the Six Sigma initiative will have
input from everyone involved in the process, making it easier for staff-level employees and
management to embrace the new process. Second, and perhaps more importantly, carrying
out the Six Sigma project means that the data to support these findings will be present,
removing any doubt about the proposed solution. This also means that if the company is
forced to take a segment of the warehouse out of commission for a period of time, the data
will show exactly why this is necessary and what the eventual benefits will be. As a result,
management can make the choice that will help the warehouse to become more efficient in
both the short term and the long term.

Many warehouses carry out obsolete and outdated process on a daily basis, partly because

alternatives. Six Sigma can help to foster meaningful changes that will benefit the entire
company, with minimal upheaval to the current level of operations.

Enhanced Communications
Although the warehouse process is focused on the efficient processing of customer and
retailer orders, that process cannot start without receiving the order. If there are any
communication issues that prevent the warehouse from receiving the order in a timely

771
 SIX SIGMA IN WAREHOUSE

manner, or if the instructions on the order are often unclear, the warehouse cannot do its
job effectively. As a result, the warehouse that gets the blame, even though the information
provided was incomplete, if it arrived at all.

Six Sigma methodology is designed to identify these breakdowns of communication. More

specifically, Six Sigma can help tailor the process so that the warehouse picking team gets
orders in the exact specifications they prefer. When communication is streamlined, the
team can get right to work fulfilling the order, which makes the retailer and the customer
happy.

This practice can also help to improve communications between the warehouse and the
-minute order changes and cancellations are
inevitable, but if handled incorrectly, they can cost the warehouse time and money,

methods of communication can let the warehouse know when an order is changed via a flag

Additionally, Six Sigma can help the warehouse to identify the ideal time to start filling an
order. Because the customer may have a grace period in which changes can be made,
jumping on a customer order right away might not always be the best idea. Using Six Sigma
to analyze the order fulfillment process can help the warehouse to zero in on the best way
to handle new orders.

Inventory Management
The proper and cost-effective management of inventory is a vexing issue for warehouses
worldwide. Carrying too much inventory minimizes the possibility of stock-outs, but it can
become costly in short order. On the other hand, maintaining a low level of inventory on
hand reduces costs, but presents a risk in terms of keeping up with demand.

Using Six Sigma can help warehouses to determine their magic number for how much

area - for example, sale items must always have higher quantities of inventory because
- but Six Sigma can give a general guideline
and philosophy that the warehouse can utilize in its daily processes.

Every business is different, and what works in one industry might not work in another.

Frequent orders may incur additional shipping costs, but can provide equal or greater
savings in terms of inventory-based expenses. I

772
 SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

do without looking at its numbers and seeing how the figures align with the actual
processes carried out by the company. The use of Six Sigma in this area is essential.

Outside Factors
Six Sigma is an ideal discipline for situations in which the organization has full control of its
operations. On one level, warehouses fall into this category. But in many ways, outside
forces can completely unravel anything a Six Sigma program sets out to accomplish.

For instance, a manager may initially decide that a Six Sigma project is worth undertaking,
only to change their mind when they see how much time and manpower the project is truly
costing the company. A rush to return to the old system may be in order, even if the
eventual benefits of the project are undeniable. The unfortunate reality is that the show

during the project, company leadership may impose its will, even if it hurts the company
over the long haul.

Another challenge comes in the form of communicating with the customer service team.
Representatives who are not properly trained in the proper methods of order transport not
only slow down production, but they may even create more work for warehouse staff.
Additionally, changes to orders can be problematic for the warehouse, even with a
comprehensive process in place.

Labor Issues
A Six Sigma initiative can only be successful with the consistent effort of the stakeholders.
This becomes an issue in the warehouse industry in two ways. First, the staff-level
employees of the warehouse might be resistant to the changes that arise as a result of the
project. The employees may fear that the company is being downsized, which would
eliminate their jobs, or they may simply be disinterested in a large project at a place in
-term.

The issue of turnover is one that plagues warehouses around the world. Due to the
relatively low wages, the mundane nature of the work and the physical exertion required
each day, warehouse staffers have a high turnover rate.
(https://www.kaneisable.com/blog/warehouse-labor-the-real-cost-of-associate-turnover)
The bigger the warehouse, the more turnover there is.
773
 SIX SIGMA IN WAREHOUSE

(http://www.emeraldinsight.com/doi/abs/10.1108/09600030710758437) While there is

undoubtedly value in an experienced warehouse employee, management is likely to see this
-cost, unskilled labor
anywhere.

The impact of this philosophy regarding overall warehouse performance is debatable.

Six Sigma project, that initiative has some serious challenges to consider. The only way to
get to the heart of a problem is to talk to experienced staffers and identify the issues they
face in their daily responsibilities. Similarly, the only way to truly know if a solution is viable
is to present it to those same employees and verify that it meets their needs. A high-
turnover situation makes this sort of fact-finding and analysis extremely difficult. A Six
Sigma project needs to rely heavily on the veteran employees who are unlikely to leave and
make those workers the basis of the investigation.

Practical Limitations
The ultimate goal of any Six Sigma project is to generate the solution that will allow the
company to operate at the maximum level of efficiency. And while it may be possible to
obtain this solution, it still might not be good enough to satisfy the desires of management.

Instead, there are simply some constraints that cannot be avoided. In particular, space and
capacity can limit the company from achieving its goals, even if the Six Sigma project is a
success.

Creating the perfect process that helps the warehouse to become streamlined, fast and
defect-free is a great end product of a Six Sigma initiative. But that solution only goes so far
when the holiday season begins and the company is inundated with orders. Though the
warehouse may be able to process and fulfill orders faster than ever, the warehouse staff
might not be able to keep up with demand.

This is a good problem for the company to have, to be sure. If the process put in place is
well-documented and easy to follow, temporary workers can be hired to continue the work
that was started during the project. However, this also opens the door for inadequately
trained employees to introduce defec
leadership to ensure that anyone who joins a smoothly operating process knows exactly
what their roles and responsibilities are. More importantly, those new pieces to the puzzle
need to know what not to do.

774
 SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

The physical size of the warehouse can also become problematic. Once again, a warehouse

means the company is doing something right. But if the company expands, there are some
tough decisions to be made. The company can continue to work beyond its capacity, putting
additional strain on the employees and the system that was created during the Six Sigma
project. On the other hand, the company may decide that additional warehouse space is
needed, potentially creating logistic problems that can cause the defects and interruptions
that the Six Sigma initiative was designed to avoid.

Although IBM has largely abandoned the hardware computing industry in favor of software
and consulting solutions, the company is still synonymous with the high-powered business
computers and servers that made IBM a household name. IBM and Lenovo (the company
iness) have been the preferred hardware suppliers for
the business industry for the better part of two decades, and will continue to be for the

dedication to responsible inventory management.

(https://www.isixsigma.com/implementation/case-studies/case-study-using-measurement-
drive-behavior-change/)

Inventory is a major issue for technology-based companies. Hardware is expensive, and

having an excessive amount of inventory on hand means a great deal of cash is tied up in
goods that are manufactured, but not yet sold. More troubling is the fact that this excess

chance that surplus inventory becomes waste. With new products and features becoming
ks.

until the order was shipped to the customer. This shipment to the customer did not
necessarily happen right away. As a result, inventory numbers remained high, even though

ses. Not only were inventory

numbers inaccurate, but outdated machines were kept in circulation, which cost the
775
 SIX SIGMA IN WAREHOUSE

company money. IBM decided to embark on a Six Sigma project to determine why inventory
numbers were so high in spite of consistent sales.

The company discovered that the problem had nothing to do with the warehouse staff, or
the internal processes carried out in the warehouse. Instead, it was the sales team that had
the most culpability. It was the sales team that directed the ordering process, and if a sale
fell through, an item that was ordered remained in the warehouse. Instead of shipping it
back to IBM for refreshing, the sales team would frequently keep the item in the warehouse
in the hopes of selling that item to another customer. Other issues included changes made
by customers after the sale, as well as salespersons ordering items early in anticipation of
potential future sales that may or may not come to fruition.

ng inventory. In
the minds of the sales team, having inventory on hand meant it was easier for them to do
their jobs. The more computers they had in a warehouse, the quicker the sales team could
move those items.

The first solution brought about by the Six Sigma project was ensuring an across-the-board
understanding of the importance of inventory. Though the sales team benefited from the
extra items in the warehouse, they did not realize the financial burden that came with these
additional machines. Once the sales team understood why inventory mattered so much,
they were able to buy in to the additional solutions generated by the initiative.

The most critical change was the implementation of detailed record keeping of sales and
the inventory tied to those tr

headquarters so that the hardware could be retooled. Doing this meant that inventory
would no longer become obsolete. It also ensured that the sales staff would be accountable
for the end results of their orders.

Before the project, IBM experienced a failure rate that is, the number of items
commissioned for shipping that never got to the customer of 50 percent in its largest
country of business. In another large country, IBM had an 80 percent failure rate. Thanks to
the changes brought about by the Six Sigma project, IBM reached a failure rate of less than
20 percent in both of these areas. Additionally, warehouse inventory levels across the entire
company were cut in half.

warehouse. However, as discussed earlier, outside factors play a major role in the

776
 SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

productivity of a warehouse. All parties involved must understand the role inventory and a
coherent process plays in the warehouse. When the warehouse can function at optimal
levels, every aspect of the company benefits.

Tips for Using Six Sigma in the Warehouse Industry

A warehouse that undergoes a Six Sigma project will see tremendous benefits. Obsolete
processes can finally be discarded, and warehouse employees will have their say in making
sure that the new process is effective and practical. In the end, everyone wins, from the
warehouse staff, to management, to the customers who receive products quicker and with
fewer defects.

project with everyone who may be affected. It should be clearly stated that any changes

typically felt by the warehouse and it staff.

essential to get buy-in from all stakeholders, particularly the staff-level employees.
These are the people who will be carrying out the changes that are made as a result of the
project. The warehouse staff has the power to make or break any change that is
implemented. By getting on the good side of these individuals early and giving them a
chance to have a say in the new process, the company sets itself up for prolonged future
success.

The warehouse has many masters to serve. However, it cannot serve any of those masters if
its own house is not in order. The project should focus only on improving the efficiency of
the warehouse. Outside factors, such as corporate directives and the whims of other
departments, should not be considered. If the warehouse is able to focus on its own issues
and come up with its own solutions, improved efficiency and reduced defects are virtually
guaranteed.

Source

https://www.isixsigma.com/implementation/case-studies/case-study-using-measurement-
drive-behavior-change/
http://www.emeraldinsight.com/doi/abs/10.1108/09600030710758437
https://www.kaneisable.com/blog/warehouse-labor-the-real-cost-of-associate-turnover
http://www.wired.com/2014/06/inside-amazon-warehouse/

777
 money. The power of
distribution is what enables businesses to recoup their investments and profit from their

business to increase its exposure, nor can that business plan for the future.

most visible online retailers. In the case of brick-and-mortar retailers, the store needs
assurance that the item is worth the financial investment and shelf space that comes with
stocking that item. Just the same, online distribution might not be the utopia it seems,
particularly with so many third-party sellers cluttering up the marketplace on sites like
Amazon and Etsy. It takes the right mix of product availability, choice of retailer and the
product itself to draw up a winning formula.

