What gets measured gets done

Unveiling the secrets of performance management

Nopadol Rompho
 Preface
Simon Sinek makes a particularly interesting point in his book Start With
Why: How Great Leaders Inspire Everyone to Take Action. The author lays
down the three important questions: what, how, and why. Most people only
communicate the “what” and “how,” failing to address the “why.” I could
have written in this preface: “This book is about performance management
(the what); and will present how it is designed and put to good use in
organizations (the how).” Then, this would be an ordinary book about
performance management techniques. However, I’d like to communicate the
“why” first. I am inspired to transform the usual jargon-filled “academic
textbook” into an engaging and comprehensible text. I studied many excellent
research reports that few people manage to get through, and saw many useful
textbooks gather dust on library shelves. My aim is to transform such content
into simple terms that everyone can comprehend, and more importantly,
enjoy.
Through my experience of teaching, writing textbooks, and researching about
performance management made, I realized that people tend to overlook
certain aspects of performance management. Apart from its being engaging
and comprehensible, I hope this book will be a successful knowledge transfer
of such overlooked aspects of performance management.
 Acknowledgments
This book could not have been possible without the assistance of many
others. The first person I would like to thank is my wife, because she inspires
me to write. She is also one of the first fans of my Facebook fanpage, and is
usually the first person to tell me that my morning post was difficult to
understand. In addition to her reviews, she also helped consolidate and edit
various articles, gather many files together, and create and format layouts. In
other words, she has been my editor.
Next, I would like to thank my daughter Paan and my son Pun for the moral
support, their smiles and laughter...
I am grateful to my parents for their knowledge and support. I also appreciate
my younger brother’s continuously exchanging thoughts with me.
I thank my research and teaching assistants for gathering the relevant data
and works. I am also grateful to Pat Juthamard, who helped me translate the
Thai version of this book into English.
In addition, I would like to thank the fans of the Performance Measurement
Fanpage for their comments, shares, and likes of my daily 7 am posts. These
have encouraged me to write this book.
Last but not least, I would like to thank you, the readers.
 Table of Contents
Chapter 1 Why measure?
Chapter 2 Benefits of performance measurement
Chapter 3 On measurement
Chapter 4 Cautions of performance measurement
Chapter 5 Get to know the system of performance measurement
Chapter 6 Things you should know before creating a performance
measurement system
Chapter 7 Design a performance measurement system
Chapter 8 Create a performance measure
Chapter 9 Set targets and collect data
Chapter 10 Prepare a performance report
Chapter 11 Do not use it if it fails a test
Chapter 12 Implement a performance measurement system
Chapter 13 Put the data from performance measure to good use
Chapter 14 Make your performance management system sustainable

Why this book?

I really love reading... Not only does it promote imagination and is a
wonderful kind of entertainment, but it also gives knowledge. A book can
change our lives. Looking around, I saw plenty of books lying on the shelves,
commonly referred to as “academic” books. I realized that these books
contain a lot of knowledge and benefits. Yet, people are not interested in
them. Why?
What is the difference between these books, and the ones on the “Best Seller”
shelf? The answer is “simplicity” and “engagement.” I have written many of
those “textbooks,” and I made sure to follow the rules and norms: they should
be supported by research, refer to academic theories, have learning
objectives, and so on.
As a university professor, I followed the format of what universities or public
organizations would define as a “textbook”, partly to obtain academic
positions, as they proved my in-depth academic knowledge. However, these
textbooks lacked simplicity and engagement, even though they might have
been useful. Completing a textbook is quite a difficult task, as it requires
research and crafted writing to express the author’s intentions. It is a pity that
very few people are actually interested in these textbooks. The readership of
these books is usually the students who are studying a particular subject. If
that subject is mandatory, then more students will study it. However, if the
subject is an elective, less number of students will study it. Most importantly,
the student’s main objective in reading these textbooks is to memorize the
content for exams. On the day after the exam, as one of the textbook writers,
I am stabbed in the heart... Those textbooks will be coldheartedly abandoned.
Well, if the exam is over, what’s the point of keeping them?
Another potential reader group may be working people who seek knowledge
about these subjects, but when they walk into a bookstore and read a few
pages, they will immediately put the book down. There are many difficult
equations and theories crawling all over the pages. They do not understand
what is written on those pages. Or if they do, they find it uninteresting, and
even after they finish reading, they do not know how to apply the knowledge
in real life.
This is what inspires this book. I had once written an academic book called
Organizational Performance Management, which is used as teaching
material for Bachelor’s, Master’s, and Ph.D. courses. That book was
supported by my research. However, I want this textbook to be beneficial to
other audiences than just my students. If it comes in the form of a traditional
textbook, the general public would hardly pick it up.
With this in mind, I began to write daily articles on the Performance
Measurement Facebook Page at
www.facebook.come/PerformanceMeasurement. There, I try to communicate
the knowledge from my textbooks in simple terms, and through interesting
real-life examples. I started with a few hundred followers, most of whom
were my students. Later on, the number was in the thousands, and shortly
after in the ten thousands. As I am writing this chapter, my page has already
gathered 60,000 Likes. This proves that something simple and interesting can
attract followers. Now, if you are ready, let us begin!
 Chapter 1:
Why measure?
Why learn about performance measurement? The first part of this book
highlights the main reasons why we need performance management in the
first place. Let’s see what those are.
#1 Performance measurement helps realize the outcome.
Yes, as simple as that. I think this is one of the main reasons why you should
measure. If you do not measure, then you will not be informed, right?
Imagine yourself driving... You would check the dashboard to see how fast
you are driving, how much gas you have left, and the heat level of your
engine. What for? So that you can make a decision. If you are driving too
fast, then you will slow down. If you are running out of gas, then you will
drive to the nearest gas station. Same goes for business. If you do not
measure, you can never know the outcomes of your actions. If you do not
know whether or not your customers like your product or service, or how
much you can sell, you are literally waiting for the day that you will run out
of business.
#2 Performance measurement can be a warning system.
Let’s revive the previous example. In addition to providing information on
the current status of your car, a car’s dashboard also serves as “a warning
sign.” Normally, I am not an observant person (please do not be like me). I
always start the car and drive off without checking the gas. Most of the time,
I will suddenly notice the yellow light of the fuel level gauge. This is what I
mean by a “warning sign.” When the yellow light appears, the driver knows
that he has to be careful, because there is not much gas left. The driver would
then look for a gas station to fill up the gas tank, because if he keeps on
driving, the car would definitely stop moving at one point.
Organizations need measurements like these, because if they just wait to see
the final result, such as profit or loss, it might be too late – akin to waiting for
the car to stop moving first, and then searching for the reason why the car
stopped working. Performance indicators, such as repeat purchases, can serve
as “warning signs” for an organization, so they can understand if their
strategy is on the right track or not. If customers return, then the strategy
should be right. Whereas, a decrease in the repeat purchase rate could be a
bad sign, signaling to the organization that its strategy needs to be reviewed
and revised before any real damage is done. It can be too late to fix things
sometimes.
#3 Performance measurement changes behavior.
The first question is “why change behavior?” For an organization’s strategy
to be successful, employee behavior should be in alignment with that
strategy. Let me give you an example. Assume that the strategy of a company
is “to make a difference” The company could never make a difference if its
employees think traditionally and cannot innovate. So, what can we do for
the employees to be innovative? I see many organizations try to communicate
to their employees that it is necessary to think outside the box. They arrange
meetings to clarify the matter, make announcements through the company’s
Intranet, or even through a notice in front of elevators. Is it an efficient way
to apply the strategy? I think it is, because employees would gradually absorb
the idea. However, if you ask me how long it would take to see the results, I
would bet very long. Have you ever read something in front of an elevator
and felt like “I need to change the way I work right now”? I think that hardly
ever happens.
So, what should we do to make these employees aware of the importance of
this matter? The simple answer is to measure everything that is important. If
you want your employees to see the importance of innovation, then try to
create performance measure metrics, such as “the number of beneficial
suggestions”, and tie them to incentives. Of course, every employee would
want the incentives, and they would realize that in order to receive the
rewards, they will need to come up with new ideas and make suggestions to
the company. Yes, at first, these behaviors might happen because the
employees are being measured and want an incentive, but eventually they
become part of a routine, and develop into “the culture” of the company.
#4 Performance measurement translates strategy into action.
During a lecture, I asked the attendees, “Whose organization has a strategy?”
Most raised their hands. However, when I asked the next question, “Whose
organization can translate everything in their strategy into action?” Raised
hands disappeared all of a sudden. This incident does not happen only in a
particular country, it happens everywhere. All organizations have strategies,
but, most of the time, those strategies are not translated into action.
Performance measurement can help, as “you get what you measure.” If you
want to turn strategy into action, then you have to measure it.
I had asked my students once if they believed that performance measurement
could really help translate strategy into action. Most of my undergraduate
students had no work experience, so they could not imagine. I asked them if
they would be OK with not having to take an exam for my course. Most of
them said they would be OK with it. Then, I would tell them that they have to
promise me that they will go back and review the material as if they were
going to take an exam. Everyone laughed, because they knew by heart that
they would not do it. An exam is a kind of performance measurement. If the
students do not have to take an exam, then they would not have the
motivation to review the lessons. Hence, performance measurement is one of
the tools that can translate strategy into action. In addition, it will also give
you the results of your action. Therefore, performance measurement is one of
the tools that encourages action.
#5 Performance measurement helps monitor trends.
Jeff and James both scored 60. Who do you think had more problems in
studying?.. Well, they got the same score, so they should have equivalent
problems. Let’s think about that again. For the past four weeks, Jeff scored
30, 40, 50, and 60, while James scored 90, 80, 70, and 60. Now, who do you
think has more problems? It seems to be James, right? His performance has
declined over the period, whereas Jeff’s performance has continued to
improve. See? Receiving the same score did not mean they both had
problems.
Apart from reflecting the current status, performance measurement can also
be the tool that informs you about trends. We can use a past trend to predict
the future so that we can develop a prevention plan. So, is your organization
Jeff or James?
#6 Performance measurement helps allocate resources.
There is a tank full of water with many leaking holes. What should we do?
Well, find something to plug those holes with. Yes, that’s right. If we can
plug all of the holes, water will stop flowing out, and the problem will be
solved. But... Life is not that simple. Let’s assume that you have a very
limited amount of clay that can be used to plug those holes. You cannot plug
every hole. So, what should you do? Think about it... If we divide the clay
into tinier pieces and try to plug every hole, in the end, it might turn out that
we could not stop any leaks, because the pieces of clay were too small for the
holes. Plus, we would lose all the clay, and the water would continue to leak
from every hole. You can solve the problem in another way by using the
limited amount of clay to plug only the few large holes so that you can stop
80-90% of the leaks, and plug the rest of the holes once you have more clay.
Even if you do not have the additional clay right now, at least a smaller
amount of water will leak. This is a better solution, right?
Same goes for organizations. Every organization has problems, and problems
never care if we have enough resources to solve them. And most of the time,
problems occur when we have few resources to spare. Using the limited
amount of resources to fix all the problems that occur at the same time might
not be the best option. Apart from the fact that it might not fix any problem at
all, it might even cost us all of the limited resources.
Wouldn’t it be better to allocate the limited resources to resolve the most
important problem first? Once that problem has been fixed, we might find
that we were able to prevent 80-90% of the damage. Once we have more
resources, then we could start fixing the rest of the problems. My point is that
performance measurement will tell you which problem is the most important
and needs your attention first. The less significant problems can be fixed
later. Simply put, performance measurement will help you prioritize your
problems.
#7 Performance measurement can teach.
Have you ever noticed that we measure different aspects of our projects
mostly before we start them, rather than after we finish? Prior to starting a
project, we measure everything – profit rate, return rate, breakeven point –
even though the numbers obtained from projections have very high
uncertainty levels (since most of them are just predictions.) Before we begin
a project, we should make predictions to make sure that our project will be
worth it, because once a project begins, a lot of money is already at stake,
which makes it hard to turn back.
A good project analysis could lead to future success. Have you ever measured
the results of your project after you were done, to see whether everything
progressed in line with your predictions or not? It is true that you cannot turn
back the time, but if you measure your finished project’s performance, you
can learn from your mistakes, and never make the same mistake twice.
George Bernard Shaw, one of the founders of the leading academic
institution London School of Economics, had said: “Success does not consist
in never making mistakes but in never making the same one a second time.”
#8 Performance measurement can be a marketing tool.
Do you believe that organizational performance measurement can be used as
a marketing tool? What?! We have always been taught that marketing is just
the 4Ps: Product, Price, Place, and Promotion. Isn’t that right? How can key
performance indicators (KPI) used in measuring organizational performance
become a marketing tool?
Here’s the thing. Normally, in an organization, we do not use only one KPI,
because we have to measure many important factors. Some KPI outcomes
might turn out good, and some bad. What turns out bad are the weaknesses of
that particular organization. The benefit of knowing the results is that we can
work on those weaknesses to resolve or eliminate the problems. On the other
hand, aspects with excellent KPI results would reflect our strengths, right?
This is where we can utilize KPI as a marketing tool. We usually see
advertisements like “100% satisfaction,” or “Best Place to Shop,” etc. These
things that we present to customers might be the results of organizational
performance measurement. So, go back and take a look at your organization’s
KPI to see if there is anything in there you should tell the world about.
#9 Performance measurement is input for incentive systems.
I believe everyone waits for a “bonus” or “reward.” We sit on the edge of our
seats hoping to get it at least once a year, right? Giving rewards or bonuses to
employees is a great motivational factor. Yet, it can also discourage
employees, if done incorrectly or unfairly. What is fair? This question is very
vital. “Equally shared” does not always mean “fair.” Think about what would
happen if you give everyone an equal amount of bonus?
The employees who work hard would be discouraged, because those who
don’t work as hard still get the same amount of bonus. And those who do not
work hard would think that what they are doing is good, because even when
they do not work, they still get the bonus. Why work at all then? In the end,
the organization would only have people who do not want to work. I do not
need to tell you what is going to happen to this organization next, right?
Organizational performance measurement is one of the tools that we can use
to help make decisions about employee incentives. Of course, we might not
be able to use performance measurement to help decide on employee
incentives 100%, because each job is different. However, at least the
performance that is reflected by KPIs can help us make “better” decisions
than those based solely on our judgment. Giving incentives can be hard, and
has always been a difficult management task. However, used correctly, KPIs
can become motivational tools that lead to continuous improvement.
#10 Performance measurement can be a benchmarking tool.
Let’s imagine you are a tennis athlete… Who would you want to train with:
the world’s number one tennis player, or your high school friend? If you play
soccer, which team would you want to play against: the world champion, or a
primary school team? The answer for the questions above depend on your
objective in competing. If you compete for a chance to win, then you would
play tennis with your high school friend, or play soccer with the primary
school team. However, if you play to improve, I think most people would
choose to play tennis with the world’s number one player, and play soccer
with the world champion.
Same goes for business. If you want to be better, you need to look up to the
ones who are better than you in the aspects that you want to improve.
Learning from the better ones is best practice. Instead of having trials and
errors all by yourself, you can learn from those who have been successful.
This is what we call “benchmarking.” No one can be better than you in every
aspect; sometimes you are better, and sometimes they are better. Wouldn’t it
be nice if you can exchange knowledge so that both parties can be successful
together. Yet, benchmarking would mean nothing if you do not know what
you are good at.
And this is another benefit of organizational performance measurement; it
reveals what you are good at, and more importantly, what you are not good
at. Performance measurement reflects your weaknesses, which can be used
for benchmarking that would eventually lead to the betterment of the
organization.
#11 Performance measurement increases motivation.
Who likes to play online games? Whether it is Facebook games, or any other
popular online game, there is one thing that makes games addictive. Do you
know what that thing is? It is the competition displayed on the leaderboard!
The tougher the competition, the more fun it is. What follows is that we will
try to win by coming up with different strategies; we will use our free time to
play the game, and feel like time has flown by. But, have you ever felt like
that when you work? (I doubt I need to wait for an answer.)
Let’s take a look at this again. If we work as if we are playing a game, how
interesting would our jobs be? When executed well, performance
measurement will inspire us to work and develop to achieve set targets.
Performance measurement can turn a boring work day into the day we are
waiting for, and our Monday mornings might never be the same again.
 Chapter 2
Benefits of performance measurement
In the previous chapter, we discussed why we need performance
measurement. In this chapter, before we get to learn what performance
measurement is, or how it can be done, I want everyone to get a clearer
picture of the benefits of performance measurement. Understand the goal
before finding a solution... So, if you are ready, let’s see what kinds of
benefits performance measurement has to offer.
#1 Help monitoring activities
What quality should management possess? A good manager does not need to
do everything by himself, without any mistakes. A good manager might not
be the person who goes home last, in order to finish every task at hand.
Rather, a good manager is someone who can create a team environment
where members work is in alignment with the organization’s mission.
How can you make your subordinates work the way you want them to? Many
people utilize various methods in controlling the way their subordinates
work. A control system that requires physical control usually comes at a high
cost. For example, if we want our salespersons to actually sell products,
instead of traveling around for fun, we might need to measure the distance
these people have commuted, and ask our customers, or even ask another
employees to follow them around. But this kind of control method requires a
lot of investment. Also, employees might get the impression that you do not
trust them.
So, what can we do? How can we make sure that our employees do not get
sidetracked, or do something that our organization would not approve of?
Performance measurement is the key to help solve this problem. If you want
to monitor the activities of salespersons, simply measure their sales volume
and accordingly provide incentives. As simple as that! Now you can manage
the way your salespersons work, and be rest assured that your employees will
comply with your organization’s goals. This method does not require a lot of
investment, and would not make employees feel like management does not
have enough confidence in their abilities.
Have you ever wondered why you have to take an exam after you finish a
course? I used to wonder that, too, and I believe many people are curious
about the same. Yes, exams can demonstrate whether we have acquired the
required knowledge of standards or not. Yes, exams can demonstrate who
knows more than who. Yes, exams can demonstrate your weaknesses so that
you can address them for betterment. However, I think there’s even more to
exams than that. Examination is a kind of performance measurement. Let’s
think about it... If there were no exams, how many people would attend
classes? How many people would pay attention in class? How many people
would review their lessons? What would have been missing are these
activities. More importantly, these activities are all beneficial. Examination is
a control tool, which, if applied well, yields great benefits. Actually, the
result of an exam itself is less beneficial than the activities that students
perform in preparation for the exams, such as reading books and reviewing
lessons.
#2 Determining your scope of work
Have you ever wondered what your scope of work actually is, and where you
can find the relevant details? Many people would say, “Well, just look at the
job description.” A job description is a good information source on your
scope of work and responsibilities. However, the problem is that job
descriptions provided by some companies include every task that could
possibly exist - pretty much everything under the sun. So, although given the
job description, your responsibilities and scope of work might still be unclear.
Or sometimes, there are cases when you will only see five tasks in the job
description, and think it is not that difficult. However, in reality, you will not
be executing any of the tasks 1 to 4, but you will always be working on the
5th one: “other assigned tasks,” which simply means that you do whatever
your boss asks you to do!
I think that many of you have experienced the above problem before. Having
KPI as one of your tools can help solve the problem one way or another,
because performance indicators tell us about the main task that we need to
achieve. If we tell our salespersons that we are going to measure the “sales
volume,” it practically implies that we want our salespersons to know that
they need to “sell products.” If we tell our production manager that we are
going to measure “defect rate” and “on time delivery,” it is like we are
telling our production manager that his or her job is to “produce good quality
products and deliver them on time.” I think it is possible that sometimes the
KPIs that we use, can perform better than unclear job descriptions.
#3 Translating strategy into action
Everyone must have heard of the word “strategy,” but few really know what
it is. Strategy is what an organization develops to achieve its goal. The basics
of strategic management involve planning, implementation, and evaluation.
The problem with most organizations is not the lack of strategy, but rather the
implementation process.
Many organizations have annual strategy reviews, and have management
teams attend a seminar to brainstorm about the best strategy. The problem
happens afterwards – when the formulated strategy is not put into action, and
rather laid on the shelf. Why isn’t strategy always translated into action? I
think this problem has many causes. One is that most employees do not know
what a strategy is. Even when organizations publicize their strategy through
various channels, it does not reach every employee. More importantly,
employees might be aware of the strategy, but still not realize its importance.
Performance measurement is one of tools that can solve this problem. As
strategy is measured, we will need to know what it really means. The terms
like ‘effectiveness’, ‘efficiency’ or ‘sustainabiltiy’ will be translated into
more operational terms hence managers will understand more clearly and
thus strategy can finally be put into action.
#4 Understanding the process
In addition to providing us with the results that can be used to improve the
way we work, performance measurement has another benefit that you might
not be aware of. Can you guess what it is? When designing indicators or
analyzing results, another thing to consider is the business process, which
turns strategy into action.
For example, an academic institution has the strategy to produce graduates of
distinction for society. Hence, the institution needs a “teaching process” to
achieve its goal. We create performance measurement systems to control the
processes that will enable us to achieve our goals. When designing any
performance measurement system, the question will arise: what is “this
process” for?
This question is particularly important, and could lead to process elimination.
Some processes might have been necessary in the past, but no longer needed
in the present, such as a government policy. However, we still follow that
same process because we had done it last year. Performance measurement
will make us question, “What is this process for?”, which would then lead to
future improvement. This is another unexpected benefit of organizational
performance measurement.
#5 Knowing your limits
How fast can you run 800 meters? Most people cannot answer this question.
The reason is that you have never measured it before, right? Performance
measurement makes us see our own limits; we can learn how fast we can run
800 meters, how much weight we can lift, and how far we can jump. Once
we know, we can improve. That is why athletes always measure their
performance when they are training or competing.
Same goes for organizations. If we do not measure, we would never know
how much we can accomplish. If we never measure, we will never improve.
Hence, another benefit of organizational performance measurement is that it
enables us to become aware of the limit of each process. For example, we can
find out how many customers we can serve per day, how many items we can
deliver to customers per month, or how fast we can respond to customer
complaints. Once you realize that, you will be able to improve your process.
Do you know the limits of your own work processes?
#6 Help in betterment
Did you know that, apart from exercising and reducing calory intake, there is
another technique to reduce weight? Weigh yourself every day. It sounds
strange – how can weighing yourself make you lose weight? Actually,
weighing yourself does not help you lose weight directly, but it does help
indirectly, because it will affect you emotionally. Weighing lets us know if
we have gained or lost weight. If we weigh more than before, the
measurement reminds us to take control of our weight, one way or another. If
we weigh less than before, then measurement encourages us to lose weight.
Many of you might have noticed that once we stop weighing ourselves for a
certain amount of time, then measure again, our weight tends to increase.
That is because there was no performance measurement to control us.
Same goes for organizations. Sometimes organizational performance
measurement does not have a direct effect on quality or the productivity of
employees. However, it imposes an indirect effect by acting as a warning
system, or as encouragement for the employees. If the result of performance
measurement turns out to be negative, we will be able to get our hands on the
problem quickly and solve it. If the result turns out to be positive, then we
would be encouraged to move forward. So, we can say that if organizational
performance measurement is done well, it can be a tool that encourages
improvement and productivity.
#7 Help in prioritizing problems
Have you ever been in a situation where many problems occur at the same
time? How did you deal with those problems? Usually there are four kinds of
management in terms of the methods with which they deal with problems.
Let’s see what kinds of management those are.
The first kind is the kind that just sits around, hoping the problems will just
go away. This approach might be effective for certain problems. But mostly,
it does not work at all. And sometimes, small problems turn into bigger
problems, much harder to fix.
The second kind arranges a meeting to solve the problem. Well, it does sound
good at first, but the thing is, meetings only convey that we have plenty of
problems, and we want everyone to participate in solving them. After we
leave the meeting room, no one knows what he or she needs to do. We only
point fingers at each other, saying, “It is your responsibility, not mine.” This
is the kind of organization that deserves a “You Do It” award.
The third kind tries to solve all problems at once. Sounds like the right way to
do it. However, there is one problem: limited resources. If we delegate the
limited resources to fix all problems at once, in the end, none of the problems
might be solved. In addition, we would lose all the resources.
The fourth kind is the one that solves the most recent problem. Do not forget
that the most recent problem is not necessarily the most important one that
deserves immediate attention. We might even risk losing resources in fixing
unimportant problems.
These four kinds of management occur in organizations with poor
performance measurement systems, or in those that do not put performance
measurement to good use. If there is a complete performance measurement
system and the outcomes of the system are utilized, apart from knowing
which process has a problem, we would also know which problem should be
prioritized. Organizational performance measurement can play a part in
prioritizing problems, and help management with resource delegation to
correctly solve problems.
#8 Forecasting
Do you want to know the future? Of course, we cannot expect 100%
accuracy in forecasting. However, we can estimate. The most popular kind of
estimation is to use historical data. We forecast future stock prices based on
historical stock prices; we forecast sports scores based on past scores,
because we believe the past is a good indicator of the future. Hence,
acquiring historical data is very important.
Organizational performance measurement is a tool that provides us with
historical data. When we have enough data, we can then utilize that data to
forecast the future. Furthermore, organizational performance measurement
helps us realize the limit of each working process, which will help us predict
the future more accurately. For example, if we know we can serve no more
than 100 customers per day, then we would know that our revenues will
come from no more than 100 customers per day. So we can say that
organizational performance measurement can be used as a tool to forecast
operational results, too. Do you want to know your organization’s operation
results for tomorrow?
#9 Giving you the courage to assign work
Shortly after I graduated, I used to meet up with my college friends. After
work, we would frequently have dinner together. After a while, we still had
our usual meet-ups, and from time to time, some of my friends were not be
available. Generally, everyone usually participated. But one day, one friend
of mine disappeared. Every time I called, he would say, “I’m very busy.” I
heard that this friend had recently been promoted as a “manager.” He used to
be able to go home at 5 pm every day, but now he goes home at 1 am or 2 am
every day. One day, my friends got curious, so we dragged that friend out to
dine with us. We asked him as to why he has to go home that late. Doesn’t he
have any subordinates? The answer was that he does have subordinates, but
he does not feel like leaving his work to his subordinates. He was “afraid that
his subordinates wouldn’t be able to deliver good work” like he does.
This point is very interesting. I believe that many managers are just like this
friend of mine. They are not willing to delegate work to other people. And
even when they allow other people to do their work, in the end, they end up
doing it themselves. Thus, the manager no longer “manages,” instead, they
“work” for their subordinates. Operational performance measurement is the
tool to help managers solve this problem.
The first reason is that performance measurement is an effective control
system. When it is available, workers are willing to work their hardest, so the
bad output that managers are worried about is unlikely to be produced.
Secondly, when a problem occurs, performance measurement will reflect that
problem immediately. Thus, managers will be able to recognize the problem
right away, and find a way to fix it. Good organizational performance
measurement is like a reliable right-hand man who controls work operations
for us so that we can have enough time to “manage.”
#10 Helping us admit and take responsibility only for the mistakes we
are associated with
There is another story I would like to tell you. A CEO of a company received
a complaint from a very important customer that his delivery did not arrive
on time. Therefore, the CEO called the production manager to complain, and
insisted that he needs to speed up production to deliver items to customers on
time. One month later, the CEO received complaints again, from the same
customer, on the same matter. The CEO was upset, and called the production
manager immediately. The manager explained that he is working at full
speed. Before the matter got any worse, the manager came up with an idea for
the company to create an operational performance measurement. It would
measure the time from when a customer order is received until the products
are delivered to the customer, process by process – receiving an order,
producing, and delivering. One week later, the results showed that the
process that took the most time was not production, but delivery, which faced
many delays. The one who deserved to be reprimanded and needed
improvement was the delivery department, not production.