These days, thanks to the advent of online sales, anybody can be a distributor. But if the
business succeeds, consumer demand will

and outs of the distribution industry comes in handy.

Distribution can come in a variety of forms. A local band selling its demo tape on
consignment at a mom-and-pop record store is one form of distribution. A clothing retailer
that signs an exclusive agreement to sell its clothes at Target stores around the United
States is another. At the end of the day, distribution is what allows a company to tell its

The level of reach and the number of distribution outlets is ultimately up to the company
selling the product. A home-based drop shipping operation might not want to attract too

demand. On the other hand, a larger manufacturer might only partner with a retailer that
has a specific number of locations across multiple states. The ultimate distribution equation
is one that satisfies both sides in the short term, while allowing for flexibility and room for
growth over the long term.

778
 SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

Developing a Universal Process

Six Sigma is an effective methodology to employ because it highlights issues that even the

Consider the case of a company that has multiple distribution channels. It sends out orders
to brick-and-mortar retailers; at the same time, it has to fulfill online orders, and it also

where employees can get confused by the many destinations of these orders, as well as the
various order sizes and specifications that are unique to each distribution outlet. In this
situation, the potential for things getting lost in the shuffle is extremely high.

Using Six Sigma, the company can develop one universal process that works in all of these

way. Even if certain retailers call for certain specifications, the vast majority of the order
preparation can be accomplished in line with all other orders, and any unique tailoring that
must be done can be incorporated into the process. Having this process set in stone also
helps companies to work with retailers and manage the expectations of those retailers

Setting up a universal process accomplishes far more than simply making the lives of the

potentially opening the door for new distribution opportunities.

Identify the Best Retail Outlets

quite this simp

turning a seemingly great distribution arrangement into a net loss for the company.
Unfortunately, figuring this out in the midst of filling orders and developing relationships

always done, not wanting to decrease the number of products available on shelves.

A Six Sigma project gives a company the opportunity to take a step back and evaluate the

company. On the other hand, another retailer might yield more revenue for the company,
but ample supply cannot be provided due to commitments to other retailers.
779
 SIX SIGMA IN DISTRIBUTION

Six Sigma not only shows the best channels for a company to use in its distribution, but it
also provides the hard data necessary to encourage management to change its
philosophies. Without the presence of evidence, it would be difficult to get leadership to
consider alternative methods for distribution. However, when the potential benefits - as
well as the lost revenues associated with maintaining the status quo -
easy to demonstrate why change is necessary.

Another hidden benefit of employing Six Sigma in this way is that it can help the company to

products in the incorrect manner. Six Sigma can help the company to clearly see how
distribution should be handled, both in terms of revenue and in satisfying demand.

Determine ROI for Distribution Outlets

determine which distribution outlets
should receive these precious few items. While reach is a consideration in this situation, the
company is best served using the stream that will provide the biggest bang for its buck.
Without the opportunity for in-depth analysis, obtaining these figures is virtually impossible.

Using Six Sigma gives a company the ability to calculate its return on investment for each

what it wants out of its relationships with retailers and vendors. Going through this analysis
will illuminate the best distribution channels when it comes to exposure and generating the
most revenue per dollar spent.

Once the company knows how much its products are worth to a retailer, it can renegotiate
its terms with its distribution partners. This negotiation is a lot easier to stomach if the
company knows exactly what it should be getting, along with the potential benefits and
drawbacks of partnering with different retailers. This is an example of one Six Sigma project

Multiple Distribution Streams in One Project

In order for a Six Sigma project to be successful, it must have accurate measurements taken
across a manageable project scope. In other words, a large company cannot consider the
entire breadth of its distribution process, as this would prove to be unmanageable in terms
of data collection and process improvement.

780
 SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

This is problematic for companies that have multiple outlets for distribution. For example,
the parameters of online distribution are completely different from those of brick-and-
mortar storefronts. A company that wants to improve its efficiency and return on
investment across the board may find it difficult to narrow its focus on just one small area
of distribution. They would find it even harder to extrapolate any new implementations to
their other means of distribution, since there is no guarantee that the new model could
work on multiple platforms.

Part of the preparation for a Six Sigma project is to use root cause analysis to identify the
heart of the problems that plague a company from a distribution standpoint. The company
can then isolate the area with the biggest problem and the most lucrative potential benefit.
Once the project is complete, the company can evaluate whether the solution would work
with other distribution streams, but they should understand that each outlet is different.
What works in one area is not guaranteed to work in another.

Dollars Versus Brand Image

Using Six Sigma to identify the most beneficial revenue streams, as well as the ROI incurred
when accessing those channels, is an endeavor any company would be wise to employ.
However, there may be a disconnect between the findings of the Six Sigma project and the
brand image that the company wants to portray.

A company that produces merchandise aimed at high-end customers might not see great
returns in working with specific retailers that fit their key demographics. However, in order
to reach this desired audience, the company must continue making its products available at
these retailers. Just the same, the company might not want to sell its merchandise at more
common retail outlets, even if their return on investment is higher. The prestige factor is

to control the image of their products and their brand as a whole.

Six Sigma can help companies to find a happy medium that enables their products to
maintain a certain status without putting the business in a financial hole. The problem arises
ir products must
be showcased in a specific way. This is just one example of why poor distribution
agreements continue to exist in spite of data that shows better options for the company.

Limited Control in Third-Party Arrangements

ace makes it easy for both large and small companies to gain
exposure. Sites like Etsy, Amazon and Sears allow companies to sell on larger platforms,
which in most cases is a mutually beneficial relationship. The retailer is able to offer a larger

781
 SIX SIGMA IN DISTRIBUTION

selection of products, while the company benefits from a far greater distribution platform

However, this relationship can quickly turn sour if the retailer makes any changes that might
negatively impact the wholesaler. Changes in service fees or reduced royalties can
significantly impact the bottom line for companies that distribute through these third-party
sites. Even small changes, such as a site layout change or a modified ordering process, can
become catastrophic if customers react so negatively that they stop making purchases.

The impact of these changes sheds light on the amount of control that is ceded to retailers
when companies distribute in this way. Retailers will always aim to maximize their profits
while minimizing costs, and their every action reflects this mindset. Modifying a relationship
with a manufacturer might only mean a small increase in revenues for a company like
Amazon, but for a small wholesaler, such a change might put that business in financial peril.
Worse, since the cost of losing a distribution outlet like Amazon is far greater than that of a
diminished revenue stream, the small business has no choice but to comply with these
changes.

ven the best laid plans can be

completely derailed by the whims of a third-party retailer. The Six Sigma project can and
should include the development of contingency plans that can help a company to anticipate
and react to changes made by third-party sites. Unfortunately, even creating these plans is
no substitute for a controlled distribution environment that allows the company to plan for
the future with confidence.

Based in Butler, Indiana, Therma Tru Doors has been a leading force in the door industry for
more than 50 years. (http://www.thermatru.com/company/overview/) In 1982, Therma Tru
Doors unveiled the first-ever fiberglass door to the public. Over the years, Therma Tru Doors
has continued to innovate and expand its audience, enabling the company to reach its

In 2001, Therma Tru Doors began experiencing difficulty in shipping products from the

Indiana.
(http://www.lasaterinstitute.com/casestudy/Logistics%20&%20Distribution/Six%20Sigma%

782
 SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

20in%20Logistics%20&%20Distribution.pdf) After an initial review, it was decided that the

utilization of trailer space in the Mexico warehouse was the main culprit. A Six Sigma
project was initiated in order to identify ways in which the trailer space issue could be
resolved.

Christopher B. Anderson and his project team set out to obtain baseline numbers and try to
improve those figures. At first, the utilization in the average trailer was 57 percent. In other
words, more than 40 percent of the space in a given trailer was unused. While the obvious
goal would be 100 percent utilization, Anderson and his team decided that a reasonable
goal for the project would be 90 percent utilization. This represented an attainable goal that
would still leave room for further improvement.

To assess the situation, Anderson sought out input from the very people who are tasked
with loading the trailers on a daily basis. These employees had great ideas for how the
process could be more efficient, but they simply never had the opportunity to make their
voices heard. Additionally, Anderson noticed that the large racks used to transport glass

took note of the fact that some items were shaped in a manner that made it hard for them
to reliably be stacked or organized within the trailer.

Improving the process required some outside intervention. Anderson recruited a

warehouse supervisor from Oklahoma, who had already undergone his own trailer space
utilization project, to come to Mexico and make some suggestions. One of his key
contributions was the suggestion that the company purchase additional pallets that could
better fit the various shapes and sizes that previously made the trailer hard to organize. This
approach allowed for more pallets to fit on a trailer, enabling the company to transport
more materials while using less space.

Solving the issue of the glass racks posed a bigger challenge. Not only did these items take
up more space than anything else, but they were also extremely fragile. However, the
solution for this problem came in-house. Anderson asked the staff-level employees if they
had any ideas for how glass could better be organized and transported. The employees
stated that the racks could be reconfigured so that they could be double stacked, effectively
doubling the amount of space available on each trailer.

The numbers showed a slight decline in trailer space utilization during the first three
months of implementation. However, this dip in productivity soon gave way to a massive
increase in utilization. Within six months of identifying the solutions to these issues, the
Mexico trailers experienced a 99 percent utilization rate. This was far beyond the goals of

783
 SIX SIGMA IN DISTRIBUTION

the project. A process for monitoring productivity and maintaining control was established
so that the utilization would never go below the stated goal of 90 percent capacity.

The distribution implications of this project were twofold. First, Therma Tru Doors was able
to increase the productivity of the average trip from Mexico to Indiana. This meant that
more goods could go onto shelves more quickly, resulting in additional revenue while
keeping costs at a minimum. Furthermore, employees at the Indiana plant had a better idea
of what they could expect when opening a trailer from Mexico. This allowed them to take a
more organized approach to their jobs, which also translated into increased efficiency.
Although distribution was the main focus of this Six Sigma initiative, the impact of the
project touched many departments in a positive way.

distribution, it is still a valuable tool that can help any company to become more efficient
and knowledgeable about the true performance of their distribution channels. A company
that carries out a Six Sigma initiative will develop an iron-clad process that will not only help
internal work flow, but it will also help companies to understand the nuances of their
operations when negotiating with retailers.

The best way to incorporate Six Sigma is to do so with a clearly defined project scope. It

project. Careful analysis should be done to uncover areas of subpar performance, and those
areas should receive the focus of the Six Sigma project.

initiative. As stated earlier, somet

financially rewarding. A company that has a clear understanding of where it wants to go is
well-equipped to weigh the positives and negatives of each distribution partnership. This
enables the company to carry out its vision while attempting to create workable
relationships with its desired retailers.

Communication is key in all aspects of business, and this is certainly true in distribution.
Those involved in the Six Sigma project must inform other departments about the project
and clearly state how they may be affected by the initiative. These stakeholders should also
come to understand that the project will ultimately help everyone, and each group should

784
 SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

be informed about how the unique benefits they will receive as a result of their
cooperation.

Source

http://www.thermatru.com/company/overview/
http://www.lasaterinstitute.com/casestudy/Logistics%20&%20Distribution/Six%20Sigma%2
0in%20Logistics%20&%20Distribution.pdf

785
There is perhaps no more important discipline in the entire retail industry than
manufacturing. Simply put, a poor manufacturing process means that products take longer
to make, and the quality of those products is questionable at best. As a result, fewer
finished products reach the marketplace, reducing the revenue brought in the by company,
which in turn causes the company to make cuts in other areas. The trickle-down effect of a
manufacturing process rife with defects, or even occasional inefficiency, can turn a
successful business into a struggling one without warning.

On the other hand, an effective manufacturing system opens a variety of doors for the
company. If the manufacturing team operates at maximum efficiency, the company will
experience savings in terms of both labor and resources. It also enables the organization to
take note of any defects that may enter the usual process.

The typical manufacturing process involves taking raw materials and turning them into
finished products. While this task is usually aided by machinery, manual intervention is
necessary for the machines to operate successfully. For instance, machines may need to be
cleaned, while others have to be reconfigured to execute another step of the manufacturing
process. When the process is complete, products assembled during a given time period will
be placed into a specific batch so that the company knows what products were made during
at various points in time. A sample from each batch is tested so that the company can
identify any issues before those products are shipped to retailers. If there is a problem with
one of the samples, the entire batch will be inspected, and other batches from similar time
periods may also be reviewed to ensure that the defects did not spread into other batches.

In order for manufacturing to go smoothly, every aspect of the process must be carefully
planned and documented in detail. This is the only way to keep tabs on underperforming
areas of the process. Many manufacturing processes experience bottlenecks - steps where
production significantly slows down, creating a backlog that adds
time. The ultimate goal of Six Sigma is to spot these bottlenecks and find solutions that
allow the entire process to function smoothly.

786
 SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

Examine Existing Processes

The initial phases of any Six Sigma project require the project team and its shareholders to
define the scope of the project and review the existing processes that have been in place.

imperative to review every facet of each action taken by the company.

he
biggest difference. Something as small as the need to clean a machine on a regular basis

ix Sigma project may

uncover that a more thorough cleaning done less frequently can increase production
significantly.

These discovery steps highlight the need to break down every aspect of the manufacturing
process. Six Sigma can help companies eliminate inefficiencies and redundancies, but it can
only do so if these smaller parts of steps are documented and examined. Some employees

reluctant to talk about aspects of the

However, by discussing these issues, they can take a hands-on approach in developing a
better process that not only increases productivity, but makes the jobs of staff-level
workers easier.

Revamp Processes to Increase Productivity

One of the most difficult parts of monitoring the quality of a manufacturing operation is

fix that job. Because the company requires that products continue to be produced, shutting
down operations to make an efficiency-based fix is virtually impossible. This is why so many

the cost of increased productivity.

Six Sigma gives companies the impetus to make large changes to processes where
necessary. When a proposed change is presented to the stakeholders of a Six Sigma project,
there is supporting data that suggests how much the company might benefit. This can help
decision makers to consider the true impact of the change. Even if a short-term decrease in
787
 SIX SIGMA IN MANUFACTURING

productivity occurs due to the reworking of the process, the overall benefit will compensate
for the diminished efficiency many times over.

Of course, there is more at stake than simply overseeing a more efficient process. A process
with a shorter lead time will naturally yield the creation of more products than the previous
method of operation. Over time, this increased haul can lead to a much greater amount of
products created, giving the company the ability to either sell the excess amount or save on
labor costs by reducing the number of hours the manufacturing plant operates each day.

Identify Cost Savings Opportunities

The big benefit of creating extra products is the financial impact of additional inventory on
hand at virtually no cost. Since these additional items will be made using the same labor and
mechanical resources as the typical load in the previous system, the only added cost comes
in the form of raw materials. This increased productivity can have massive implications for
the company.

As is the case with any business, internal improvements that save money should end up in a
less expensive product for the consumer. If a company finds a way to make a cheaper

the customer is a great way to gain market share and provide the company with additional
revenue - income above and beyond what the company would generate with the old
process.

The average staff-level employee might not be too concerned with the price the customer
pays for a product he or she produces. But seeing the big picture is an important part of any
Six Sigma initiative. If the company is able to show manufacturing employees how their
money-saving efforts can benefit the business - and ultimately, themselves -
much more likely to support the project.

Diverse Array of Processes

To an outsider, manufacturing is often thought of as a singular function. However, as
anyone who has ever seen a manufacturing process up close knows, the steps required to
create each unique product is different. Even multiple product lines of the same item that
have differing features can have unique processes.

788
 SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

This presents a problem as it relates to Six Sigma. Since the typical project scope of a Six
Sigma initiative is limited to one process, a company with multiple products - and multiple
inefficient manufacturing processes - may have difficulty drawing conclusions on a large
scale. In other words, the solutions they find as a result of the project may not be able to be
extrapolated to other manufacturing processes.

In most cases, since at least one flawed process can be overhauled, the project is still worth
undertaking. While the entire solution cannot be replicated in every other process
performed by the manufacturing team, the lessons learned along the way may be
applicable to other activities that need improvement.

Outdated Machinery
Six Sigma can accomplish many things, but the analysis by itself is unable to transform an
organization. It takes strict adherence to the recommendations of the project to see results,
and sometimes the implementation of a revamped process requires an investment of
financial capital. And even if a business is able to hire a project team to run a Six Sigma

through to its ideal conclusion.

to the prohibitive cost of upgrading machinery, many companies try to get by with the
ct might
identify this old machinery as the bottleneck for a broken process.

Therein lies the dilemma. The project makes it abundantly clear that the company will
benefit tremendously from upgrading its machinery. The employees, tired of using old
equipment to carry out modern tasks, are excited to see this new technology that will make

afford to buy new machines.

In this case, the impact is a two-pronged effect. Because the old equipment has been

upgrade is made. At the same time, employee morale is likely to take a dip as staffers
soon. The company is then presented with a
serious dilemma - either go on with the process as is and continue to experience difficulties,
or go into significant debt to buy a machine that may not pay for itself for many years.

789
 SIX SIGMA IN MANUFACTURING

Although the project has achieved its preliminary goal of identifying the weak link in a

manufacturing companies take every time they embark upon a Six Sigma project.

Encouraging Buy-In
From an executive level, implementing Six Sigma in a manufacturing facility always seems

realized from executing a project and improving an inefficient process. However, at lower
levels, employees might not be so enthusiastic about the idea of a Six Sigma project.

One of the main aspects of any Six Sigma initiative is that it must receive complete buy-in if
the project is to be a successful one. Just as a process is only as good as its weakest link, a
Six Sigma project is only as effective as its least interested participant. And in an industry
-level employees manning the machines are
the people with the real power in any project.

The Six Sigma initiative can only work if the lowest-level workers want it to succeed. While it
might not be obvious to the executives, these staffers may have some good reasons why
ourcing and downsizing are
feared in every organization, manufacturing employees may be afraid that a more efficient
process could mean that their services would no longer be needed. That may be all the
motivation these individuals need to rebel against this initiative that seems far more like a
threat than an asset.

result of a Six Sigma project will ultimately be to their benefit. Additionally, these workers
need to understand that not only are they needed after the solution has been
implemented, but they will also be an important part of creating the solution. At the end of
the day, the ultimate goal of a Six Sigma project based around a manufacturing process is to
make the jobs and lives of those staff-level employees executing the process easier. The
- and the more likely

A manufacturing company based in the Midwestern United States was plagued with an
inefficient manufacturing process. (https://www.isixsigma.com/implementation/case-

790
 SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

studies/case-study-increasing-machinery-throughput-without-more-machines/).The
-hour around-the-clock operation that burned out
employees, many of whom had to work 12-hour shifts on the weekends.

The main role of the manufacturing employees at the plant was to clean machines that

as ten times per hour. Although the employees were skilled at identifying issues and
cleaning them, the company felt it could benefit from a more efficient approach, and it was
decided that a Six Sigma project would be the best way to identify ways the company could
improve its manufacturing process.

Under the existing setup, the company operated at an efficiency rate of 65 percent. The

me to 16 hours a day, creating two eight-

hour shifts that would enable employees to approach their jobs with a more positive
attitude. This model would also help the company to save money; for every one percent
increase in efficiency that was achieved, the business would save $250,000.

To start the project, the company collected data over several data points - the machine in
use, the type of product, the name of the individual operating the machine and the time of
day. After six months, the organization was able to begin to review its figures.

One of the things that was obvious right away was that there were significant differences
between the parties involved in the operation. Some types of products performed far better
than others, and the same was true fo
was to figure out how these data sets worked together.

The company ran a series of regression analyses to determine the relationships between
these factors. The first analysis showed that there was a strong correlation between
efficiency and the number of machines worked by each operator. In order to reach the goal
of 70 percent efficiency, each operator could only work on one machine at a time.

Another regression analysis was run, this time incorporating all of the data points collected
during the Measure phase of the project. While several combinations of variables proved to
be statistically significant, the two biggest factors were the time of day and the machine
operator. Further investigation showed that the main reason for time being a contributing
factor was the drastic drop in productivity that occurred during shift changes twice each
day. Additionally, the project showed that each operator tended to work best on their

791
 SIX SIGMA IN MANUFACTURING

preferred machine; however, since operators generally worked on whatever machine was

Given the data, the solution was a rather simple one. Operators would be assigned to the
machine on which they were most efficient. This one change not only brought efficiency to

percent after a six-week implementation period. Because of the change, the company was
able to save $1.7 million, and the employees had the opportunity to enjoy the more relaxed
schedule that they had earned.

This case study is a classic example of a situation in which many companies would have
been tempted to invest money in new machinery. However, investing in human capital can
often be just as effective, if not more so, than simply buying new machines. Companies
should consider all options when making improvements, especially those that allow their
employees to shine.

Manufacturing is a vitally important aspect of the operations of any retail-based business.

-run manufacturing
team can make the entire company more successful through cost-saving measures,
increased quality and a larger inventory of products that the company can sell.

When implementing Six Sigma, choosing a manageable and purposeful scope for the project

successful conclusion will yield positive insights that can impact other divisions of the
manufacturing operation. These findings will make subsequent Six Sigma projects within
that company easier to manage.

Getting staff-level manufacturing employees involved early and often will help them to take
ownership of the initiative. The success or failure of the project lies in their hands, and
-

Tracking all data is a major

manufacturing industry, where something minor can end up being the bottleneck that
derails an entire process. No piece of data is too insignificant to track, and as the case study
showed, no potential solution is too simple too consider.

792
 SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

Source

https://www.isixsigma.com/implementation/case-studies/case-study-increasing-
machinery-throughput-without-more-machines/

793
Food Service

Nobody wants to cook every meal. To those people who want a night out, or to those who
home-cooked meals, the food service industry is a godsend.

The United States Department of Agriculture defines food service as any meal or snack that
is prepared for immediate consumption. (http://www.ers.usda.gov/topics/food-markets-
prices/food-service-industry.aspx) While items can be brought home, the majority of the
meals consumed as a result of the food service take place at commercial venues like
restaurants and fast-food outlets.

There are many different types of products offered by the food service industry. Cafeterias
-tasting options, but they do provide
affordable meals for people without disposable income. Basic fast-food restaurants take

types of restaurants. Sit-down restaurants vary from family-friendly and basic to elaborate
and very expensive. The mix of price point, quality and customer experience varies
depending on what type of message the restaurant wants to send to its patrons.