From this story, you can see that organizational performance measurement
helped the production manager keep his job, and also helped identify the real
cause of the problem. This is another benefit of organizational performance
measurement.
#11 Changing organizational culture
What is organizational culture? It is what people in an organization value and
follow, and also what an organization gives importance to and uses as rules
of practice – similar to human behaviors. Nowadays, organizational culture is
created because it is necessary for the long-term survival of the organization.
Organizations with strong cultures have clear operational direction.
Now, let’s test that. If I mention the word “innovation,” which organizations
come to your mind? Is it something along the lines of Apple or Google? That
is because these organizations reflect “innovation” through their new
products. In order to establish a strong organizational culture, an organization
needs to communicate what employees should give importance to. This is
where organizational performance measurement can help.
Think about it. If Apple and Google, which are companies that emphasize
innovation, have KPIs like the percentage of decrease in expenses and expect
it to decrease 5% every year, we would not be able to witness as many new
products from these two companies, because R&D would have budget
constraints. Creating the wrong performance measurement system can
barricade organizational culture.
On the other hand, assume performance measurement is in alignment with
organizational culture - for example, if our organization were to promote a
culture of innovation, and some of our KPIs were the number of new
products or the percentage invested in R&D, employees would pay more
attention to innovation until such behavior eventually turns into a strong
organizational culture.
#12 Benefits for employees
Organizational performance measurement is not only beneficial for
management, but also for employees. Let’s take a look at how employees can
benefit from organizational performance measurement. The four main
benefits for employees are:
1) Employees will know their scope of work, and understand work
objectives. Performance measurement can communicate important
matters that should be conveyed to employees.
2) Employees can see how their contribution helps the company succeed,
and receive recognition by that success. Performance measurement
shows us the results, and when results are represented, there will be
recognition. No performance measurement, no recognition.
3) Employee evaluations will be more transparent. Performance
measurement results can improve the evaluation process, because they
reduce bias, and enable more transparent and fair assessments.
4) Employees will be empowered. Performance measurement puts
management at ease, because with it, employees will clearly understand
the goals. Performance measurement is already a kind of control tool in
itself, enabling management to “feel comfortable” in assigning more
work. Also, it can give early warnings as problem arise. Hence,
management would no longer be afraid to empower their employees,
because they would know that if their employees make mistakes, they
will receive warnings and can fix the problems early on. Empowered
employees will improve how they work and feel more satisfied with their
jobs.
Now, you can see how performance measurement benefits both management
and employees.
 Chapter 3
On measurement
Having discussed the “why” and the “benefits,” let’s take a look at “what”
performance measurement is, and how many kinds of measurement there are.
Also, I am going to answer the interesting question, “Can we measure
everything in this world?” So, stay tuned.
What is measurement?
Since this book is about “performance measurement,” we need to know what
“measurement” is first. Can you answer that question? “Measurement” means
assigning a numerical scale to the size, value, or other characteristic of a
tangible or intangible object. For example, measuring “the number of
persons” (tangible) or “satisfaction” (intangible).
Do you know the most important aspect of measurement? Take a minute to
think about it. Do not peek below. The most important aspect is
“comparison.” All measures are relative. Without a reference for comparison,
all measures are meaningless numbers.
Let’s assume someone tells you that you scored 56 on an exam. Should you
be happy about it? You cannot answer that without knowing the full score.
That means that you are already looking for a comparison. Now let’s assume
that the full score is 100. Now, many of you would say that you should be
sorry about it because you scored slightly over half. But just wait… What if I
can find another comparison value showing that the average score on this
exam is 15? Now, did the whole world just get brighter? Well, the score is
little more than half, but it is way above average, right? Now, what if I give
you another comparison value? The top score is 56. Well, well, that would be
you! Do you feel like jumping up and down? See? 56 means a lot when there
is a reference point, and this can help in decision making. Hence, we can say
that measuring without comparison is almost useless for decision making,
right?
Let’s get to know different kinds of measurement.
Now that we know the meaning and origin of performance measurement,
let’s familiarize ourselves with different kinds of performance measurement.
There are two kinds of standard measurement:
1) Standardized measurement uses standard scales commonly adopted
around the world. For example, we measure length in meters, weight in
kilograms, and time in minutes. These scales are internationally accepted. If
you measure 1 meter anywhere in the world, the length will be the same. If
you measure 1 kilogram in any country, it will weigh the same. If you
measure 1 minute anywhere, it will last the same.
This kind of measurement makes comparing easy and clear. We can say for
certain that 10 meters is longer than 1 meter, 10 kilograms weigh more than 1
kilogram, and 10 minutes is longer than 1 minute. These facts hold true no
matter where you are in the world.
2) Relative measurement does not use standardized scales like the first type.
An example is “customer satisfaction.” Organizations of the world do not sit
and agree in consensus to use the same measurements or questions for
customer satisfaction. If we were to do that, I believe many organizations
could not reach an agreement, because different organizations have different
kinds of customers and highlights. Hence, having standardized measurement
in such a case would be difficult and inappropriate.
Therefore, every organization should have its own customer satisfaction
measurement; comparisons should be done within the organization itself. For
example, comparing last month’s level of customer satisfaction with this
month’s will be meaningful, because we will have used the same instrument.
However, we should not compare this type of value across organizations
because instruments and questions will be different. It would be “comparing
apples and oranges.” You just cannot compare them, right?
We can also divide measurement into two more types based on the
measurement method:
1) Direct measurement is used to measure tangible things, and the outcome of
the measurement would be what we want to know, e.g., number of
employees.
2) Indirect measurement is used to measure intangible things, so we would
need to measure their outcomes, e.g., measuring employee morale through
absenteeism because we might be assuming that absenteeism is caused by
low morale.
Now, many of you might be wondering which type of measurement is the
best one. There is no definitive answer to that question. It depends on the
situation. For example, you have one water pipe at home, and you want to
compare its length to the one your friend in the other country has, to see
which one is longer. In this case, the first thing that you would do is to use a
standardized measurement, e.g, measure in meters so that you can compare
the result. However, if these two pipes are in the same room, you do not even
have to measure. You can just simply put them right next to each other to
find out which one is longer.
Same goes for business. Ask yourself before you start measuring:
1) What question are you trying to answer?
2) What kind or method of measurement can you use in answering that
question?
3) What kind or method of measurement would be the most efficient and
reliable?
Ask these questions each time you are about to choose the kind and method
of measurement to use.
Can everything in this world be measured?
This question is very intriguing. The answer is “yes and no.” It all depends on
what you want to measure. If you want to measure with a 100% accuracy that
no one can contest, then you cannot measure everything in this world,
especially abstractions, such as love, goodness, and loyalty. These things are
all abstract, and their definitions might vary from person to person. For
example, some people might measure “the level of love” by the number of
times people call each other, by the value of the presents they give each other,
or the number of times they miss each other. Those measurements cannot
reflect “love” with 100% accuracy.
However, such measurements might be useful if the main objective is to
acquire data for decision making. For example, if people never call, never
give gifts, and never miss each other, then these things would reflect a certain
thing, right? In fact, “immeasurable” implies that something cannot be
measured directly. However, things can be measured indirectly, meaning we
can measure their outcomes. Measurements like the number of calls, the
value of presents, or the number of times two people miss each other, are
what we believe the outcomes of love to be (and you might be able to
recommend better measurements than mine.) These might not be 100%
accurate, but somewhat useful in decision making processes.
In the world of business, measuring things that are (or seem to be)
“immeasurable” is quite normal. This kind of measurement happens all the
time without us noticing it (TV ratings, customer satisfaction surveys, the
quality of a movie by the count of awards received). All of these indicators
measure abstract things. Hence, we cannot measure directly, but rather
indirectly. For example, “popularity” is intangible, so we measure the
“viewing rate,” or the so-called “TV rating” instead, because we believe that
if one likes a TV show, he or she would turn the TV on to watch it.
Rating measurement is simple. One can simply install a measurement box at
the representative sample’s houses. Once a participant watches a show on a
particular channel, this box will collect the data. Rating is the percentage of
the sample watching a show at a certain period of time. We believe that high
ratings “indicate” the popularity of a particular show. Or, the number of likes
or shares would “indicate” the popularity level of a post on a Facebook page.
When we measure what seems to be immeasurable, most of the time, we only
use two gauges like “yes” or “no” – yes or no, love or hate – and we count
that as measurement. Let’s say before we go to see a movie, we want to know
if it is good or not. In the old days, we used to look at reviews from two
famous reviewers: Roger Ebert and Gene Siskel. At the end of their reviews,
they would give the movies a thumbs up or a thumbs down. Thumbs up
means that the movie is worth watching. Thumbs down means the movie is
not good. This is a very simple kind of measurement. A movie that receives a
rating of two thumb-ups meant that both reviewers agreed that it is a good
movie, so most people would be interested in that movie (because it did not
happen very often.)
Now you can see that performance measurement is not a complicated thing.
If you know what it means, then you can use it. Like in the case of movie
reviews, even when a movie receives two thumbs-ups, it does not necessarily
mean that everyone will like it. But at least, it lets you know that two famous
reviewers like the movie, which is useful.
Quantitative measurement vs. qualitative measurement
In addition to categorizing performance measurement by type and process,
we can also divide it into two forms:
1) Quantitative measurements are mostly direct measurements that do not
require much interpretation, such as counting and quantifying.
2) Qualitative measurements are mostly indirect measurements that are based
on human decisions, such as measuring employee satisfaction.
We can observe examples of both forms of measurement everywhere. For
example, they were used during the Olympics. The measurement used in
sprinting or broad jump competitions was mainly quantitative; whereas, for
gymnastics or diving competitions, measurement was qualitative.
Have you ever noticed that the sports that usually have problems with scoring
are the ones that use qualitative measurement, like gymnastics? That is
because it requires judges to give scores, which can be biased. However, that
does not mean that qualitative measurements are not reliable. We can use
different techniques to reduce bias; for example, we can have more than one
judge. Once all scores are in, we can cut the highest and the lowest score out
to reduce bias, or we can even issue detailed scoring regulations. For
example, if a gymnast makes a certain mistake, points might be deducted.
These methods can be utilized to reduce bias in qualitative measurement, and
provide more accurate results.
Same goes for the world of business. Most of the time, when people say that
something is “immeasurable,” it means that it cannot be measured directly. If
you cannot measure something directly, then the question you should ask is,
“What outcome does that task produce?” From there, you can measure the
outcome, indirectly. If there is no outcome that can be measured, it means
that the thing cannot be measured, which implies that the particular task
should not be attempted in the first place, because it does not affect your
organization in any way.
Performance measurement development
Having familiarized ourselves with the definition of performance
measurement and performance measurement systems, let’s now learn about
the development of performance measurement before getting into further
details. In fact, we might have to rewind to hundreds of years ago, because
humans began measuring profit and loss since the establishment of the
monetary system. However, to talk about performance measurement for
organizational development, we would have to visit some time after World
War II, the mass production era.
The notion of that era was to produce large quantities to reduce the cost per
unit. Once the cost per unit is reduced, customers are more interested in
buying, because they appreciate cheap things. Hence, it should not come as a
surprise that one of the most vital organizational performance measurements
during that era would be productivity.
Productivity is a performance measure obtained by dividing the output by
input. The higher the value, the more likely we are to succeed, because we
have used less input to produce a large quantity of output. This performance
measure had always been the most important organizational measurement
until the 80s, when people began to question whether measuring productivity
alone would be enough, because we have to wait until everything is finished
to find out the value of productivity. This measure is not that useful, because
it cannot fix anything. It is like finding out where a hole is by looking at the
rearview mirror of a car that has already fallen into a hole.
Wouldn’t it be better if we could find a measurement that can predict the
future? This question is the origin of organizational performance
measurement as we know it. Sink and Tuttle were one of the first people to
write a book about organizational performance measure, called Planning and
Measurement in Your Organization of the Future. The book inspired the
notion of organizational performance measurement.
During the transitioning period of the 1980 - 2000, many organizational
performance measurement concepts were introduced: Balanced Scorecard,
Six Sigma, Economic Value Added, and Activity-Based Costing. After the
year 2000, less organizational performance measurement frameworks were
introduced. However, the focus had shifted to putting existing frameworks
into use. More organizations had adopted performance measurement, and
integrated it into their management systems. Hence, people started to use the
term performance “management,” instead of performance “measurement”.
Today, performance measurement is becoming more significant. It is related
to various branches of study, which perceive “performance” in different
ways. For example, marketers might understand performance as brand
awareness, while production managers would see it as utilization percentage,
and financiers as profit. The definition of performance will evolve. And I
truly believe that one day the knowledge on organizational performance
measurement will be a branch of study itself, and eventually, it will be no less
important than finance, marketing, and production.
 Chapter 4
Cautions of performance measurement
Performance measurement is quite similar to taking medication. If you take
the right medicine, your will be cured. However, taking the wrong medicine
might cause other damages to your health, instead of healing you. Even the
best medicine has side effects, and so does performance measurement. If you
measure it poorly or incorrectly, there will be side effects. And if you do not
realize your mistake, sometimes it can become a fiasco. Hence, in this
chapter, I would like to tell you about the possible side effects of using the
wrong performance measurement method.
#1 Don’t use performance measurement to punish.
One thing that we need to comprehend is that the term “performance
measurement” itself has a negative connotation. Think about the times when
you were a student. A teacher walks in and says, “We have a pop quiz
today.” How did you feel? I think most students would feel uneasy. Or, at the
workplace, your boss walks in and says, “Today I’ll assess your
performance.” The feeling would be common, something along the line of
stressful.
The right amount of unease or stress is a good thing, because it motivates us
to work harder. However, we should be careful about not being too stressful
or anxious, because it will have more negative effects than good. Hence, we
should not associate our organizational performance measurement with
punishment, but rather with rewards. Associating performance measurement
with punishment will reinforce negative effects. Although punishment might
yield faster results, you might lose more than you gain.