The type of restaurant has a significant impact on the way the business is run. While fast-
food restaurants take for granted that people will come and go quickly, sit-down
restaurants must manage additional factors related to customer flow. Table turnover, wait
times and the potential for additional courses being ordered by a given party all interfere
with the normal flow of operations. When a restaurant of this type experiences a dip in
attendance or revenue, it can be difficult for them to understand exactly why their numbers
are changing. Careful analysis is required to determine if the change is related to the food,
the customer experience, wait times or some other factor.

The food service industry has had to change with the times, navigating the always fickle
waters of customer tastes and preferences. The modern climate poses perhaps the biggest

ever in terms of the health impact of food, as well as matters relating to the quality of food.
People are increasingly demanding fresher ingredients and higher quality meals. While
some are willing to pay a premium for better ingredients, many want better quality at the
794
 SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

same cost. Additionally, specialized diets have become commonplace in modern culture.
-free, vegan or organic options may alienate segments
of the marketplace.

for companies to know where they should focus in order to improve. Six Sigma can be
tremendously helpful in helping a business understand what they do best and how they can
improve their products, as well as their internal processes.

Ideal Serving Size

Serving size is just one of many things that the average cu
but poses a major problem for food service operations of all sizes. In fact, even some

true cost of their average portion.

Serving size can be a competitive advantage for some food service companies. Restaurants

This can also be a form of advertising for them. Knowing that people often bring leftovers to

experience can serve as promotion for that restaurant. However, the restaurant might not
know for sure that this is the right way to go. They may be shocked to find that this
approach could actually lose money. On the other hand, a food service location that offers a
smaller portion will likely never know that its inadequate serving size is alienating its
customers. Instead, those people will simply never return.

The only way to know for sure is to analyze the serving size in relation to its cost to produce,
the price the customer pays and the potential benefits of a smaller or larger portion size. Six
Sigma can be very useful in this area. Instead of putting off this analysis, as many companies

concrete data that shows exactly how each unique restaurant should handle the serving size
dilemma. In addition to simply running numbers, Six Sigma gives the company the chance to
talk directly to its customers to see what they really think. Between the internal number-
crunching and customer feedback, restaurants have a great chance to settle on an ideal
portion size that satisfies customers while maximizing revenues.

795
 SIX SIGMA IN FOOD SERVICE

Minimize Food Waste

A 2005 study conducted at the University of Arizona stated that nearly 10 percent of all
food in fast food restaurants was wasted, while more than 3 percent of food in sit-down
restaurants ended up in the garbage .
(http://www.endfoodwastenow.org/index.php/issues/issues-restaurants) Not only is food
waste a big problems in terms of corporate responsibility and environmental concerns, but
it can also seriously cut into the margins of the food service industry.

An estimated 4 to 10 percent of food becomes wasted before it ever reaches the

-preparing foods on a

prepared are also reasons for food waste, but the difference is that these two categories
are preventable.

Like all businesses, the food service industry is plagued with issues that it would like to

plagues most restaurants. Problems like spoiled food may be easily solved by ordering in
smaller quantities more frequently, but the average restaurant is too consumed with filling
current orders, leaving it unable to consider other possibilities.

Implementing Six Sigma gives these companies the impetus to begin considering the
reasons for their food waste, as well as potential solutions to these problems. One of the

problem might be. Instead, Six Sigma incorporates the input of the people closest to the
process - the people who work with the very food that ends up spoiled - and allows those

it can help a company significantly cut into its food waste dilemma.

Freshness Initiatives
Today, everyone wants fresh ingredients and high-quality food items. The problem is that as
the level of quality goes up, so does the cost. If all of these costs were passed on to the
customer, even the most loyal patron would be taken aback and begin to look elsewhere.
the restaurant to find its own internal way to get the best ingredients in a cost-

This is an area in which many restaurants have been hesitant to change. After all, their
business works right now using their current level of quality. Why scrap something that

restaurants and food service outlets to understand that the lack of quality is the reason. All
796
 SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

they know is that they have fewer customers than they once did. Worse, even if the
-quality food.

revenue is affected. Six Sigma can help restaurants find ways to bring their food quality to a
reasonable level without drastically increasing costs. Better yet, Six Sigma may be able to
help zero in on exactly what aspects of a food service operation need a quality makeover,
eliminating the need to improve everything all at once. This makes the change more
manageable and more affordable, meaning that customer prices will remain affordable.

Challenges of Using Six Sigma in the Food Service Industry

Changing Menus and Public Tastes
The emerging trend of customers preferring fresh ingredients and healthy options is a
microcosm of what food service establishments have had to deal with for the past century.
Restaurants, fast-food chains and even school food courts have had to recognize trends and

hit to the bottom line.

consumer, especially considering how frequently those tastes change. Six Sigma is highly
effective in terms of shoring up internal processes and making sure that all processes are as

who suddenly finds itself rushing to add healthy options to a menu full of cheap fast-food
items.

Food service restaurants must always stay close to the tastes of their customers so that they

magically predict customer preferences, it can certainly help a restaurant become efficient

Unskilled Labor and High Turnover

While there are some servers and managers who have aspired to be full-time employees,
much of the food service workforce is young and unskilled. Those who are older, yet still
remain working non-management roles in fast-food restaurants, largely do so because of a
lack of other options. The average age of fast-food workers is 29, and of the adults in that
group, many receive public aid.
(http://www.politifact.com/punditfact/statements/2013/nov/04/bill-maher/bill-maher-
average-fast-food-worker-29-most-are-pu/)

797
 SIX SIGMA IN FOOD SERVICE

-food industry is barely above minimum wage.

Servers at sit-down restaurants have the potential to make more money, but the vast
majority of their income is derived from tips from customers. One or two slow days in a

many are so quick to abandon this industry; anyone who sees an opportunity to make more
money and learn something new is usually very willing to take that chance.

Because of this high rate of turnover, and the unskilled labor that typically fills lower-paying
positions in the food service industry, Six Sigma is hard to implement. Six Sigma relies

establishment. Adding to the problem is that new employees, particularly those who might
not be staying very long, might not be inclined to buy into the proposed changes. A 16-year-

making enough money to support their hobbies.

The increasing number of full-time adults who work in fast food provides a glimmer of hope
that Six Sigma can help food service restaurants of all types. However, until turnover
.

Public Relations Crises

have undergone Six Sigma projects find themselves disgusted to find out about the true
conditions of their workplaces and processes. However, most of these businesses have the
ability to keep these findings under wraps so that the public never knows just how dire
things really are. The worst case scenario is a product recall, in which anyone who has
purchased a faulty product receives a new one free of charge.

This cannot be said of the food service industry. If the public becomes aware of any issues

employees, the perception of that restaurant is irrevocably damaged. A rat spotted in a fast
food establishment is enough to turn people off of that restaurant forever. Worse, if one
branch of a chain restaurant is found to have an issue, all of the other branches feel the
impact.

Adding to the problem is the fact that restaurants undergo safety inspections on a regular

business will suffer drastically - even if the Six Sigma project was working on fixing that very
flaw.

798
 SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

trust. Ideally, Six Sigma should be implemented before a preventable occurrence does
irreparable damage to the restaurant, as well as the other locations related to that
establishment.

A Taiwanese food company specialized in producing pork buns for mass consumption.
(http://www.academicjournals.org/article/article1380706320_Hung%20and%20Sung.pdf)
The preparation required for the buns to be steamed, then frozen. The frozen buns were
then sold, after which the buns had to be re-steamed in order to be served. However, after
re-steaming, people found that the buns were often defective. Common defects included
shrinking, cracks in the buns and additional objects being found in the bun. The company
decided to use Six Sigma to find ways to reduce the defect rate.

The company began its analysis by trying to identify which of the buns was the most
problematic. While all of the buns the company produced suffered from defects, the small
custard bun
defects. Additionally, while there were various types of defects, the one that was most
common was shrinkage. Therefore, the company decided to focus its energies on fixing the
small custard bun so that it would not shrink nearly as often.

Early analysis showed that the baseline defect rate for the operation was 0.405%. The goal
of the Six Sigma project was to reduce this number by 70 percent to 0.141%. Once all
stakeholders came to an agreement regarding what constituted a defective bun, the
company could begin measuring and analyzing data.

The company chose nine main data points to track. The main analytical tools used were
ANOVA analysis and main effects plots. Of the nine factors tracked, five proved to be
statistically significant - temperature, color of dough, shift, amount of ice water and
steaming time. Subsequently, factorial analysis was used to determine the main culprits in
causing shrinkage. The analysis showed that the volume of ice water was the biggest reason
for the shrinkage, at a rate more than double that of the next closest factor.

Using the data collected, an optimization tool was used to calculate the best possible recipe
that would incorporate just the righ
to click. In the six months that followed the implementation of the revamped recipe, the
company only exceeded its goal of a 0.141% defect rate once. In fact, the other five months
799
 SIX SIGMA IN FOOD SERVICE

were well below the goal rate. Encouraged by the results, the company carried out further
Six Sigma projects to further improve its performance in the food service industry.

The food service industry can be a difficult place for revamping processes and planning
ahead. However, Six Sigma can be highly beneficial for those who use it wisely.

Because inefficiencies can have a wide variety of causes in food service, considering all
possibilities when implementing Six Sigma is advised. Brainstorming is a great idea for
generating ideas as to why a restaurant is underperforming. As the case study shows,

Getting the real scoop about a process is imperative for success in Six Sigma. Usually, the
higher-
floors. Speaking with the lower-level employees that deal with food and customers on a
daily basis will yield the best insights ab

Defining a manageable scope for a Six Sigma project can help an establishment to fix one
very important problem instead of getting bogged down by every problem that the
restaurant faces. Fixing one major flaw can have a trickle-down effect that impacts other
processes in a positive way. Identifying the aspect of the business that can benefit the most
from one change is key for success in Six Sigma.

800
Construction is one of the most valuable and lucrative trades that exists. There are very few
people equipped to perform the tasks required of construction workers, which makes
working with the construction industry a necessity for many people around the world. Even
laypeople with technical know-how are often more comfortable hiring someone to do the
job right the first time, not wanting to risk making a costly mistake.

Construction comes in a variety of forms. Most people immediately think of workers on the

more to construction than road work. Construction also covers the building and
restructuring of homes and businesses. Adding onto existing buildings, fixing internal
structural errors and refinishing outdated elements of those buildings are just a few of the
tasks that fall under the jurisdiction of construction.

Many construction employees work for contractors, which are project leaders that are well-
versed in communication, workflow and time management. Contractors must also purchase
and transport the raw materials that are necessary to get the job done. Additionally, they
must deal with customers and adhere to rigid time deadlines that may change during the
duration of the project. Ultimately, if something goes wrong with the project, even months
after its completion, the contractor is the one blamed for the error.

The construction workers who handle the nuts and bolts of a project are skilled in manual
labor. However, they also perform high-
commensurate with their work. According to U.S. News & World Report, the median
construction worker salary in the United States is less than $32,000.
(http://money.usnews.com/careers/best-jobs/construction-worker/salary) Construction
workers usually become members of unions, through which they receive their benefits.
Given the health risks and potential long-term health ramifications associated with
construction, good health benefits are essential for any construction worker.

The scope of construction jobs vary from site to site and from region to region. The bigger
job, the more money paid to the contractor. Construction jobs can be as small as knocking
down a wall in a basement, or they can be as big as building a billion-dollar sports stadium.
No matter the size of the job, any construction business could benefit from being more

801
 SIX SIGMA IN CONSTRUCTION

efficient. Six Sigma is able to help the companies in the construction industry in their quest
to reduce defects and increase productivity.

Reduced Labor Costs

Aside from doing a good job, the ultimate goal of any construction job is to get done on
time. Running past schedule has many ramifications, such as added costs and a damaged
reputation for the contractor. An organized workflow is necessary to ensure that the job is
completed on schedule, if not ahead of time.

Incorporating Six Sigma can help contractors and construction companies to streamline the
planning process. Collecting data from previous construction jobs is a vital element in
creating plans for future projects. When each member of the team is in a role he or she
enjoys and excels in, the entire operation benefits. The work is done more safely, more
efficiently and with more care.

put these pieces in place. Six Sigma will not only help with data collection, but it may also
recommend new planning ideas that will enable contractors to schedule jobs more
efficiently. This added efficiency will increase productivity, not to mention the quality of the
work.

Furthermore, seeing the proper alignment of a labor force may highlight areas where labor
is no longer needed. Perhaps there are some redundant processes, or maybe the work
e savings can help the
business to purchase new machinery or provide more flexible schedules to employees. The
surplus money can also encourage the contractor to offer lower prices for better quality
work.

Reduce Injuries
Safety is paramount in the

wherever possible.

802
 SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

Injuries impact a construction crew in many ways. First and foremost, suffering an injury is
devastating for the employee, who now has to miss work and may be facing a lengthy
ensation benefits,
which is not always an easy task. The injury also has a residual effect on the other members
of the team, who might find themselves a little less enthusiastic about engaging in the risky
parts of their job. It also affects the contractor, who not only has to soothe the concerns of
the injured employee, but must also find an equally effective replacement capable of
working with the existing team. After all, the job has to get done by the guaranteed end

The use of Six Sigma can help construction companies to go further than simply installing
safeguards. Now, the company and the Six Sigma team can work together to rework
processes as much as possible in order to mitigate the amount of risk undertaken by
construction workers. As is the case with all Six Sigma projects, data analysis will be
accompanied by insightful feedback provided by the very workers who perform dangerous
tasks on a daily basis.

Due to the nature of the industry, not all risky activities can be removed. However, the
company can incorporate Six Sigma in order to make the job as safe as possible for its
employees.

Create Defined Processes

One of the more frustrating aspects of managing construction is that most of the people on
a given construction crew have different work backgrounds. This means that each of them
learned how to do their jobs differently. While this might not seem like a big deal on an
individual basis, having an entire team of people that work differently from each other can
create massive headaches when laying out processes.

This disconnect can also result in jobs being inadequately completed, with the mistakes only

ways of doing their w -

workers. Additionally, there are sometimes language barriers, particularly between an
English-speaking contractor and a Spanish-speaking crew. When such miscommunications
are downplayed, the
or diminished productivity from employees.

These conflicts highlight the need for defined processes that are largely set in stone. When
they know exactly how to execute

803
 SIX SIGMA IN CONSTRUCTION

management. Additionally, laying out processes in this way makes it easier for contractors
to bring in new labor after someone leaves a project.

must be adhered to. Fortunately, Six Sigma allows for the necessary analysis and reworking
to take place. Temporary dips in productivity are offset by periods of sustained efficiency.
Solutions that may take time to implement are accompanied by supporting data showing
why change is necessary. Most importantly, a new process from which everyone involved
will benefit is created, creating a workflow that eliminates many of the barriers that
prevented maximum efficiency.

Cost of Machinery

is second to none. The only problem is if the bottleneck is proven to be a piece of outdated
machinery. This puts the entire project, as well as the company who commissioned the
project, in a tough position.

n
actionable plan that would help the company to better structure its construction jobs and
lower costs. Instead, the leadership team receives a bill for a new machine that could cost
tens of thousands of dollars. And although the new equipment would help
simply not an option for a cash-strapped construction company.

This puts the business in a no-

new piece of machinery, or they go without the new equipment and continue to operate at
a suboptimal level. Furthermore, the company has to deal with disillusioned employees who
thought they were getting a great solution, only to find out that the status quo will continue
to remain intact. This can have a subsequent impact on morale, potentially making
productivity even lower than it was before the project began.

Six Sigma projects always aim to provide the best possible solution at the lowest cost.
-prohibitive for the
business.

804
 SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

Difficulty Testing New Processes

Even the best Six Sigma processes go through a great deal of testing, as well as a period of
expected inefficiency while the solution is being implemented. Though this may work in
o carry out in construction.

Construction jobs come with high costs and big expectations. For the price paid, the person
hiring the construction team wants the job done their way, as quickly as possible. That
person would not be very happy if he or she was told that the construction team was

likely be very quick to hire a new team altogether.

In such a high-stakes business, nobody wants to be the guinea pig, and understandably so.
That puts extra pressure on the Six Sigma project team, as well as the contractor and his or
her employees. No matter what solution is presented as a result of the project, it has to
work, and it has to work the first time. If the solution fails, the project has failed. Worse yet,

with significant costs, not to mention the aggravation that comes from dealing with an
angry customer.

Whenever possible, a contractor should arrange to have a revamped process tested on a

small scale before making it part of the regular routine. This in itself might cost money, but

Increased Rates
Sometimes, Six Sigma can unwittingly have an adverse effect on a company. A business uses
Six Sigma to improve its processes; as a result, additional quality measures are implemented
to ensure that the job is done quickly and effectively the first time. These added parts of the
process, while helpful in ensuring quality and efficiency, force the company to increase the
cost of the product or service.

In many cases, the recipient of this product or service is happy to pay the additional money,
kn
construction. People want a good job done, but they also want to pay as little as possible. In
the end, money is usually the factor that wins out.

A construction company that incorporates Six Sigma and irons out its broken processes
should expect to find itself in a great position. And yet, it may find itself unable to attract
new business because of the new price tag associated with its services. Even though the

805
 SIX SIGMA IN CONSTRUCTION

company does better work than ever and its employees are happy and safe, nobody wants
to hire this construction team simply because someone else will do the job for less money.

safe
construction company is its own reward. Over time, enough work will come in to
compensate for being passed over by cost-conscious customers. That said, the initial
rejection is bitterly disappointing for companies that have strived to improve their
processes.

A construction company was hired to build 405 villas in Saudi Arabia.

(http://www.iglc20.sdsu.edu/papers/wp-content/uploads/2012/07/102%20P%20033.pdf)
Since the project required the company to build so many of the same structure in a short
period of time, it was decided that pre-constructed walls and floor panels would be used to
build the villas. A year into the project, the company realized that it was 25 percent behind
schedule. A Six Sigma project was undertaken in order to examine why the project had
fallen so far behind its expected level of productivity.

The Define process of the project showed that the focus of the project should be the
number of panels sent to the construction site each day. The initial measurement indicated
that 18 panels were being delivered, while the goal of the project was to increase that
number to 75 panels daily.

The company had many potential factors for change; these factors covered the scope of the
entire project. Instead of one massive change dictating the future of the project, the
company and the Six Sigma team made several smaller changes aimed at increasing the
production and delivery of the concrete panels.

The management of the project changed, with four foremen hired to oversee six areas of
production each. Additionally, a full-time project quality engineer was brought on to ensure
that the quality of the panels was consistent. Defects in the panels themselves would only
serve to add further delays and costs to the project.

To construct the panels more quickly, the company incorporated the use of a ready-mix
Cranes and a concrete bucket
were also purchased in order to facilitate the more expedient production of the panels.

806
 SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

Most importantly, a night shift was added in order to help the company to get back on
schedule.

In addition, the inventory system was overhauled. The stockyard was reorganized in order
to make it easier for employees to find the parts they needed. At the same time, the
inventory level was reduced to a maximum of 250 units per item. Electronic inventory
management was also implemented so that the company could easily re-order materials as
needed.

Lastly, the company took the time to clearly define the process, the stakeholders in each
part of the process and who was responsible for each aspect of the production. The
company also began measuring cycle times in order to implement benchmarking and
measure productivity.

Many changes were needed to make this process a viable one, both in terms of making it
efficient and getting back on the established schedule. Fortunately, the changes proved to
be h
panels per day to the construction site. Furthermore, internal control charts proved that the
process was indeed in control, proving that the solution would be sustainable for the
foreseeable future.

Given the diverse nature of projects undertaken, construction might not seem like the most
obvious industry that would benefit from Six Sigma. However, its need for constant
precision and productivity makes the construction industry a candidate for the process
improvements that Six Sigma is known for delivering.

Although the company in the case study used their Six Sigma project to make widespread
changes to their process, this is often the wrong approach to take. Looking for specific
improvements that will make massive change with minimal effort is the best way to
incorporate Six Sigma, particularly in a high-pressure, results-oriented industry like
construction.

In some cases, a Six Sigma project will indicate that investment into machinery is necessary
to improve efficiency and reduce defects. However, there are often other solutions that can
improve workflow without requiring such a large investment of capital. These alternate
807
 SIX SIGMA IN CONSTRUCTION

fixes might not present the improvements that the new machinery would have, but they
may come at a much more reasonable cost.

Each construction process is different, and even the best Six Sigma process improvement
might not be able to be replicated across other functions. However, there are always tips
and positive developments to be derived from a process improvement initiative. Finding
ways to incorporate the improvements made via Six Sigma into other processes can help
the company to become more well-rounded and productive, which results in a better
experience for everyone.

Source

http://www.academicjournals.org/article/article1380706320_Hung%20and%20Sung.pdf, ht
tp://www.politifact.com/punditfact/statements/2013/nov/04/bill-maher/bill-maher-
average-fast-food-worker-29-most-are-pu/,
http://www.endfoodwastenow.org/index.php/issues/issues-restaurants
http://www.ers.usda.gov/topics/food-markets-prices/food-service-industry.aspx
http://www.iglc20.sdsu.edu/papers/wp-content/uploads/2012/07/102%20P%20033.pdf
http://money.usnews.com/careers/best-jobs/construction-worker/salary

808
Six Sigma is primarily associated with the business world, but it has applications that reach
far beyond producing consumer goods. Six Sigma can help improve any process, no matter
how large. Even an organization as secretive and as tradition-based as the military, stands
to benefit from a Six Sigma project.

The military may be best known for physical combat, but the actual responsibilities of the
military go far beyond fighting wars. Above all else, the military is charged with protecting
its country by any means necessary. This means more than simply fighting when force is

before it happens so that the best plans for defense can be devised. For this reason,

these groups to keep the military safe from sneak attacks, ambushes and unannounced
strikes. Their role in aiding the military is paramount.

In addition to the more exciting aspects of the industry, the military must also deal with the
same issues as any other job category. Recruitment and turnover are just two of the more
difficult internal tasks the military must handle. Furthermore, the military must also process
a great deal of information, including the current load of individuals serving, new initiatives,
updated equipment and revised processes. Since the military is a subsidiary of the
government, any change is usually accompanied by a fair amount of red tape. Military
agencies must find ways to incorporate this lead time into their actions, which can
sometimes work to the detriment of the operations they are in charge of carrying out.