Imagine your boss announcing, from now on, the company will have KPIs,
and whoever fails to meet KPIs will be fired. What would happen?
Employees might go on strike, or spread rumors that the company wants to
downsize. Employees might see KPIs as their enemy, instead of a tool that
would improve the organization. Nothing could be worse than turning
performance measurement into an enemy of employees. Apart from the fact
that there will be no improvement, it might even hurt the organization. So, be
careful!
#2 Be careful not to cause stress.
Have you ever read the news about some college student jumping off a
building because he or she felt disappointed? Sometimes when you read the
details, you just do not want to believe your eyes. The student committed
suicide because he or she did not receive an A on an exam, or did not gain 4.0
like she previously did!
I remember a college student being caught cheating on his final examination.
At first, I thought he had done so due to his really low midterm exam score;
he must have been afraid of getting an F. But it turned out that he had
received a really high midterm score, and he cheated on the exam because he
was afraid of not being able to get an A. I even made a joke to the students
who received low scores: “Well, if it were you guys, I wouldn’t have been
surprised.” However, one of students replied, “I wouldn’t cheat on exams,
because even if I have a cheat sheet, I would still get an F.” My point is that
performance measurement might cause stress. We tend to overlook good
performers, because we know by heart that they are good, so we think that we
do not need to worry about them. In fact, this is what you should be worried
about.
When we are stressed out, we try to find a way out. If we choose the good
way, then it is alright. However, if we decide to go with the wrong way, it
can harm the organization. For example, if you measure salesperson
performance by the amount of sales they make, you might forget about the
top performing salespersons, and focus on the low performing salespersons
instead. But top salespersons can also be stressed out. For example, imagine
this top salesperson who is always number one in sales rankings, but this year
he is not. “Over-stimulation” usually happens to the best of us, to “over-
achievers.” When he is afraid that he will no longer be number one, he will
try to find a way out. Finding a good way out such as improving his own
sales skills is totally fine. However, if he chooses a bad way out, like cheating
or stealing a colleague’s sale, then you should beware. As such, instead of
stimulating improvement, performance measurement might cause unwanted
behavior that might harm the organization.
#3 Be careful not to create a culture of excuse.
Another thing you should be cautious about is to not let poor performers stay
as they are for a long period of time. The simple reason is that no one wants
to be rubbed in the face that they have failed. Even when you do not say it to
their face, they might already feel like a failure because they could not reach
the goals. So, what would be the consequence? The low performer might try
to improve himself. Self-improvement is a desirable outcome of performance
measurement.
However, this method might be difficult, and require such effort that some
people might not choose it. Rather, they might end up choosing the second,
much easier method: making up excuses such as “the performance
measurement system is not good,” “the goal is too difficult to achieve,” or
“the resources are too limited.” This is the easier way-out than self-
improvement, as it requires only words.
If things turn out like this, there will be no improvement. Let’s use health
checkups as an example. The results show that your cholesterol level is
higher than the standard… What should you do?:
1) Start working out, and reduce unhealthy diet to be healthier.
2) Blame that the cholesterol test is bad, or the standard is too low, and
continue to live your life as is.
I believe most people would choose number 1, but in the workplace, we
might observe many people choosing number 2. Sometimes, measurements
might be flawed, and deserve rectification. If the performance measurement
is really flawed, then we should improve the system. However, if this is just
an excuse, then we should pay attention to the problem. We should not blame
the person, but rather we should help those who could not reach the goal. I
believe that most people do not fail intentionally, rather, they fail because of
certain problems. Performance measurement will help identify these
problems so that everyone can work together to fix the problems to improve
the organization.
#4 Be careful not to let the goal become the ceiling.
Goal setting comes along with performance measurement. A good goal can
motivate people to work. In contrast, an inappropriate goal will not only
discourage your employees, but also become an operation ceiling. What does
operation ceiling mean? Here’s an example. If you set an easy goal, then once
your employees have achieved that goal, they will no longer be motivated.
For example, you tell your salespersons that if they can achieve one hundred
thousand dollars of sales revenue, you will take them on a trip to Europe.
However, it turns out that the sales revenue reached one hundred thousand
dollars in June. What do you think would happen next? Of course,
salespersons will not be motivated for the next 6 months, because the revenue
goal is too low that it has become a ceiling. If you continue like this, you will
lose more salespersons. You might also end up losing your top performing
salespersons (otherwise they would not be able to achieve the goal this fast).
Therefore, you have to be cautious when setting a goal.
#5 Beware of delayed results.
What do you expect to gain from performance measurement? Many people
would say “good results.” Well, that is correct. But what if the results turn out
to be bad? For example, if you know you did poorly on an exam, would you
still want to know the result? I bet many would say no. But deep down you
do want to know the result, right? No matter how bad it could be, you still
want to know, because once you know, you can make decision.
For example, finding out that you scored very low on an exam is still useful,
because you will decide whether you should withdraw or try harder to pass
the exam. Imagine that you are a freshman, and you have to take an exam for
each course. However, the professors never tell you what your grades are,
and all you can do is to continue to study. Now, it is your last semester as a
senior, and they will tell you all at once: what your GPA is, the list of courses
that you passed and failed, and whether or not you managed to graduate. I
believe you would feel anxious throughout your college years. Performance
measurement makes people who are being measured expect the results. If
possible, the results should be delivered fast, and on time.
#6 Beware of perfection.
Performance measurement does not have to be 100% accurate. What?! Why
would you say such a strange thing? Performance measurement must be
correct. Yes, it has to be correct. But it does not have to be 100% correct.
Most of the time, perfection requires huge investment, and 100% accuracy is
just not worth such huge investment.
Let me give you an example. Let’s say we want to measure customer
satisfaction by using a 5-scale survey, where 1 = very dissatisfied, and 5 =
very satisfied. Let’s assume that we had measured customer satisfaction last
month as 3.5, and for this month we got 4. Now someone tell us that our
current customer satisfaction measurement is not detailed enough, and the
scale should be smaller. For example, instead of asking 20 questions, we have
to ask 200 questions; and instead of using a 1 to 5 scale, we have to use 1 to
10. What does this mean? The most important thing is investment in data
collection. Asking 200 questions and using a scale of 1 to 10 would make it
nearly impossible to find respondents who would be willing to complete our
survey. We would have to pay them to fill out the questionnaire. Hence, data
collection would incur more cost. For example, we might need to hire or
provide incentives to respondents. If we give each respondent 5 dollars, and
we want to collect 400 samples, then we need to invest 2,000 dollars.
What can we get out of this? Many might say that we will get a more accurate
result. Yes, it will be more accurate. For example, we would learn that, in
fact, last month’s customer satisfaction was 3.5234, while this month’s is
4.3245 (using scale of 1 to 5 for easier comparison.) Wow, that is more
accurate… However, does it really increase the value of the data we have?
Last month’s score of 3.5 improved to 4 this month (old measurement),
compared to 3.5234 and 4.3245 respectively (new measurement). Our
interpretation would be nearly the same. We would also get the same
implication that customer satisfaction has improved. The only difference is
the investment in data collection; the latter incurs 2,000 dollar more.
Such amount of investment might not affect large organizations so much.
However, still, if repeated many times, we would lose quite a lot of money.
The same amount of investment might have more impact on Small Medium
Enterprises (SME). In conclusion, I do not mean that we should collect
incorrect data, but rather, that performance measurement does not need to be
100% accurate. Sometimes, 80-90% accuracy is enough, because the
comparison is more important than the actual number.
#7 Beware of depending solely on quantitative measurement.
Many people prefer quantitative measurement that shows numbers or
amounts of money, because it is clear. Whereas, qualitative measurement
relies on human judgment, and many people don’t like it. In fact, measuring
by human judgment is nothing new; it is called subjective measurement.
Most people are unsure about using subjective measurement, because its
results can be contested since people might have different points of view.
For objective measurements, such as the number of employees, the result will
always be the same; no matter who measures it, we will get the same number
of employees as a result (unless, of course, someone miscalculates.) Whereas,
for subjective measurements like employee satisfaction, can easily be
challenged. Although the average employee satisfaction index is high, many
people might think that it does not reflect overall employee satisfaction. Well,
that might be true. When measuring intangible things, we have to admit that
we will not be able to find a performance measure or measurement method
that would provide us with true values that cannot be challenged. The more
important question is, “Can we put measurement to good use, despite it not
being 100% accurate?"
For example, if we observe that certain aspects of employee satisfaction has
been decreasing, can we utilize that data? The answer is yes, we can. I
believe that even though measurement based on human judgment might not
be 100% accurate, it is still more useful than not measuring anything at all. In
fact, this kind of measurement has not been recently invented. Businesses
have been using it for a really long time, and many organizations achieved
success due to their implementing this kind of measurement. Performance
measurement will always be of use, one way or another.
#8 Beware of incorrect measurement.
How can you know whether your current organizational performance
measurement is effective or not? I think this is a very interesting question.
Usually, we would consider two dimensions to determine the effectiveness of
our organizational performance measurement.
The first dimension is validity. Are the things that we measure and the things
that we want to know one and the same? For example, I want to know my
weight, and I’m using a ruler to measure my height instead. This
measurement has no validity, because the result will be a value of height
instead of the weight that we wanted to know. Some of you might think, “But
who would do that in the world of business?” Well, let me give you an
example from the business perspective. Let’s say we want to know about our
customer loyalty, what should we measure? Most people would measure
“repurchase rate.” It is true that loyal customers would repurchase our
products; however, customers who repurchase are not always loyal. Why
not? Because repurchase derives from many reasons, such as convenience.
I am a frequent customer of a certain bank, because there is a branch near my
workplace. But I am not that loyal to this bank, because I know for certain
that if this branch moves elsewhere and there is a branch of another bank to
replace it, I would immediately switch. Hence, measuring customer loyalty
solely with repurchase rate still has a validity problem. Besides repurchase
rate, we should also measure recommendation rate or other things to increase
measurement accuracy.
The other dimension is reliability. If you use the same measurement multiple
times during the same period, you should get similar results. If results are
similar, it means reliability is high. If not, then that measurement has low
reliability. For example, I believe everyone can rank three of their best
friends from the closest to the least closest. If I ask you to do it once more,
the result would still be the same. This means the reliability is high.
However, if I ask you to rank 100 of your closest friends, I might get a list of
100 names the first time. If I ask you to do it again, do you think the list will
be the same? I think not. This is what we call low reliability.
Validity and reliability are both very significant to performance
measurement. If you want to measure, then measure right. Am I right?
 Chapter 5
Get to know the system of performance measurement
In the first four chapters, I have discussed the reasons behind using
performance measurement, its benefits, origin, and cautions. Now, I want to
introduce the “system” of performance measurement. Let’s see what the
system looks like and what it is used for.
Let’s get to know the system of performance measurement.
Before we dive into designing performance measurement systems, let’s
familiarize ourselves with “organizational performance measurement
systems” first. Organizational performance measurement system is a
management tool, composed of two main parts:
1) Organizational performance measures
2) Supporting infrastructure
This is where things get more interesting. Many of you might think of only
the performance measures, aka. KPIs, when talking about organizational
performance measurement systems. Performance measures are vital parts of
organizational performance measurement systems. However, you should
never ignore the fact that the system also consists of a supporting
infrastructure. There are two main supporting infrastructures: the data
collection system and the reporting system.
Sometimes you might see a failed organizational performance measurement
system that is not a product of poor performance measures, but rather an
erroneous supporting infrastructure, such as incorrect data or poorly prepared
reports. Knowing the definition of an organizational performance
measurement system, we would know how to fix the problem at its cause,
which will help improving organization management.
What are performance measurement systems used for?
For the most part, a performance measurement system is a tool that helps the
management of an organization. Then the question is what are the main
responsibilities of management? There are three. The first is to formulate a
long-term plan, aka. a strategy. Next one is to deal with day-to-day operation
problems, because everything does not simply go 100% according to plan.
Hence, “problems” occur, and it is the management’s job to deal with them.
The last one is to improve work processes. Just because there is no current
problem, it does not mean that management can stop working, as
management should continue to improve work processes. See? Being a
manager is not a simple task. This is why managers deserve assistants to help
accomplish their tasks. Let’s move on to the next topic and see how
performance measurement systems can help management accomplish
different tasks.
Devise a strategy to create results.
As mentioned earlier, one of the main responsibilities of management is to
create a strategy. Devising a strategy requires two important procedures:
1) Setting long-term goals and implementing intentions, and
2) Making sure every employee understands the goals and is willing to
work hard to achieve them.
Generally, organizations have a vision, mission, and values. These three
things are an organization’s compass.
Vision tells us what an organization would like to accomplish in the long-run.
I would like to use a large global company like Coca-Cola as an example.
Coca-Cola’s vision is based on six important pillars: people, portfolio,
partners, planet, profit, and productivity.
Mission answers the question as to why an organization exists. Why has the
organization been established? Coca-Cola answers this question simply with
three sentences: to refresh the world, to inspire moments of optimism and
happiness, and to create value and make a difference.
Organizational value are the principles that guide an organization’s conduct,
because the organization believes that these principles will play an important
role in the long-term survival of the organization. Simply put, values make up
the personality of the organization. Coca-Cola’s values consist of seven
aspects: leadership, collaboration, integrity, accountability, passion, diversity,
and quality.
You can see that every global organization has a clear vision, mission, and
value. Once an organization has a vision, mission, and value, it must
transform these three things into something tangible. And that establishes the
organization’s main objectives. In the end, these objectives will be translated
into operational activities. One of the most vital things in transforming
strategy into practice is to establish “business processes.” But what exactly
are “business processes”? It is a defined sequence of actions that would help
an organization achieve its predefined objectives.
For example, a university aims to produce goodhearted graduates for the
society. If the university sets the objective but does nothing about it, then of
course it would not be able to produce goodhearted and skilled graduates for
the society. Thus, the university has to define a certain business process, and
that process is teaching and learning. Therefore, we can say that an
organization is a consolidation of various processes. How can we push these
processes forward and achieve the predefined objectives? This is where
operational performance measurement will come in handy with the simple
principle of “you get what you measure.”
If you want to develop a business process the way you want, then you have to
create a control system. However, using physical control would incur higher
costs. Let’s revisit the previous example. How can you be sure that the
teaching and learning process designed by the university will produce
goodhearted and skilled graduates? If you rely on physical control, for
example, you hire a person to listen to the instructors and observe the
students, and have them report on the process. You would have to invest a
huge amount of money in hiring a lot of these people. Also, both the
instructors and the students would feel uncomfortable with this method.
However, if there is no control at all, there is a risk of instructors skipping
classes or not teaching their best, and the students not paying attention
(because there is no performance measurement.) In the end, this process
would fail. But if a good performance measurement is in place, such as a way
for the students to assess their instructor on whether he or she is good at
teaching and works hard or not, then the instructors would not be able to
neglect their duties. The instructor has to assess the students as well through
examination - this way students will pay attention in class. Such teaching and
learning process will prevail, because it can create goodhearted and skilled
graduates.
Therefore, operational performance measurement is one of the tools that can
help manage an organization. When good business processes are predefined,
the strategy will be effective. Now let’s see what we can measure in order to
control the processes. Normally, we would measure input, process, and
output. Let’s use the manufacturing process as an example. Input measures
are as follows:
Measure input quantity; for example, the quantity of raw materials
used in production
Measure the quality of input materials; for example, the quality of
raw materials
Measure the quality of input procedural quality; for example, the
timeliness of raw materials delivery
Measure other environmental factors related to input; for example,
price of raw materials (these are the things we cannot control but
have an effect on the organization)

Process measures are as follows:

Measure the amount of resources used; for example, the resources

used in production, or the machine power or labor used in
production
Measure the losses as a result of the process; for example, the defect
rate

Output measures are as follows:

Measure outcomes; for example, the quantity of finished goods

Measure the quality of outcomes; for example, the quality of
finished goods
Measure the quality of output process; for example, the timeliness of
product delivery to customers
Measure productivity which is an economic measure of output per
unit of input

These measures will control and ensure that the procedures would work
smoothly. If a problem occurs, you would be informed and be able to deal
with it right away, which would then lead to organizational success.
Help in day-to-day operations
Apart from devising a strategy, what else do managers do in their lives?
Asking a question like this is asking for trouble. But I do not mean it that
way. On the contrary, I asked that question as a lead-in to the main
responsibility of management and to the how-tos of putting performance
measurement to use. Managers do not spend their time only on formulating a
strategy or creating a long-term vision. They also have to oversee and ensure
that everything is operating effectively and efficiently.
For example, in production control, managers need to make sure that
production runs smoothly, without any problems. But how can a manager
know what’s going on in production? Performance measurement is the tool
that will help the manager find out whether everything is running smoothly or
not. With performance measurement, managers do not need to gather
information from various sources, which is time consuming. In addition,
when performance measures indicate a problem, such as a decline in the
timeliness of production, managers will be immediately notified of the
problem so that they are able to fix it on time. Many performance measures
might even demonstrate the roots of the problems. Generally speaking,
organizational performance measurement is the manager’s right-hand man, as
long as the system implemented is correct and relevant.
Help in organizational development
Besides working on long-term plans and preparing daily reports, managers
have to improve and develop current business processes. Many managers
think “Solving daily problems takes the whole day. How can I have enough
time to improve procedures?” The catch is you are spending time to fix daily
problems because you did not try to improve your process in the first place.
For example, if you never care to improve your production and leave the
machines broken, you would feel anxious that you will not be able to produce
good quality products or produce goods on time. You have to spend your day
seeking for products from somewhere else to be delivered to customers.
Taking a step back to improve the process will definitely be worth your time.
To develop a better process, you have to work harder only once. After that,
you might find that you have more time to work. Performance measurement
will show you what you need to improve, so that you can solve the root cause
of the problem.
However, I have to admit that performance measurement is rarely used for
this purpose. Performance measurement for organizational improvements is
like getting a health checkup. You do not get a checkup because you are sick,
you get it because you are afraid of getting sick, right? If the result shows that
you have certain problems, like high level of cholesterol, you will start to
improve yourself by exercising more or eating a healthy diet. Organizations
need checkups, too. Have you checked the health of your organization yet?
 Chapter 6
Things you should know before creating a performance measurement
system
If you have made it to this chapter, by now, you probably want to know how
you can design a good performance measurement system. Wait a little longer,
do not rush it. Before you can design a good system, you need to know what
a “good” system is first. It is like when you want to design a house. Before
you start designing, you have to know what your ideal house is like. So, let’s
take a look at the important factors that you need to know before designing a
performance measurement system.
#1 A performance measurement system must eliminate existing
problems without creating new ones.
I have a story to tell you. A company has a problem delivering its products to
customers on time. To fix this problem, they create a performance measure:
the percentage of late delivery. This performance measure divides the number
of products that could not be delivered on time by total production. If a
customer orders 100 pieces, they produce exactly 100 pieces. However, only
80 pieces are delivered on time, and not the other 20. Hence, the percentage
of late delivery equals 20/100 = 20%. The CEO aims to reduce the
percentage of late delivery to less than 5% within a month. If this goal is
achieved, then he will give out rewards. One month later, it turns out that the
percentage of late delivery is reduced to 2%. However, the company still
receives complaints from their customers about late delivery.
What happened? The company still could not deliver the products of the
same quantity on time even though the percentage of late delivery decreased?
Why? Let’s stop here and take two minutes to think about it.
The answer is that the employees tried improving the percentage of late
delivery, however, such action was not in alignment with what the
organization wants. Do you know what they did? Let’s take a look at the
calculation again. The number of products that could not be delivered on time
is divided by total production. Hence, there are two ways to reduce the
percentage: (1) Reduce the numerator (the number of products that could not
be delivered on time), or (2) Increase the denominator (total production).
Which one do you think is harder? The employees of this company chose
option 2, because option 1 demands hard work. Whereas, with option 2, they
only need to produce more than what the customer orders. In this case, they
produced more than needed. For example, when the customer ordered 100
pieces, they produced 1,000 pieces. And since they did not rush production,
they could manage to produce only 80 pieces on time, while the other 20
were still delayed. Yet, the percentage of late delivery was reduced because
the value of parameters in the equation have changed: 20/1,000 = 2%, instead
of 20/100 = 20%.
This kind of problem occurs because the performance measure creator does
not pay attention to the overall picture, and creates additional problems like
over production. So how can we fix the late delivery problem in this case?
The first method is to adjust the performance measure so that it is aligned
with what we intend to measure. Using the above example, we can calculate
the percentage of late delivery by dividing the number of products that could
not be delivered on time by the number of products the customer ordered,
instead of total production. With this formula, even if the employees produce
more, the performance measure will not be affected. The only way out for
them to reduce late delivery would be to produce and deliver faster.
The second method is to involve other performance measures. For example,
we can include an inventory level measure, so that when the employees
overproduce in order to improve the late delivery percentage, the inventory
level will also increase. When other performance measures are included,
employees would not dare do what the organization does not desire. In this
case, the employees would not overproduce.
The third method is to create a restriction for the performance measure. For
example, we want to reduce the percentage of late delivery and not allow
overproduction. This way, employees will act according to the organization’s
goals. You can see that designing a performance measure is not an easy task.
You have to give careful consideration to it. Otherwise, the performance
measure will become the troublemaker.
Let me give you another example. There is a city with ferry service in the
United Kingdom. The problem is that the ferry captain receives a fixed
salary, so he does not have the motivation to work hard. As a result,
customers begin to complain that the captain is not eager to work. Despite
having full passengers on board, the captain continues to chit chat, and does
not hurry back to pilot the ferry. Once customer complaints reach the ferry
service company, management begins to invent a performance measure that
would solve this problem. The performance measure they come up with is
“the number of trips per day,” because they believe this will make the captain
work harder. Now, I bet everyone is able to guess what’s going to happen
next.
Yes, once you count the number of trips, you will get numbers of trips; “You
get what you measure.” What happens is that the ferry captain does not wait
for passengers to board the ferry, because he knows the more “trips” he
makes, the more rewards he will receive, and he has to achieve the required
number of trips. Hence, waiting for passengers to get on board is a barrier to
his goal. So, customers find that now the captain would not wait for anybody.
Once he had 3-4 passengers on the ferry, he would leave right away. See how
a performance measure that has not been carefully considered can affect an
organization very badly? In the end, this company had to adjust their
performance measure to be based on the revenues received from the trips so
that the goal of the captain and the organization are the same, which is to
generate profit.
Let’s take a look at another example. This happened in a supermarket
somewhere in the United States. The problem with this supermarket was the
long waiting line, especially during prime time. The long waiting line
problem is nothing new. However, customers were particularly dissatisfied
with the sluggish cashiers. As the cashiers were receiving a fixed hourly
wage, they did not have to work fast; no matter if they worked fast or slow,
they would get the same wage. When the cashiers worked slowly, the
customers in line felt dissatisfied because they did not understand why the
cashiers would not work faster, given such a long waiting line. Most of the
time, the cashiers even chit chatted with each other.
To solve this problem, the supermarket’s operation manager set up a
performance measure for cashier performance. This performance measure
focused on the time a cashier spent in servicing a customer. But the problem
was that counting the amount of service time was unfair for the cashiers,
because the cashiers that get to serve the customers who buy few items would
obviously spend less time, and vice versa. Therefore, measuring by dividing
the total service time by the number of purchased items was more rational.
The total service time would start when a cashier begins to scan the first item,
and end when the cashier scans the last item. Then they were able to divide
the total service time by the total number of scanned items. In addition, the
management of this supermarket also found the statistical value of the
performance measure, and defined a goal of spending less time in scanning
an item that rewarded the cashiers.
One week later, the performance measure showed significant improvement;
time spent per item was greatly reduced. Technically, the time spent in the
waiting line should have gotten shorter, right? However, that was not the
case. The time spent in the waiting line got even longer than before. How is
that possible when the checkout time per item is reduced? How come did the
waiting line get longer?
Here’s what happened… The performance measure started measuring when
the first item was scanned, and ended when the last item was scanned. It is
true that the decreased value reflects the reduced amount of time, when
compared to the checkout time of the same quantity of items. However, it
does not reflect the total waiting time. Once this performance measure was
implemented, the cashiers were under the pressure of having to improve the
value (the value had to be reduced). Thus, the cashiers began to take certain
actions to reduce this value. However, these actions became the reason why
the waiting line got longer.
What did the cashiers do? Firstly, when a customer approached the counter
and took items out of her cart, normally the cashier would start scanning the
items right away. However, because of this performance measure, they did
not do that. They would wait until the customer finished unloading the cart,
because they knew that the clock would start ticking when they would begin
to scan the first item. The customer’s unloading her items very slowly was
affecting this performance measure.
Another possible situation in this scenario is when a customer unloads his
cart for the cashier to start scanning, but then he realizes that he forgot
something. In the past, the cashier would continue to scan the other items
while the customer ran to grab the thing he forgot. With this performance
measure in place, the cashier would not do that, because whenever he starts
scanning, the clock starts ticking. If the customer takes a long time, the
cashier’s performance measure would indicate that he is doing a bad job. So,
we would see cashiers standing and waiting for their customers to come back
before starting the scanning process.
Thirdly, what if the customer realizes that he forgot something after the
cashier starts scanning? In this case, the cashier would not be able to prevent
the customer from going back and grabbing something he wants. But at the
same time, the cashier cannot stop scanning, so he has to continue what he’s
doing. But what might happen is that once the cashier finishes scanning, he
might finish the transaction without waiting for the customer to return, not to
prolong his total transaction time. Despite the fact that he might see the
customer already walking back, he would still finish the transaction to stop
the clock, and then start a new transaction once the customer is back at the
counter. Finishing the previous transaction means that the customer has to
pay and sign his credit card for that transaction before the cashier can start
scanning again for the new transaction. Therefore, the waiting time would be
longer.
Now you can see that designing a performance measure is not an easy job. It
is true that in order to reduce waiting times, many actions have to be taken –
from increasing the number of checkout counters to other matters. But what I
am trying to say is that if you design a performance measure without
considering other factors, the performance measure can create new problems
without you even realizing it.
#2 A performance measurement system must explain the performance
gap.
Once you have received your organization’s performance measurement
results, you must ensure that you know what the result means. And more
importantly, you need to know how you can improve it. Normally, you would
receive two values from any performance measure. The first value is the
actual value, and the second one is the expected value. Most of the time,
these two values are different. You might find that the actual value does not
equal the expected (or target) value. The gap between the actual value and the
expected value is called the “performance gap.” It is very important that you
should be able to explain the performance gap once you have a performance
measure in place.
Let me give you an example. Assume that we have a measure called “the
time wasted without creating added value to production.” Of course, we
would want the value to be zero (expected value). However, once we receive
the results, we realize that we have wasted 100 hours without creating any
additional value. The performance gap in this case is 100 hours, or simply
put, there is a 100-hour difference from the desired value. Just reporting the
100 hours would be meaningless for the workers, because they would only
know that they have wasted 100 hours without knowing any of the reasons
behind it.
Hence, a good operational report should contain useful information that can
be utilized in problem solving or further development. Therefore, being able
to explain the difference is very vital. However, it is not necessary to explain
the difference 100% accuracy; 80-90% would do for a start, and later on we
can try to achieve about 95%.
From the aforementioned example, apart from being able to tell the amount
of time wasted, we can also tell that out of those 100 hours, 50 hours were
caused by a broken machine, 30 hours by human error, and 15 hours by the
late delivery of raw materials. Now we can use the result to solve problems. It
is much better than only knowing that we have lost 100 hours, am I right?
#3 A performance measurement system must include sufficient data.