The military and defense units of any country are synonymous with obtaining key pieces of
intelligence, operating efficiently and identifying problems before they arise. Six Sigma can
help in all three of these areas, as well as countless others. In fact, Six Sigma is becoming
commonplace in military operations around the globe, becoming an important tool in
keeping the world safe. (http://www.villanovau.com/resources/military/six-sigma-for-
defense/)

809
 SIX SIGMA IN MILITARY/DEFENSE

Decreased Processing Time

Most civilians see the military largely as a combat-
inside look at what happens behind the scenes. People outside the military might be

is in any other industry.

While difficulty in getting things done is found in

comes to the military. In the business world, not being able to bring in a new hire for two

However, in the military, a lengthy delay can literally be a matter of life or death. Since
disaster can strike at any time, the necessity for deployment may be just moments away.

been properly tra

being ready. Instead, it puts the entire military operation - and the country defended by
that nation - at risk.

Six Sigma has been proven to assist with the processing of paperwork and speeding up of

-saving.

Improved Repair Time

As is the case in the civilian world, equipment in the military industry is in constant need of
maintenance and repair. From combat weapons to behind-the-
always something that needs fixing.

This poses two potential problems. First, and most obviously, the repair of broken combat
weapons or intelligence systems is imperative. The absence of up-to-date equipment in
these areas could be catastrophic. The second issue is not as obvious, but might be even
more important. There needs to be an internal system to categorize and prioritize the repair

come in. A high-level intelligence system used by the CIA is far more important than the
radar detector on a tank.

Additionally, the military needs to have a strong sense of assurance built into its repair
work. These repairs must work the first time, every time. Therefore, a fool-proof testing

810
 SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

system must be put into place so that military personnel can test repaired equipment
is also gives soldiers the assurance that comes with knowing
that their equipment is in good hands.

Using Six Sigma gives the military the tools necessary to organize its repair effort, as well as
reduce defects inherent in the existing process. A well-executed system for repairs saves a
great deal of time and money while simultaneously providing quality assurance for the rest
of the military operation. Implementing this approach is just as important as battle tactics,
intelligence briefing or any other function carried out by the military.

Identify Security Breaches

Defense intelligence operations are designed to identify potential security issues and to
thwart those threats before they become bigger problems. While every potential security
concern must be
prioritizing more serious threats, making sure that those crucial matters get the full
attention of the team.

This issue of security breaches highlights the importance of having a defined process set in
stone. It gives defense employees the best chance at success because it shows these
individuals exactly what they should look for each day. When everything is lined up in a
specific way at all times, and then something deviates from that usual pattern, employees
will know for sure that something is amiss. As a result, they will be able to take the correct
defensive action.

Incorporating Six Sigma can be very advantageous for the defense industry. Not only can it
help the unit to identify, track and stop threats, but it can help to pinpoint where suspect
activity comes from. This knowledge will help the defense to move beyond merely
responding to threats, allowing the department to develop policies and processes that will
prevent those threats from occurring.

Emergency Situations
Process improvement initiatives do a lot to benefit any company in any industry. However,
even the most helpful improvements take time to establish and implement. While it might
seem like this time is available at the start of a project, plans can change at any minute,
especially in a high-stakes industry like the military.
811
 SIX SIGMA IN MILITARY/DEFENSE

The military is trained to respond to emergencies, and any moment can bring about an all-
hands-on-deck emergency that requires immediate action. In the event of such a situation,

Although the military has become increasingly friendly towards Six Sigma, the reality

collection that occurs in Six Sigma projects is always helpful, the ideal solution that is
realized
that this is a risk that comes with the territory, and that they should weigh the positives and
negatives before beginning a project.

Ordered Initiatives from Superiors

In the business world, Six Sigma is utilized by a company that wants to become more
efficient, and that company is generally able to carry out a Six Sigma project at their leisure.
se of an
emergency that necessitates action.

have the final say in how things are run. For example, a newly elected president that strives
to focus on a specific area of military operation may supersede the desires of military
leadership. Even if a Six Sigma project yields data that clearly shows that a new process will
improve the workflow of the military, such a change can quickly be overruled by someone
with the power to get their way.

In this way, the military is just like a regular business when it comes to Six Sigma. The
numbers can tell many things, but without complete buy-in from all stakeholders, the
project cannot succeed.

Lack of Funding
Like all other branches of the government, the military must operate on a pre-set budget.
Once again, like many businesses, the military must find ways to achieve the most good
while using minimal resources. This is one area where Six Sigma can be particularly helpful.

Unfortunately, sometimes a Six Sigma initiative is impossible to implement because there

the long run, it does cost money to undertake the project. The additional resources incurred

812
 SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

decisions come from many levels above military leadership, and those numbers are derived
so that other branches of the government can have budgets of their own.

Adding to the problem is the potential for elected officials to slash the military budget.
These promises may be made before candidates are close enough to the situation to truly
understand why the military needs its funding. However, because a promise was made, the
elected official must find a way to stay true to his or her word, regardless of whether or not
- and regardless of whether Six Sigma initiatives have to be put on hold.

Despite the potential for budget cuts at any point, the military should always strive to
maximize productivity and reduce the occurrence of defects. Six Sigma projects should be
conceived and executed without worrying about the possibility of funding decreases. As is
the case with all military operations, anything can change at any time.

Due to the confidential nature of military operations, there are very few Six Sigma case
studies available that include information regarding improved processes or internal metrics.
However, Six Sigma does have an important role in military operations. The United States
military has greatly benefited from Six Sigma, and it has disclosed information about several
of its successful initiatives. These improvement projects include the following:

- The United States Army began its use of Six Sigma in 2005. Within three years, the Army
had completed more than 1,000 Six Sigma projects, with another 1,600 projects in progress.
The Army had trained 2,000 workers in Six Sigma methodology, with the total savings
incurred due to Six Sigma reaching nearly $2 billion. (https://www.minitab.com/en-
us/Press-Releases/Department-of-Defense-Saves-Billions-with-Lean-Six-Sigma-and-
Minitab/)

- The U.S. Navy started using Six Sigma in 2006. Between 2006 and 2008, the Navy certified
more than 5,000 officers as Six Sigma green belts or black belts. The total savings brought
about by Six Sigma projects conducted in 2006 and 2007 was $450 million, which
represented a 400 percent return on investment.

- The U.S. Air Force has also begun utilizing Six Sigma, boasting a Six Sigma staff of more
than 500 black belts. To date, the Air Force has completed several Six Sigma projects, and
the results have been so positive that the Air Force decided to reduce its staff of
813
 SIX SIGMA IN MILITARY/DEFENSE

officers by 40,000 employees, which was expected to be completed without any sacrifice to
the Air Force's current level of operations.

- The Red River Army Depot Repair Facility in Texarkana, Texas has undergone a series of
fruitful Six Sigma projects. The facility has used Six Sigma to increase its ability to produce
High Mobility Multipurpose Wheeled Vehicles (Humvees) by more than 7000 percent, going
from three units a week to 32 in a day. Furthermore, the cost of each vehicle has been cut
in half, with an $89,000 cost of production turning into $48,000 per vehicle as a result of Six
Sigma. The main tools used to bring about these changes were the implementation of an
assembly line process, improved time and personnel management and enhanced methods
for keeping parts in stock. Front-line staffers were particularly helpful in suggesting and
implementing these changes. In addition to the workflow modifications, the Red River
Depot began placing stickers on the job site reminding workers that their work ultimately
affects the lives of soldiers who fight to defend their country .
(http://www.dtic.mil/dtic/tr/fulltext/u2/a493547.pdf)

- The Pine Bluff Arsenal, located in Arkansas, used Six Sigma to find a better way to produce
and repair gas masks. Thanks to their improvements, the rate of gas mask production has
increased by 50 percent, while repair time has decreased by 90 percent.

- The Corpus Christi Army Depot saw Six Sigma drastically overhaul their previous process
for helicopter repair. Their Six Sigma project showed that the group was overhauling
engines far too regularly. As a result, the depot began overhauling engines after 900 hours
of work, as opposed to the previous baseline of 309. In the end, the overhaul time was
reduced by nearly 65 percent, while the reliability of the overhaul rose to more than 90
percent

The trend of Six Sigma usage in the military is a positive development. Although several
projects have returned outstanding results, the industry is still merely scratching the surface
of what Six Sigma can do to help. Future projects are sure to have major positive
ramifications that will benefit all aspects of the military.

Because Six Sigma has become so popular in the military, units are encouraged to use any
and all resources at their disposal when working on a project. Although budgets are often

814
 SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

subject to the whims of politicians, there are plenty of ways military personnel can improve
their own processes, as well as the functions of others.

overhaul everything at once. Choosing the proper project scope will help military units to
get the most out of their projects. Projects should be big enough to make a huge difference,
but not so big that management gets bogged down by the over-analysis of data and
processes.

onsider all possibilities when coming up with

solutions. Any process improvement, no matter how small, can provide tremendous benefit
to the unit and, by extension, to the entire country. Encouraging lower-level soldiers to
speak up and identify their pain points will help the project team to come up with solutions
that help everyone.

815
At first glance, the government might not seem like the poster child for Six Sigma
implementation. The
truly have to answer to its superiors or its constituents. And yet, in many ways, the
government is actually the perfect forum for Six Sigma process improvements.

ly calls to mind elected officials that are responsible for

helping the people in a given geographical area. These individuals are tasked with many
duties that are unseen by the average person. Organizing local activities and appearing at
community events barely scratch the surface of the responsibilities of an elected official.
Politicians must work with groups such as law enforcement and special interest
organizations in order to facilitate changes that are demanded by the citizens of their local
area. Additionally, any progress made by the politician is subject to further review by other
politicians, which results in many proposed changes never seeing the light of day.

The workflow of politicians would be less difficult if not for the non-stop stream of outside
considerations that interfere with their duties. Lobbyists and fellow politicians want support
for their own causes, which prompts them to make pitches to other elected officials to
garner support. As a result, politicians are forced to devote time, energy and money to

right thing to do from a political perspective.

Elected officials must also work diligently to raise money, both for their own re-election
campaigns and for the parties to which they belong. Cold-calling in order to request
donations is a harsh reality for many lower-level politicians, and fundraising dinners are the
norm for higher-level officials. These politicians are also required to acquire endorsements
from important people in the community, including newspapers and unions. With so much

on promises for the future in order to gain support, as opposed to showcasing their track
records from previous jobs.

While elected officials may be synonymous with the popular definition of government,
there are plenty of other people that keep the industry running. Everyday people that take
jobs working in the government also have a say in how well the industry serves its people.
Civilians working in the government have a wide variety of responsibilities, including serving
816
 SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

as court officials, DMV clerks, IRS auditors and countless other positions. The politicians in

initiatives.

The need for government employees to work with elected officials, and the potential for
disagreements along the way, highlights the need for Six Sigma in government. Additionally,
Six Sigma may be able to help politicians to manage the many demands on their time.

Eliminate Red Tape

While the government is commonly associated with being the ul

decision made by government members involves outside parties. Citizens may be asked to
vote on certain changes, while others simply require approval by higher authorities. These
added steps add significant lead time to the execution of approved items, often resulting in
added costs and unhappy citizens. (https://www.isixsigma.com/industries/government-
non-profit/six-sigma-government-focusing-customer/)

Incorporating Six Sigma can help any government office to increase its efficiency. One of the
main ways in which Six Sigma can make a notable difference is by identifying common areas
where progress initiatives stall out. It may turn out that a certain office is a culprit in
delaying new actions, including improved processes that will help everyone. Officials can
then work with these offices to develop new ways of gaining approval that will speed up the
implementation of new programs and processes.