You cannot start your car, so you take it to an auto shop. A technician tells
you that if you let him overhaul your car, he can guarantee that the car will be
able to start. Well, that would leave you with empty pockets. A technician
should rather inspect your car in detail, and identify which systems or parts
don’t perform well, and fix those. That would be a wise move costing much
less than overhauling your car. This problem occurs when your performance
measure is not specific enough. If you only measure input and output, and
nothing in between, you will not know the cause of the problem.
Another example… We have two products, Product A and B, but we measure
the total sales instead of breaking it down to sales of A or B. What could
happen? You might see that the total sales might be stable across time.
Looking at the stable total performance, most people would assume that the
sales of both Product A and B are stable. Yes, it is possible that the
performance of A and B are both unchanged, explaining the total
performance’s remaining constant. However, that might not always be the
case. It is also possible that A’s sales is declining, while B’s is improving.
Yet, when the sales of the two are added up, total sales appears to be
constant.
Same goes for stock market indices. When indices remain constant, it does
not mean that the stock price of every company in the market remains
constant. It is possible that the prices of half of the stocks have increased,
while the other half have decreased. Therefore, we can conclude that
performance measures that are not detailed enough might hide both
opportunities (like the increase in sales of B) and threats (like the decrease in
sales of A). As a result, the organization might miss out on many things.
#4 A performance measurement system must be timely.
Can you remember what you had for dinner yesterday? I think most people
would remember. Let me ask you another question. What did you have for
dinner 15 days ago? Can you remember? Most people would not be able to
remember, unless it was a special occasion like your birthday or the day your
boyfriend or girlfriend broke up with you. Why can’t you remember? I don’t
even have to ask. Who would be able to remember that long ago?
Reporting the results of our performance measures too late reduces their
value, because we cannot remember what had been happening during that
time. Think about it… We get the three-month old value of the percentage of
defects, and we notice that it is higher than usual. Would you be able to
remember what was happening three months ago? Our memory is limited. If
we cannot remember what happened, how can we solve the problem?
To make matters worse, if we only realize the problem of three months ago
just now, we have no idea how much damage was done during the past three
months. This is why I usually compare performance measures to fruits and
vegetables. They can be rotten or expired. If we report the results too late,
they might be useless. Now is the time for the important question. What is
considered “timely” in today’s business world? There is no clear answer to
that. But I can give you a brief answer. Measuring weekly should enable us to
send out reports within the first days of the coming week. Or, if we measure
daily, we should be able to deliver a report by tomorrow. This is what would
be called “timely” in today’s business world.
#5 It takes time to see the results of a performance measurement system.
Have you ever faced a problem in analyzing the results of an organizational
performance measure? Apart from being able to tell if the result is better or
worse, we should also be able to tell what makes it that way. For rational
analysis, we should look at the changes in other performance measures as
well. For example, this month’s customer satisfaction has declined. But why
has it declined? If we observe other performance measures like the rate of
timely delivery, we might find that the rate has decreased. Then we’d know
that the decrease in customer satisfaction is caused by the declining rate of
timely delivery. This would be an easy case to analyze. However, sometimes
analysis could be difficult, because it involves a period of time. What does a
period of time have to do with this? Let me give you another example.
Let’s assume that our sales performance measure is constant or decreasing.
But when we look at other performance measures, we notice that the
advertising expenses have increased. Well, that is puzzling. How can an
increase in advertising expenses make sales decrease? It does not make much
sense, right? In fact, drawing a conclusion like this would be inappropriate,
because we did not consider the period of time it would take for advertising
to create an effect on sales. “A period of time” starts when we begin to adjust
our operations, and ends when an effect is observed throughout a
performance measure “learning cycle.”
Let’s use the same example again. The advertisement that we paid for might
require more than a day or two for it to affect sales. It might even take a
month. Therefore, if you find that your recently launched advertisement
campaign does not affect sales the next day, do not panic. It requires more
time. Thus, you should be aware of the “period of time” when analyzing the
results of performance measures. Simply put, we need to know or estimate
the learning cycle of each performance measure; otherwise, we might make
wrong decisions. In the case of the above example, if we rush things, we
might decide to cancel the advertisement campaign and start to work on
something else without knowing if the advertisement was actually working or
not.
The last concern is what to do about performance measures that have a long
learning cycle. We might end up wasting a lot of money before we find out if
they work or not. Let’s use the same example again. Assume that the learning
cycle of the performance measures for advertisement is one month. Do we
have to air the advertisement for one month? If the learning cycle is very
long, you do not have to wait. If you want to measure using sales, then you
might have to wait a month. But if you are unsure whether the advertisement
works or not, you can try using other performance measures that have shorter
learning cycles. For example, you can measure the number of views of that
advertisement – you should get the result within a week. If we have already
advertised for a week, but no one has seen our advertisement, then you do not
have to wait one month to know for sure that it will not stimulate sales. This
is just an example. You can see that if we really pay attention to our
performance measures, they actually have plenty of uses.
#6 Performance measurement requires consistency.
When should we use organizational performance measurement?
1) When an organization has problems
2) When an organization does not have any problems
That is worth thinking about. Most of the time number one is the case. For
example, you could not deliver goods in time, so you begin to design a
performance measure of late delivery to solve the problem. Or, there is a
problem with production, and you create a performance measure of defects to
deal with the problem. Some people might argue that their organization never
has any problems, until they start to use performance measures, and the
performance measures become the troublemaker. Sometimes that is the case.
The answer to the above question is “all of the above.” Business processes in
organizations are like machines – the processes are repeated over and over.
For example, you produce goods and deliver them to customers, and collect
the money every day. When you stop doing it, it is like you have stopped
doing business. Hence, in order for the processes to operate smoothly, we
need to control them. Therefore, performance measurement is a necessary
tool. But the problem is that many people only use performance measurement
to solve problems. Once the problems have been fixed, they stop using that
performance measure. A risky action...
Let me give you an example. Let’s assume that we have a problem with
production. We are producing a lot of defects. One solution is to create a
performance measure for measuring the percentage of defects so that we
become aware of the problem and find a way to solve it. Yet, once the
problem is solved, we stop using that performance measure. You can guess
what is going to happen next, right? The performance measure is no longer in
place, we lose control, and eventually the problem resurfaces. Many people
would say, “Then, reinstall the performance measure.” Yes, you can start
measuring again, and the problem will be solved again. But if you stop
measuring again, in the end, the same problem will return. Can you guess
what would happen if this persists? Performance measurement would
eventually lose its importance. Even if you start measuring again, the
problem will not be solved. Why? Everyone knows that you will stop
measuring soon, as you always do. It is akin to developing drug resistance.
That is why I always emphasize that you should always measure what’s
important. Inconsistent measurement includes not only data collection, but
also data analysis, data presentation, and implementation.
#7 A performance measurement system must produce comparable
results.
As mentioned earlier, what’s more important than the outcome value of
performance measures is their comparability. It is very difficult to make
good use of outcomes if they cannot be compared. Outcomes that are hard to
compare can cause confusion. For example, if we design “number of defects”
as a performance measure, and find that the “number of defects” has
increased from 100 last month to 200 pieces this month. It seems like the
number of defects is much higher. Also it looks like our production process is
deteriorating. Maybe yes, maybe not. The higher number of defects might not
be caused only by the poor production process, it might also be a result of
more production. The more you produce, the more defects there will be.
Let’s say that last month we only produced 1,000 units, and had 100 defects.
However, this month, we have 200 defects, and produced 10,000 units of
goods. Now we can see that the higher number of defects is not a result of a
low quality production process, but of higher production. If we take a look at
the “defect rate” instead, it was reduced from 10% (100/1,000) to only 2%
(200/10,000). In fact, our production has improved. This is why we usually
create performance measures in ratios or percentages, which is easier to
compare than unit-counts. Be careful when you use unit-counting
performance measures, such as the number of employees who resigned, the
number of defects, etc. to avoid problems like this.
#8 A performance measurement system must not cause unnecessary fear.
Most organizational performance measurement system problems are
“human” problems. Designing an appropriate performance measure, and
considering various relevant factors before creating the measure are both very
vital. However, despite flawless performance measures, problems can still
occur if user attitudes have not been adjusted.
The term “performance measurement” itself makes people feel
uncomfortable. Imagine you are still a student. You probably would not like
these words: "There’s an exam today," "We’re going to have a quiz before
class is dismissed," "I’ll assign you a project to measure your performance.”
These expressions can cause stress. Same goes for work; we would feel the
same way. “If you cannot achieve this KPI, you will be punished.”
These words often create fear and stress, which raises the risk of undesirable
employee behavior, e.g., protesting against all kinds of changes. Sometimes
these fears become the source of false rumors in the organization, such as
"they are going to use KPIs to fire us," "these KPIs are aimed at reducing our
salaries." These rumors can become a barrier to organizational improvement.
Hence, we should deal with these fears before they create problems. Fears
related to performance measurement result from two main reasons:
1) Afraid of being punished, scolded at, embarrassed, etc.
2) Afraid of colleagues getting punished, fired, etc.
Number 1 and 2 can be equally damaging. Therefore, managers have to be
aware of these fears if they plan to implement performance measurement in
their organizations. Fears should be dealt with the right way. Before putting
performance management to use, management needs to inform employees
regarding the rationale and benefits of performance measurement. More
importantly, they should not use performance measurement as a tool to
punish their employees. The enemy of performance measurement is the
organization’s problems, not its employees.
#9 A performance measurement system must not create the “you do it”
culture.
Have you ever heard of a “you do it” organization? A “you do it” culture
implies that employees refuse to do their jobs. This usually is the cause of
failed performance measurement, because no one wants to be responsible for
anything. But who should take care of performance measurement?
“Accountability” is a simple principle; the person who is responsible takes
both the compliments and the blame. Thus, performance measures should be
controlled by an accountable person.
For example, who should be accountable for measuring defects? First, we
should ask ourselves who has the power to control and reduce defects. The
answer is the production department, right? Then the production department
should be accountable for this performance measure.
I have another suggestion about how to select the accountable persons. For
the best result, you should assign only one person or one department to be
accountable for a particular performance measure. The more people involved,
the less accountable a person involved will be. In other words, ‘we are
accountable’ actually means ‘no one is accountable’.
Have you ever noticed that none of the organizational performance measures
receives enough attention? This is because everyone assumes that there will
be a person who is already working on it. So go back and take a look at your
organization’s current performance measures, and ask this simple question:
“Who is accountable for this performance measure?”
#10 A performance measurement system must be correct.
What does it take for an organizational performance measurement system to
be considered as a failure? Many would respond, the system would be
considered a failure if the organization generates less revenue, incurs losses,
or even goes bankrupt. Well, it does not have to be that bad. We can say that
a performance measurement system has failed, when the system is not
accepted and used. This is already considered as a fail. Normally a failed
performance measurement system is caused by these main reasons:
1) The users do not understand the measure, or
2) There is a mistake in the system (e.g., measuring the wrong thing)
If number 1 is the case, then the solution is to train the users on what the
performance measure is, and how to use it. But if number 2 is the case, then
you need to adjust the system. For example, you can design a new
performance measure that is more concise and can better solve the problems.
Introducing a performance measurement into an organization is like teaching
a new language to everyone. We have to inform people as to why they need
to learn the language. Also, we need to hire a knowledgeable instructor of
that particular language in order for everyone to learn the new language the
right way.
The last question is, how can we know that the current performance
measurement system will not fail? Well, we can notice from the decrease in
inquiries or complaints. However, the best proof that can indicate the success
of this system would be seeing everyone benefit from performance
measurement in terms of work development.
#11 A performance measurement system must be “simple” and
“relevant.”
If you want organizational performance measurement to succeed, I have
some tips to give you: make it simple and relevant. Many organizations
invest a lot of money and manpower to design flawless performance
measures and excellent databases. However, no one pays much attention to
these measures. In the end, the performance measurement system fails.
The most common mistake is not making sure that the users understand and
feel that the performance measures put in place and the data they produce is
relevant. Otherwise, no matter how good the performance measure, or how
precise the data might be, no one will utilize it. “Simple” means that we have
to create a performance measure that is easy to understand. Avoid jargons,
and use familiar words that are already a part of the culture. Good
performance measures would not raise questions like “What is it?” In cases
where you cannot avoid jargons, you must explain the language to its
potential users. Do not assume that they would ask if they do not understand.
Most of the time, they will not ask, because they do not want to look less
knowledgable (why didn’t they know such a simple thing?), or maybe they
are just simply too lazy to ask.
Another tip is to be “relevant.” Everyone wants relevant information, right?
A simple example would be Facebook, because Facebook provides tailored
information. Your Facebook page is different from other people’s. And your
friend’s Facebook is also different from other friends’. Yet, everyone still
receives information from friends and acquaintances. This is what makes
Facebook world famous. Sending a report of performance measures to other
people is no different. First, you have to ask yourself if the report is relevant
to them. Marketing department would want information regarding sales, more
than the percentage of defects. If we send the percentage of defects to the
marketing department, it is very likely that they would not use the data.
I have a simple rule that you can use to test the effectiveness and efficiency
of your report. A good report should make readers understand the main
contents in 10 minutes. If it takes longer, it is very likely that they would not
utilize the data, which might indicate that the data is not “simple” or
“relevant” enough.
 Chapter 7
Design a performance measurement system
Time flies… You’ve already read six chapters. Now is the time for the
chapter you’ve been waiting for. How can we design a performance
measurement system? Let’s learn how many performance measurement
designs and processes there are, and you will realize that designing a
performance measurement system is not that hard at all.
Let’s learn how to design a performance measurement system.
There are three approaches to designing organizational performance
measurement system:
1) Top-down approach
2) Bottom-up approach
3) Integrated approach
The “top-down” approach is when executive management designs overall
organizational performance measures first, and then performance measures
for each department are created. The advantage of this method is that the
management is already well-aware of the organizational strategy, hence the
designed performance measures would be aligned with the organizational
strategy. In addition, the departmental performance measures would also be
in alignment with the organizational performance measures since they were
all designed by the same person (or the same group of people.) However, the
notable weakness of this approach is that the employees would be
dissatisfied, because they were not involved in the development process.