Six Sigma can also highlight the true costs that come as a result of these delays. For
instance, a delay in the execution of an increase for automobile registration fees can
ultimately cost a town or county millions of dollars. Not only do these delays present the
image of a disorganized government, but they also hurt the citizens of that community.
Even if people feel as though they dodged a bullet by not having to pay the increased
registration fee this year, they will ultimately have to pay additional money in other areas,

The elimination of red tape allows the government to run more efficiently and
transparently. This makes it easier for the community to have more faith in its government,
which ultimately will benefit all elected officials at re-election time.
817
 SIX SIGMA IN GOVERNMENT

Improved Usage of Funding

One of the most difficult aspects of government is getting things done with a limited

simply print money at will. Financial prudence is important in government, just as it is in any
business or industry.

A branch of government that incorporates Six Sigma will have a better idea of where their
money is going, but the applications of Six Sigma go much deeper. Thanks to the quality
measures taken and the data collected over the course of the project, the government will
have a better idea of how much of a return each investment provides. A government group
that spent $500,000 to renovate an ice skating rink may be pleased to find that residents
have attended skating sessions at that rink twice as often over the past three years.
Discovering the financial potential of improving recreational facilities may inspire those
officials to further enhance other government-owned properties, such as golf courses and
batting cages.

This ability to track spending and return on investment may have tremendous benefits for
politicians. Instead of merely thanking donors for their contributions, elected officials can

was positively impacted as a result of their generosity. Such personalized shows of

appreciation and consideration would be impossible without the data-driven philosophy of
Six Sigma.

Time Allocation
Government officials have a lot to do. In addition to dealing with the red tape that swallows
many of their desired implementations, officials have to take time to meet with their
constituents, maintain a presence in the community and attend important local events,
such as graduations. This is above and beyond the responsibilities typically associated with
government, including passing laws, identifying avenues for process improvements and
lobbying for the necessary support to get their initiatives pushed through the system.

These many activities take a great deal of time. With so much that needs to get done, taking
a step back to evaluate the typical workflow seems impossible. Once again, Six Sigma comes
to the rescue, prompting a re-evaluation of existing process and identifying ways to free up
time without sacrificing productivity.

Just as with financial distribution, politicians will have the chance to see exactly how many
hours each week they devote to the various aspects of their jobs. Furthermore, as they did
with funding, they will be able to see just how productive each of those hours turned out to
be. For example, a politician might find that fundraising for eight hours a week is marginally
818
 SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

less effective than seeking donations for ten hours a week, proving that eight hours per
week is the ideal amount of time to spend fundraising. Those extra two hours can be re-
allocated to a function that needs more attention and that would generate a bigger return.

This approach to time management could be revolutionary for elected officials and any
government employee that has too little time to carry out too many responsibilities. A
responsible distribution of time commitments may result in a more efficient government

Checks and Balances

The United States Constitution was written with an emphasis on checks and balances. In
other words, none of the three branches of government - judicial, executive and legislative -
should be able to obtain too much power. In the event that one of those three branches
becomes more prominent than the others, the remaining two branches have the
wherewithal to use their power to check the growing branch. On paper, this ensures that no
one branch of the government can garner too much influence.
(http://www.usconstitution.net/consttop_cnb.html)

United States. Republican officials are

inclined to vote for Republican measures while voting down Democratic initiatives, and vice
versa. The selling point of staying within party lines is one of solidarity. If a congressman is a
Democrat and he or she wants to get a specific action passed, he or she will receive much
more support if that congressman typically supports his or her party on a regular basis.

Note that these decisions often have little to do with the actual preferences of the
individuals in q
in and veto measures they actually like just to stay in good standing with their political
parties. Once again, the idea is that the entire party will stand behind a politician if he or
she works with that party on their ideas. However, the potential always remains for a

ea.

819
 SIX SIGMA IN GOVERNMENT

Limited Amount of Time

Elected government officials come into office with the confidence of the people and their

Terms of service for elected officials ensure that there are a variety of voices contributing to
solving the problems posed by the people of a given area. They also make sure that no
politician can rest on his or her laurels. They need to generate results, and they need to do
so in enough time that voters can take note of the improvements before the next election.

Six Sigma can be helpful to government officials in so many ways. However, the solutions
-sensitive. Gathering data,
evaluating a broken process and devising ways to improve that process all take time, as do

up to each individual serving to make sure the future incorporates the updated processes
and efficient workflow that have been implemented through Six Sigma.

Getting Buy-In
The unfortunate reality of politics and government is that not everyone is willing to respect
the wishes of other politicians. Every politician has his or her idea of how things should be,
and any elected official is eager to see those ideas implemented as soon as possible.

At the same time, politicians with a great vision for the future might see their ideas
throttled by the people above them. This can happen for a variety of reasons. The
aforementioned party line issue may change the minds of politicians; additionally, personal
philosophies may take center stage, or senior politicians may simply want to take credit for
the ideas of people that are lower on the food chain.

No matter what the situation may be, the fact is that not every good idea comes to fruition,
cessary support. This is
unfortunate, yet unavoidable because of the Catch-22 involved. Six Sigma requires total
buy-in from all stakeholders who may benefit from a project, but most senior politicians
the projected results of the
initiative. This results in improvements that are partially carried out - or not carried out at

820
 SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

all - through no fault of the Six Sigma project team. This can happen even if the team asked
for permission before embarking on the project.

grown steadily over the years - Houston currently boasts more than 2 million residents -

the city found itself facing a $140 million deficit in July 2015.
(http://performance.houstontx.gov/sites/default/files/Industrial%20Engineer%20-
%20Case%20Study%20-%202015-02.pdf)

2010s. However, when the deficit grew to such a high number in 2015, it was time for
Houston to get its most tenured employees involved. Since many city employee had put in
between 20 and 30 years on the job, this change was necessary in order for those
employees to keep their jobs and keep contributing to the welfare of the city.

department was responsible for confirmation and payment associated with renting city
equipment to other groups. Unfortunately, the department suffered from rampant
inefficiency. It typically took Outside Services 111 days to resolve invoices received from
vendors. This was a far cr

demand answers.

The Six Sigma team implemented a series of process improvement measures that helped
them to develop a revised process for handling the outside vendors. These methods
included process mapping, eliminating waste and developing a system for verifying that
Outside Services was actually responsible for a given invoice. The improved methods proved
highly successful. After beginning with an average cycle time of 111 days, Outside Services
brought its average turnaround time down to 2.5 days. Additionally, the group was able to
reduce its staff from six employees to one.

However, this did not mean that those extra employees lost their jobs. In fact, Houston has
re-invested a variety of resources into its Six Sigma projects, including human capital.
821
 SIX SIGMA IN GOVERNMENT

Sigma green bel

The long-time employees of Houston not only get to continue serving the community, but
they are able to do so in new and exciting ways that have invigorated the entire staff of city
employees. The future is bright for Houston, as it is for any city that implements Six Sigma
and continues to invest in its solutions.

Six Sigma is a great resource for both government employees and elected officials. Since any
improvement in government operations has a trickle-down effect that ultimately helps
many people, even the smallest Six Sigma initiative can have large-scale benefits.

Inefficiency in the government hurts everyone - elected officials and government staffers all
feel the strain of broken processes, and nobody faces the impact of these processes more
than the citizens of these affected areas. Bringing up any ideas that can help to improve the
flow of government - ideas, money, time, etc. - stands to benefit operations tremendously.

Government officials may be reluctant to undertake Six Sigma projects out of fear that they

way to use Six Sigma to its fullest potential. Talking to everyone involved with a process -
from the most important people to the lowest-ranking officials - will help bring about a
solution that benefits all parties involved.

Politicians must always remind themselves that the purpose of any Six Sigma project is to
streamline internal operations in order to help the people that truly need assistance. The
focus should never deviate from that objective. Re-elections and raising money must come
a distant second to what really matters.

Source

http://www.villanovau.com/resources/military/six-sigma-for-defense/
http://www.dtic.mil/dtic/tr/fulltext/u2/a493547.pdf
https://www.minitab.com/en-us/Press-Releases/Department-of-Defense-Saves-Billions-
with-Lean-Six-Sigma-and-Minitab/
822
 SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

http://performance.houstontx.gov/sites/default/files/Industrial%20Engineer%20-
%20Case%20Study%20-%202015-02.pdf
https://www.isixsigma.com/industries/government-non-profit/six-sigma-government-
focusing-customer/

823
 Binomial Distribution Control Plan, 235
1 Probability Function, 358 CORREL Formula, 317
1-Way ANOVA, 566 Box Plots, 214 Correlation, 310; Positive vs.
Brainstorming Tool: Affinity Negative, 311
Diagram, 642 Correlation and Regression:
5 Brainstorming Tools: Data Requirements, 331
5 Whys, 86 Brainwriting, 652; Correlation and Regression
5S Approach, 56 Figuring/Role-play, 652; Analysis, 221
Force Field Analysis, 645; Correlation and Regression:
Interrelationship Diagram, Uses, 332
A 644; Nominal Group Correlation Coefficient:
Activity Network Diagram, 641 Technique, 649; SWOT Formulas, 315
Affinity Diagram, 642 Analysis, 650 Cost Benefit Analysis, 227
Alternative Hypothesis, 373 Brainwriting, 652; Card Cost of Poor Quality (CoPQ),
Automated x-bar Control Method, 653; Pool Method, 119; Equation, 120
Charting: Minitab, 417 653 Cost of Quality (CoQ), 121;
Alpha Value, 399 Business Process Equation, 122; Managing,
Analysis of Variance (ANOVA), Reengineering (BPR), 35 123; Related to Sigma
565 Level, 123
Analyze Phase: Fishbone C Critical Path Diagram, 151
Diagram, 206; Graphical Critical to Customer
Analysis, 212; Root Cause Capability, 241 Characteristics (CTC), 114
Analysis, 206; Statistical C-Chart, 424 Critical to Quality
Analysis, 220; Tollgate Coefficient of Determination, Characteristics, 114
Checklist, 223 321 Critical to Quality
Analyze Phase, 158, 163 Confidence Levels, 639 Characteristics (CTQ), 114;
Correlation Coefficient, 314 Identifying, 115
Change Implementation, 231 CTQ Tree, 116
B Check Sheets, 649 Cumulative Density Function
Bar Chart: Minitab, 520 Chi-Squared Goodness-of-Fit CDF: Minitab, 483
Bar Charts, 247; Excel, 247; Test, 293 Customer: Focused
Stacked, 252 Continuous Data, 400 Improvement, 15
Barriers to Support, 19 Continuous Process Customer Experience
Baseline Metric, 198 Improvement, 16 Management Method
Boxplots: Minitab, 513 Control Chart, 238 (CEM), 40
Brainstorming, 86 Control Charts: Minitab, 424 Customer Feedback, 75
Brainstorming Tools, 641 Control Phase, 159, 165, 234; CUSUM Chart (Cumulative
Beta Testing, 9 Control Plan, 235; Tollgate Sum), 454
Checklist, 242
824
 SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