The second approach is the “bottom-up” approach. This method allows
employees from each department to participate in designing their own
performance measures. After that, the designed performance measures are put
together to create organizational performance measures. The advantage of
this approach is that the employees would believe in the performance
measures, because they were the designers. If they still have doubts in their
own measures, then I do not know what else can be done... On the other end,
the weakness of this method is that it is difficult to consolidate the measures
created by different departments, as they have different directions and not
relevant to the organizational strategy, most of the time.
The last one is the “combined” approach. It begins by having the executive
management design the organizational performance measures, then letting the
employees design their own departmental performance measures. However,
there needs to be a theme that guides departmental performance measures to
be aligned with organizational performance measures. The pros and cons of
this approach fall somewhere between the first and the second approaches’.
Now back to the important question, which approach is the best one? The
answer is, there is no best one. It all depends on the characteristics of each
organization. Take a look at the pros and cons, and you should be able to
select the best one for your organization. Still, the most popular approach is
the top-down approach as it is the easiest way to ensure that the measures fit
the organizational strategy. So, give it a try.
Design a system in four steps
Before I introduce the main steps of designing a system, let’s take a look at
the definition of “organizational performance measurement system” once
again. Organizational performance measurement systems consists of:
1) Performance measure
2) Supporting infrastructure, which is composed of data collection
systems and reporting systems
Therefore, creating a performance measurement system requires the creation
of both the performance measure and the supporting system. We can
conclude that there are four main steps to building a performance
measurement system:
1) Determine what to measure and create a performance measure.
2) Set a target value, and set the data collection system.
3) Design a reporting system.
4) Adjust the system before launch.
If all four steps are properly executed, you will have established your
organization’s performance measurement system. In the next chapter, I will
get into details of each step.
 Chapter 8
Create a performance measure
The first important step in designing a performance measurement system is
specifying what you should measure. This is the step that creates a lot of
misunderstandings. The most common one occurs when people brainstorm
on what they should measure, and end up with around 100 performance
measures, calling them all “KPIs.” In fact, the K in KPI is an abbreviation of
“Key,” which means important. Is it even possible to have more than 100
important things? Let’s see how we can select what we have to measure.
“Strategy” comes first.
Another significant factor in designing an organizational performance
measurement system is to understand the organizational strategy, because
the organizational performance measurement will become one of the tools
that transforms strategy into action. Therefore, if we do not know what the
strategy is, then we would not be able to design a performance measurement
system. If you do not know your goals and how to achieve them, of course,
you would have no idea how to measure your success rate. Here’s an
example... Let’s assume that an organization uses a strategy of
differentiation; the performance measures should be the number of new
products, the number of suggestions, and the R&D expenses. Whereas,
another organization that emphasizes a low cost strategy, then the
performance measures should monitor other things, such as the cost per unit
and the number of defects. Different strategies need different performance
measures. Therefore, understanding the strategy should be the top priority,
before designing an organizational performance measurement system.
Consider stakeholder needs.
Another thing that we should not ignore when creating an organizational
performance measurement system is to study the needs of stakeholders. Who
are the “stakeholders” of the organization? Many people would only think of
customers. Yes, customers are stakeholders, however, they are not the only
ones. Stakeholders of an organization include the owner, the management,
business alliances, employees, financial institutions, business partners, the
public sector, as well as the competitors and the society.
Many of you might wonder why competitors are considered as stakeholders.
Let’s take a closer look at the word “stakeholders.” It implies that they must
have gained from or lost something to the organization. What do the
competitors have to gain or lose? Get it? The customers! If they only take our
customers, then they would no longer be our competitors, because we would
be running out of business. Or, if we convert their customers, they are no
longer our competitors, because they would no longer be in the business.
What about the society? How could the society be a stakeholder? Think about
it. If we open an oil refinery in a community area, then of course, we would
need the community’s approval. Otherwise, conflicts can arise. In exchange,
we would need to improve the community’s standard of living.
As you can see, all stakeholders give to and take from our organization. Here
is what is important... To ensure the success of our organization, we need to
satisfy the stakeholders; and in turn, we must gain something back from
them. Therefore, a successful performance measurement must be created by
taking these people and their needs into consideration. We have to study the
needs of stakeholders by conducting a focus group or a survey, and figure out
how many levels of needs there are. We can divide those needs into three
levels: basic quality, performance quality, and excitement quality.
The basic quality concerns what people already expect to have, a must. So,
they usually do not express those needs. Despite the fact that they might not
have mentioned it, do not conclude that they do not want it. For example,
what do you expect the car you are about to purchase to be like in terms of
quality? Nobody would say that they want the car to run, because that is the
basic quality that every car must have.
The second quality is performance quality. This level of quality is clear-cut,
and can be measured. It is the common answers that people give. Let’s use
the same example again. What do you expect a car to be like? You might say
you want a fuel-efficient, powerful, and luxurious car. These are performance
qualities.
The highest level of quality is excitement quality. If we do not ask about this
quality specifically, we would not get the answer, because most people would
not even think of it. For example, what kind of system would you want your
car to have, that you cannot find in today’s cars? Your answer might be
autopilot, hydropower, etc. These things excite us, right? Studying the needs
of stakeholders can make us understand their needs and how to satisfy them,
so that we can design a more precise performance measure.
Create objectives and critical success factors.
Once we know the organizational strategy and have already studied the needs
of stakeholders, we must convert the strategy and stakeholder needs into the
organization’s main objectives. An organization’s main objectives are
answers to the question: “what objectives should be achieved for our
organization to be successful?” For example, objectives of an organization
could be:
1) Generating profit
2) Increasing customer satisfaction
3) Satisfying employees
Setting main objectives will give us a clear direction as to what we want to
achieve. After determining our organization’s main objectives, we should
define our “critical success factor” that will help us achieve those main
objectives. For example, if the main objective is to generate profit, then the
critical success factor would be to increase the number of branches, or to
generate more sales per branch. Simply put, critical success factor is the
cause, and the objective is the effect. However, you should create
organizational performance measures only for very important objectives,
because sub-objectives will emerge by themselves once we have transformed
the main objectives into departmental ones.
Designing performance measures
Let’s take a look at the main steps of designing organizational performance
measures. Mostly, when we talk about organizational performance
measurement, we would brainstorm what we should measure. Brainstorming
is a must, however, it is not the first step. Generally, the process would start
like this:
1) Define business processes, and assign the responsible person in
creating the performance measures for each business process.
2) Arrange a performance measure design training session for those
responsible individuals (in number 1) to set them off on the right
path regarding the proceses and performance measures.
 3) Brainstorm to identify the appropriate performance measures. Each
performance measure must be aligned with the main organizational
objectives, as well as the critical success factors.
4) Consolidate all performance measures to establish organizational
performance measures.
It does not sound that difficult, right? Creating performance measures might
take months, or even years if the organization is large and complex.
However, if it is an SME, then these processes can be completed within a few
weeks. So, try to adapt these steps to your organization.
Performance measures must be related to each other.
As mentioned earlier, good performance measures must be aligned with an
organization’s main objectives and critical success factors. But that is not
enough. What I want to add is that once you have designed performance
measures, you must ensure that they are related to each other. Let me draw a
clearer picture. Assume that one of the main objectives is to make employees
happy. In order to make employees happy, we might think that we should
give them proper benefits. Can you see the rationale behind this? When we
offer proper benefits (critical success factor), the employees would be happy
(main objective). There can be more than one critical success factor, for
example, anything apart from benefits).
Now, how can we create a performance measure that would indicate the
happiness of employees, which is the main objective? We might measure
employee happiness by an employee satisfaction score that we could gather
from surveys. But how can we measure proper benefit, which is the critical
success factor? We can measure the percentage of employee benefit
expenses. Once we have created performance measures, we can re-check
whether the two performance measures are related to each other or not. From
the above example, if the percentage of employee benefit expenses has
increased, the employee satisfaction score should increase as well. Generally
speaking, if it makes sense, that means the performance measures work.
However, if it does not make any sense at all, then we should consider
switching to a better performance measure.
Quantitative or qualitative performance measures
This matter will concern the types of measurement mentioned earlier. We can
divide performance measures into two categories: quantitative and
qualitative. Quantitative performance measures can be measured directly
without adhering to human judgment (e.g., production time). Whereas, most
qualitative performance measures cannot be measured directly, because they
usually involve intangible things that require human judgment. This kind of
subjective performance measures usually raise questions.
Let me give you a classic example. I believe that many of you might have
conducted research, and distributed questionnaires to measure customer
satisfaction. If you have not, I believe that you might have completed this
kind of survey as a customer. Let’s take a look at the scale first. The widely-
used scale is 1 to 5, where 1 means the “least” satisfied, and 5 means the
“most” satisfied. Then, the researchers would analyze the data, and draw a
conclusion on whether the customers are “rather satisfied” or “rather
dissatisfied.” No one would ask, “What is the difference between ‘least
satisfied’ and ‘somewhat dissatisfied’?” Or, “What is the difference between
‘most satisfied’ and ‘somewhat satisfied’?” Think about it... Let’s assume
that I am a very rigid person who usually gives low scores. My 2 might mean
that I am “very satisfied” (because I usually assign 1s). However, my friend
is very easygoing. He gives 5s to everyone, unless he is very upset, then he
might give 4s. When we conduct a survey, we simply average the scores
between these two people; even though someone’s 4 might have less value
than another’s 2.
As you can see, measuring intangible things is not easy at all. In order to
reduce bias in performance measurement due to different standards, we
should illustrate “an example of a behavior” instead of just saying least or
most. In surveys that require very precise results, researchers might specify a
behavior corresponding to each scale level. For example, for 5, the most
satisfied, the associated behavior might be, “You are so satisfied that you
have written a letter to the CEO of the organization to express your
satisfaction.” When you see this kind of message, you would know that even
though you are satisfied with the service, you are still not satisfied enough to
write the letter. Then, you will not give a score of 5. For 1, the least satisfied,
an example behavior might be suing the company. So, if you are upset with
the service but not upset enough to sue the company, then you should not
select 1.
This technique can help adjust the standard of each person, because they now
understand the meaning of each score. But of course, as long as we are
human, bias will continue to exist. We cannot completely eliminate bias, but
at least with this technique, we can get a more precise result. The limitation to
this method is time constraint, because surveys like this require details and
time. In order for the respondents to complete the survey, we might need to
give them incentives, which means the investment in data collection would
be higher. You have to consider the trade-off between the result and the
expenses, to see whether it is worth it or not.
Financial or non-financial performance measures
We can look at performance measures from another perspective, and
categorize them into two main categories: financial performance measures
and non-financial performance measures. You can tell these two forms apart
by looking at the unit of measurement. If the unit is monetary, then it is a
financial performance measure. Whereas, if the units are things like the
number of defects or the number of complaints, then the performance
measure is non-financial. Sometimes it could be both; for example, revenues
per employee, where the numerator is financial but the denominator is non-
financial.
You do not need to choose which kind of performance measure is better,
because both of them are used in reality. Naturally, in for-profit
organizations, non-financial performance measures are the leading indicators,
while financial performance measures are the lagging indicators. Since we
should measure both the leading and the lagging indicators, we need both
kinds of performance measures.
Result or process measures
Let’s assume that we want to create a performance measure for a salesperson.
Which performance measure do you think is more appropriate: the number of
sales by that salesperson or the activities of that salesperson that would help
generate more sales (e.g., the number of customer visits, the amount of time
spent in responding to inquiries or addressing complaints, etc.)? Which one is
better?
The problem with the first choice is that sales is difficult to predict, because it
is influenced by various factors. For example, the decrease in the number of
sales might be caused by poor economic conditions; it might not necessarily
imply that the salesperson lacks the necessary skills. And vice versa – an
increase in the number of sales might be due to improving economic
conditions, and might not necessarily be a result of good sales skills. If this is
the case, then why would we use a performance measure that the salesperson
cannot control to measure his skills?
Then, maybe the correct choice is the second one. Just you wait... The second
one measures the activities that the salesperson performs to generate more
sales. This is indeed something that the salesperson can control. However,
what an organization wants is sales, not the number of customer visits. If you
use this measurement, then the salesperson would visit 3-4 customers per
day. What should we do then?
Before I answer this question, let’s take a look at the difference between the
two performance measures. The first measure is an example of a result
performance measure. It emphasizes solely the results, and is very popular in
the West, especially in the US. This measure empowers employees to come
up with their own processes, and measures only the end-result. If the
salespersons say that sales decreased because of the bad economy, the answer
they will get is that the effect of economy will compensate each other when
the economy gets better. It is a random effect, which means it can be good at
one moment and bad the next. So, you do not have to pay attention to it,
because it will eventually cancel each other out.
The second kind of performance measure is called process performance
measure. It puts emphasis on the process, and is more popular in the East,
especially in Japan. The idea behind this performance measure is that one
should be measured based on the things under his control. That is what is
considered as fair. What if the salespersons follow each process, such as
visiting customers frequently, solving problems on time, etc., but sales still
decrease because of poor economic conditions? Should we reward the
salesperson for the decrease in sales? The answer is yes, because had the
salesperson not followed the process, the sales figures might have been even
worse. Therefore, the decrease in sales, in fact, helped the organization
because the salesperson worked his hardest, and followed each process
correctly.
In contrast, if the sales figures of another salesperson increases although he
did not follow the process – he did not visit the customers or correct any
complaints –, despite the increase in his sales, he does not deserve any
rewards, because the increase in sales is not a result of his performance, but
of other factors like a recovering economy.
Now let’s get back to the question: which one is better? That would depend
on the culture and the values of each organization. In reality, we do not have
to choose one; we can use both, because each performance measure has its
own pros and cons. If we use them together, we can get better results.
Tips for creating effective performance measures
There are a few techniques I would like to introduce. You should consider
these eight things if you aim to create good performance measures:
1) Start by studying the main objectives and the critical success factors
of your organization.
2) Try to look beyond financial performance measures.