D E Review of Statistical Terms,

395; Steps, 384; Test
Data: Continuous, 186; EWMA Chart (Exponentially Selection, 375
Discrete, 185; Interval, 188; Weighted Moving Average), Hypothesis Testing, 372
Nominal, 187; Ordinal, 188; 454 Hypothesis Testing, 231
Ratio, 189 Excel: CORREL Formula, 317;
Data Analysis ToolPak, 319 Data Analysis tool, 269;
I
Data Driven Processes, 8 Interaction Plots, 629;
Data Sampling: Non-Random, Process Maps, 665 I & MR Chart, 454
197; Sequential, 197; Interrelationship Diagram, 644
Simple Random, 196; F Improve Phase, 224; Tollgate
Stratified, 197 Checklist, 233; Training,
Deviation, 61 Failure Modes and Effect 232; Transition, 232
Dotplot, 511; Minitab, 510 Analysis (FMEA), 182; Improve Phase, 159, 164
Decision Making, 8 Revisiting, 234 I-MR Chart, 443
Decisions, 8 Factorial Experimentation, 598 I-MR Chart, 420, 434
Defect, 43 Fishbone Diagram, 206 In and Out of the Box Method,
Defects, 25; Product, 8; Flow Chart, 657, 667 178
Variance, 60; Variation, 60 Figuring Brainstorming (Role- Input, 99
Defects per Million play), 652 Interaction Plots, 629
Opportunities (DPMO): First Time Yeild (FTY), 81 Interval Plots, 519
Equation, 80 Force Field Analysis, 645 Is/Is Not Matrix, 178
Defects per Unit (DPU), 81
Define Phase, iv, 169; Toolgate G J
Checklist, 179
Define Phase, 157, 160 Gage R & R: Minitab, 543 Jidoka, 25
Define Phase: Tools, 175 Gage R&R, 192; Attribute, 192; JumpStart, 40
Delta Value, 400 Variable, 194 Just-in-Time Manufacturing
Design of Experiment (DOE): Geometric Distribution, 369 (JIT), 58
Response, 637 Graphical Analysis, 243
Design of Experiments, 223 L
Design of Experiments (DOE), H
593; Confidence Levels, 639 Lean Concepts, 43
Design Phase, 159, 166 Histograms: Excel, 285 Lean Process Management, 33
Designed Experiment: Steps, Histogram: Minitab, 502 Lean Six Sigma, 25; 5 S
600 Histograms, 277 Approach, 56; Just-in-Time
Discrete Data, 401 Hypothesis Testing, 221, 300, Manufacturing, 58
DMADV, 21, 156 383; 1-Proportion Test, Likert Scale, 80
DMAIC, 21, 156 375; 1-Sample T-Test, 377; Linear Regression Analysis,
DMAIC vs. DMADV, 168 2-Proportion Test, 376; 2- 321
DMAIC vs. DMADV, 157 Sample T-Test, 379; Chi-
DPMO, 10, 23 Square Test, 378;
Likelihood of Error, 374;
Mann-Whitney Test, 382;
One Sample Wilcox, 381;

825
 M Nine Boxes Model, 37 Pareto Principle, 68
Null Hypothesis, 373 Payback Analysis, 228
Measure Phase, 161; Nominal Group Technique, P-Chart, 421
Collecting Data, 185; 648 PDCA Cycle, 23
Tollgate Checklist, 204 Non-Normal Continous Pie Charts, 255; Excel, 257;
Measure Phase, 158 Distributions: Beta Minitab, 520, 527
Measurement Phase: Distribution, 348; Cauchy Poisson Equation, 365
Accuracy, 189; Gage R&R, Distribution, 344; Gamma Pilot, 230
192; Resolution, 190; Distributions, 350; Laplace Probability Distributions, 483
Stability, 192 Distribution, 346; Logistic Probability, 275; Choosing the
Measurement Phase: Choosing Distribution, 345; Right Distribution, 491;
a System, 189 Triangular Distribution, Formula, 275
Measures of Success, 155 352; Uniform Distribution, Probability Density Function
Meaurement Phase: Collecting 347; Weibull Distributions, PDF: Minitab, 483
Data, 196 341 Problem Function, 84
Minitab, 469; 1-Way-ANOVA, Non-normal Continuous Problem Statement, 88
570; Calculate Multiple Distribution: Lognormal Process Capability, 242
Probabilities at Once, 489; Distribution, 339 Process Improvement, 33
Column or Row Statistics, Non-normal Continuous Process Map: Symbols, 657
479; Descriptive Statistics, Distributions: Exponential Process Maps, 656
550; Design of Experiments, Distribution, 336 Process Owner, 101
605; Graph Menu, 496; Non-Normal Continuous Process(es), 95; Components,
Graphs and Quality Tools, Distributions, 336 103; Data, 102; Events, 100;
495; Histogram, 503; Non-normal Discrete Input, 99;
Hypothesis Test Assistance, Distibutions: Binomial Interdependencies, 96;
589; Hypothesis Test Steps, Distribution, 352; Poisson Major Components, 97;
562; Hypothesis Tests, 554; Distribution, 360 Output, 99; Process
Interaction Plots, 629; Non-normal Probability Owners, 101; Resources,
Interface, 470; Menu, 476; Distributions, 335 96; Tasks, 100; Time, 95
Random Data Tool, 477; Normal Curve, 292, 336 Project Charter, 169
Regression Analysis, 553; Normal Distribution, 291 Project Viability Model, 131
Scatter Plot, 497; Stat Normal Probabilibty
Menu, 533 Distributions, 274 Q
Muda: Types of,, 43 Normal Probabilities, 307
Muda: Conveyance, 50; NP Chart, 422 Quality, 113; Appraisal Costs,
Correction/Rework, 46; 122; Cost of Poor Quality
Definition, 43; Inventory, (CoPQ), 119; Cost of Quality
O
47; Motion, 48; other (CoQ), 119, 121; Costs of
forms,, 54; Over- Output, 99 Prevention, 121; Critical to
processing, 51; Customer Characteristics
Overproduction, 44; Type I, P (CTC), 114; Critical to
55; Type II, 56; Waiting, 52 Quality Characteristics
Pareto Chart, 70; in Excel, 71; (CTQ), 114; Critical to
N Minitab, 538 Success, 125; CTQ Tree,
Pareto Chart, 212; Uses, 74 116; Definition, 113
Net Present Value (NPV), 229 Pareto Principle, 68
826
 SIX SIGMA: A COMPLETE STEP-BY-STEP GUIDE

Quality Improvement Distribution, 778; Phase, 234; Coorelation

Programs: Types of,, 33 eCommerce, 762; and Regression, 331;
Engineering, 723; Field Measure Phase, 182;
R Services, 729; Finance, 700; Measurements of Success,
Food Service, 794; 173; Selection, 126; Team
R Chart Control Chart, 418 Healthcare, 692; History, Member Roles, 172
R value, 314 21; History of,, 27; Six Sigma Project: Charter, 169
Random Sampling, 415 Hospitality, 706; Human Six Sigma Sponsors, 145
Root Cause Analysis, 206 Resources, 711; in the Six Sigma Team: Black Belt,
Root Cause Verification Warehouse, 770; 146; Building a, 142; Check
Matrix, 209 Information Technology, Sheets, 649; Members, 144;
Regression, 310 717; Manufacturing, 786; Process Owners, 145;
Regression Analysis, 325 Metrics, 80; Principles, 15; Responsibility Chart, 646;
Rolled Throughput Yield (RTY), Retail Industry, 755; Sales Six Sigma Leaders, 146;
82 and Marketing, 735; Team Sponsors and Champions,
Rummler-Brache, 37 Management, 142; When 145; Team Member Types,
Run Chart: Minitab, 533 to use,, 41 149; Timekeeper, 147
Run Charts, 198; Excel, 200 Six Sigma Certification: Black Six SIgma Team: Project
Belt, 30 Manager, 147
S Six Sigma Certification, 29 Six Sigma Tools, 8, 641
Six Sigma Certification: Green Solutions Selection Matrix, 224
S chart Control Chart, 419 Belt, 30; Master Black Belt, Scatter Plot: Minitab, 497
Sample Size: Calculations for, 31; White Belt, 29; Yellow Statistical Analysis, 455
398, 401; Importance of, Belt, 29 Swimlane Diagram, 668
396 Six Sigma Champions, 145 Swimlane Process Map, 657
Sample Size Tests: Minitab, Six Sigma Construction, 801 Stakeholder Analysis, 175
403 Six Sigma Project: Problem Standard Deviation, 61
Scatter Diagrams: Excel, 260; Statement, 88 Standard Devation: Equation,
XY, 259 Six Sigma Project: Phase-based 62; in Excel, 66
Scope, 92 Timeline, 150 Standard Devation: Reason for
Scope Creep, 92 Six Sigma Project: Analyze Calculating, 67
Scrum, 39 Phase, 206; Budget, 154; Starburst Brainstorming, 650
Sigma Level, 241; Calculation Documentation, 232, 638; Statistical Process Control, 22,
of,, 12; Comparison of,, 11 Enterprise Level Selection, 239; Control Charts, 23
Sigma Levels, 25 127; Improve Phase, 224; Statistics: 2k vs. Multi-level
Sigma Level, 13 Milestone Meetings, 153; Factorials, 635;
SIPOC Diagram, 103; Milestones, 172; Assumption, 628; Central
Brainstorming, 106 Objective/Goal, 91; Pilot, Limit Theorem, 344; Chi-
SIPOC Diagram: Creating a, 230; Process Level Squared Goodness-of-Fit
103 Selection, 139; Scheduling, Test, 298; Control Charts,
Six Sigma: Application of,, 28; 149; Scope, 171; Scope and 239; Coorelation and
Basic Concepts, 60; Call Scope Creep, 92; Regression, 310;
Center, 749; Challenges, 18; Stakeholders, 171 Descriptive, 274;
Customer Service, 742; Six SIgma Project: Celebration Descriptive vs. Inferential,
Definition, 8, 10; and Reflection, 242; Control 274; Graphical Analysis
827
 Tools, 243; Hypothesis Total Quality Management Voice of the Customer (VOC),
Testing, 372; Inferential, (TQM), 34 75
274; Interactions, 624; Type I Error, 397 Variation: Common Cause vs.
Main Effect, 624; Non- Type II Error, 398 Special Cause, 453
normal Discrete
Distibutions, 352; Non- U W
normal Probability
Distributions, 335; Normal U Chart, 423 W. Edwards Deming, 23
Curve, 22; Normal Walter Shewhart, 22
Distribution, 292; Power V Waste, 17
and Sample Size, 405;
Presenting Data, 457; Value Stream: Definition, 15 X
Probability, 275; Process Value Stream Mapping: Seven
Capability, 242; Sample Flows, 676; Symbols, 677 X Y Scatter Diagrams, 259
Size, 396; Sigma Level, 241; Value Stream Mapping, 676 X-bar Control Chart, 262
Standard Deviation, 61 Variation: Definition, 17 X-bar Control Chart, 418
SWOT Analysis, 650 Verify Phase, 159
Verify Phase, 167
Visual Controls, 238
T
Voice of the Customer (VOC):
The Central Limit Theorem, Campaigns, 75; Tools, 77
344

828