3) Try to find cause and effect relationships between designed
performance measures.
4) Include both leading and lagging indicators.
5) Include both quantitative and qualitative performance measures.
6) Consider the possibility of acquiring data and the validity of the
created performance measures.
7) Try to acquire information from within and outside the organization.
8) Use simple language and avoid jargon.
Try to include these eight tips in your checklist when you are working on
creating performance measures. Or if you already have performance
measures, then re-check them considering these tips. If you lack one of them,
you can use it to improve your performance measures.
 Chapter 9
How to set targets and collect data
There is a saying that goes, “goals are meant to be smashed.” But before we
can smash them, we need goals first. Now let’s see how we can set goals to
create results. Once we have the goals, we will take a look at how we can
collect data for performance measures. Do not forget, “Garbage In, Garbage
Out.” If the input is garbage (meaning it is bad), the output will be garbage.
How to set target values
One of the most challenging parts of performance measurement for me is
setting target values. After a performance measure is designed, the target
value will be the indicator that tells us when we have succeeded. Without
target values, we would be walking aimlessly, without a final destination.
Thus, setting target values is a very important thing.
So let’s talk about the time frame in setting target values. Good target values
require time frames. Normally, we would divide target values into two levels
(or if you want more levels, it is totally up to you): short-term and long-term.
Here is another question without a definite answer: what is considered as
short-term or long-term? In fact, it all depends on the industry, and other
factors. In the past, when we talked about long-term, mostly it would mean
around five years. However, today’s long-term is becoming shorter and
shorter; most of the time, it would mean around three years. Some companies
even consider one year as long-term. All in all, it depends on how fast a
particular industry is changing. For now, I will consider three years as long-
term in examples. When setting target values, we have to think about where
we want to be in the next three years first.
Here is an example so you can get a clearer picture. Our defect ratio is 6,000
ppm (parts per million). How many defects should we have in the next three
years ? We can start by benchmarking with our superiors. Now, there might
be another question: will they give us any information about their practices?
Agreeing on information sharing would be a win-win situation. For example,
if they are willing to share certain practices with us, then we would share our
good practices with them, which is a possible case. Or sometimes we might
benchmark with the companies that are not our competitors. These could be
companies that are operating in other countries and have different target
customers than ours.
Once you receive the information, do not forget that most of the information
will be current. For example, the best of the industry has a current defect ratio
of 2,000 ppm. If we want to be number one in the industry, we have to be
aware of the fact that the industry leader will continue to improve itself as
well. However, it would be harder for them to improve than us, because they
are already the best. Let’s assume that in the next three years, their defect
ratio will be 500 ppm, so we set that value as our target, meaning that our
long-term target regarding defects is 500 ppm. This is how one can retrieve a
long-term target value. However, many people might raise the question,
“Who can I benchmark with if no one is willing to share information with
me?” or “I’m already the best in this, it would be hard for me to benchmark
with someone else” (You are very confident, aren’t you?). Do not panic. If
there is no one for you to compare yourself with, you can set up your own
targets. You can look at past data, and ask your employees to brainstorm on
how much further you can improve in the next three years (long-term target).
In the end, you will have your long-term target.
Now let’s talk about short-term targets. For example, what should the target
be for this year? Let’s say our long-term target is to have a 500 ppm defect
ratio, but our current ratio is 6,000 ppm. Let’s go with the most
straightforward method. If your current defect ratio is 6,000 ppm, and in the
next three years you want it to be 500 ppm, then the ratio needs to be
decreased by 5,500 ppm. Therefore, on (linear) average, we must reduce
5,500/3, or approximately 1,833 ppm, per year. Hence, after the first year, our
defect ratio should be 6,000 – 1,833 = 4,167 ppm, or 4,200 ppm if rounded
up.
However, for the ratio to reflect the reality, maybe we need to consider what
needs to be improved or developed first. Most of the time, the early stage of
improvement is easier than the later stages. If you ever tried to lose weight,
you would deeply understand what I am talking about. In the early stage, you
can lose 2-3 kilograms only by reducing your calory intake. But later on, your
weight just refuses to decrease, and you have to start exercising. Same goes
for performance measures. In the early stage, it would be easy to reduce the
defect ratio, because no matter what you lay your hands on, you can simply
improve it. But in later stages, it would get more and more difficult, because
the obvious problems have already been dealt with. So, in order to reflect
reality, our first year target should be higher than the targets of later years.
This way, we would be able to create short-term targets for each year that is
relevant to the long-term target.
Targets are meant to be adjusted (if necessary).
Should we adjust our target values? Many people are concerned that if they
allow target adjustment once, the target values might lose their reliability.
This is true if we adjust our targets without decent reasons, or if organizations
simply adjust their targets because they want to achieve a certain target value.
In this case, you do not need a target at all, or you can just set a target once
you know the result. In other words, shoot an arrow first, then put the
bullseye where the arrow has landed. You just do not need a target; sound
familiar? But if you do not allow any alterations at all, even if when the
predefined targets are impossible to achieve, then your employees would
have no motivation to work. Well, in such cases, you should allow target
adjustment.
Now the question is in which cases should we allow target alteration? Targets
should be adjusted when important incidents occur. First is when there is a
strategic change. An adjustment in strategy has a direct effect on the target
value of performance measures. For example, if we are making most of our
sales through salespersons, we might set our target value to be an X number
of sales per salesperson. However, if we add more distribution channels later
on (e.g., modern trade channels), that means our sales strategy will change.
At least, we need to take a look at our predefined target value for
salespersons to see whether it is still valid or not.
Next is when there is a change in customer trends. Customer preferences are
ever-changing. Sometimes, change happens very fast, but sometimes only
gradually. For example, in the past, we used cameras to take photos, but now
cameras are replaced with smartphones. Hence, we have to keep our target
values (especially long-term) in check as trends change, to see if they need
any adjustments.
The third incident is when there is an important event that greatly affects the
current situation, such as 9/11, the tsunami in Japan, or wars. These things
will affect our sales and other performance measures. We must review our
targets to see whether or not they are still valid in such cases. Furthermore, if
there is a request to adjust target values, we have to consider why we need
the adjustment, because if we alter our targets without any proper reasons, in
the end, those target values would lose their meaning,
There should be many levels of target values.
Let’s talk about the risks of having “All or Nothing” one-level target values.
Assume that I set a target of 1 million dollars for my salespersons, which will
be rewarded with a three-month bonus plus a trip to Europe. However, if they
cannot achieve it, they will get nothing at all. Is this target exciting and
attractive? Well, if today is the first of January, then maybe yes, I would be
able to motivate almost all of the employees, because everyone wants a three-
month bonus and a trip to Europe. But what if some salespersons have
already achieved one million dollars in sales by June, by the first half of the
year? What would happen to their motivation to work? Of course, their
motivation would decrease, because they have already achieved the target.
There is no need for them to generate more sales, because they’ve already
gotten all of the rewards. If you think about it, these employees who have lost
their motivation are likely to be the company’s top salespersons; otherwise
they would not be able to achieve the target within the first half of the year.
As time passes – July, August, September... – we would lose these
salespersons.
Let’s say it’s October now. How many salespersons do you think would still
be selling this company’s products? Should be less than the beginning of the
year. No, that is not enough... What? There is more? There will be another
group of salespersons who haven’t reached the goal but also have lost their
motivation. These people are the ones who are way behind the target value.
For example, by October, they only manage to generate 300,000 dollars in
sales. They know that if they can only make 300,000 dollar sales in 10
months, they would not be able to achieve one million within three months.
These people would give up, and the company will end up with very few
salespersons by the end of the year. The rest of them are all on vacation.
See how having a one-level target value could be risky? And if that is so,
how can we fix this problem? Let’s see what could happen if we have a
multi-level target value. Will it deal with the problem? And how? Using the
same example, instead of setting the target value to be “All or Nothing,” let’s
have a go with three levels. For example, if they can achieve 800,000 dollars
in sales, they will receive a one-month bonus; one million dollars gets them a
three-month bonus; and 1.5 million dollars gets them a five-month bonus plus
a trip to Europe. This way, if our top salespersons have already achieved the
one million dollars in sales within the first six months, they will not stop
making more sales because there is a higher target waiting for them. Or if it is
reaching the end of the year and some salespersons have only managed to
generate 700,000 dollars in sales, they would not stop, because despite not
being able to get the out-of-reach biggest reward, there are still other smaller
rewards waiting for them.
Multi-level target values motivate more effectively than the one-level target
values. But the thing is, achieving the highest level of target value decreases
motivation. That is correct, and that is why we have to set the highest level to
be challenging but achievable. This will help motivate everyone.
8 techniques for setting target values for performance measures
This is a checklist for those who are about to define target values for their
performance measures:
1) Target values can be adjusted if there is a change in strategy,
customer trends, and circumstances that affect performance
measures.
2) Multi-level target values can motivate employees better than one-
level target values.
3) The highest level of target values should be challenging but
achievable.
4) Target value and rewards should be linked; the harder the target, the
more attractive the reward.
5) Employees should be involved in setting target values, because they
are the ones who will achieve it. If they do not feel like they own the
target value, it would be hard to keep them motivated.
6) Past target values could be useful in setting new target values.
7) If there is no past data (maybe it is a new product), then customer
and expert opinions could be a good source of reference in setting a
new target value.
8) If competitor or industry average value is available, it could be used
as a reference in setting a new target value.
Try adopting these techniques, and they might prove to be very useful.
Garbage in, garbage out
After we have a target, it is time for data collection. Have you ever heard of
“Garbage In, Garbage Out?” Its meaning is very straightforward: bad input =
bad output. It implies that if you have incorrect input, even advanced analysis
and complex mathematics will not help make anything better, because the
result will still be useless. Same goes for performance measurement. If we
have carefully selected the best performance measures, but then we input
incorrect data, the results will be incorrect. It will not matter at all how
excellent our analysis is.
Therefore, information accuracy is one of the most significant factors in the
performance measurement process. First things first. We need to ensure that
the data is correct. Be careful when you ask your employees to punch in the
data, so you do not end up getting a blank sheet of paper. This can mean two
things: (1) Such data does not exist, and (2) The data exists but the employees
refuse to input. Most of the time, number 2 is the case. But how can we
encourage employees to punch in the correct data?
Let me tell you a story. In order to measure the percentage of defect,
production employees have to fill in the number of defects, because they are
the ones who produce defects and know best about it. However, it turns out
that by the end of the day, the company only receives blank papers. There are
two possible cases: (1) There is no defect today, or (2) There are defects like
every other day, but the employees did not fill in the forms. I would suggest
that number 2 is most likely the case, not number 1. If you do not believe me,
then wait for a couple of more days. If number 1 is the case, there should be
some defects on the next day because it is very unlikely that there will be no
defects at all. If this were the case, then there would be no point for us to
create the performance measure in the first place.
Firstly, we need to learn why the employees did not fill in the form. Maybe
they simply do not understand it, or maybe they just do not see the
importance of doing it. Actually, most of the time, it is because of fear. When
employees need to provide information, especially a negative one like
defects, they would naturally be afraid that once management receives the
data, it might be used against them. For example, the management might
reprimand them, reduce their salary, or even fire them. When employees are
afraid, they will not provide true data. This is very important, because if we
receive incorrect data, the performance measure would be useless, no matter
how well it is designed. What we should do is to take away their fear,
especially when what they fear is not even real. We have to inform the
employees that the number of defects is not being collected to serve as a
punishment. We do not collect the data to fire the employees, but rather to
stop defect problems. We have to make them understand that performance
measurement system is the tool that will help management eliminate
problems, not employees.
Data are valuable, do not throw them away.
Have you ever noticed that data collection is not an easy task? Some people
do not want to provide any information, because they are afraid that the
information might come back to hurt them. Some people want to cooperate,
but they do not understand what kind of data is needed. We can say that it
takes “blood and tears” for some organizations to retrieve all the data they
need.
Now let’s assume that we have successfully retrieved all the data. The next
question is, what should we do with it? This is also a vital step. The data that
we have acquired must be arranged and be accessible, because these data
have to go through data processing to be converted into various performance
measure values, which would then be used to create reports. Is data collection
software a must-have? The answer is that it depends on how complex your
organization is. If it is highly complex and is a large organization with a huge
amount of data, then data collection and management software should make
everything easier for you. But for SMEs with few data, MS Access or Excel
should be more than enough. More importantly, it is likely that the data that
is collected during the early stages might not be 100% complete. No matter
how well you prepare, there will still be misunderstandings, incorrect input,
etc. Do not panic. If something is wrong, then fix it. For example, you can
add more explanation to the form. If we just let it be, it could be harmful as
the creditability of the performance measurement system will go down with
it. And once your employees think the system is erroneous, do not expect
them to use the data from the system ever again.
Before getting your hands on data collection
When we plan to measure our organization’s operations, the first thing that
we need to consider is data collection. Many organizations start off by
designing a form for data collection. But hang on... Did you know that most
of the time, the process by which you ask your employees to fill in the data
becomes a barrier in data collection? Sometimes it even results in employees
objecting to the idea of performance measurement.
Why is that so? Because sometimes the form that we use is the same as many
other forms, so it turns out that the employees end up having to fill in the
same data over and over again. So I recommend that you take a look and see
if the data you need is already stored in your database or not. In other words,
talk to your IT department first. If your organization is using an ERP
(Enterprise Resource Planning) system, it is likely that you already have the
data in your database, but it is collected for other purposes. If this is the case,
then cool! You simply need to retrieve the data and analyze it to receive the
performance measurement result that you are looking for. However, do not
expect to have every single data in your database. That would be too much to
expect. Most qualitative performance measures (measurement of feelings and
emotions) are not stored, which brings us to yet another important point. If
you do not have the data, do not abandon the performance measure. The
reason why we come up with a performance measure is not because we
already have data on hand, but because that performance measure is
important. If you do not have the data, then you need to collect it. You can
conduct a survey, an interview, or anything. As long as it is necessary, you
must do it.
What should we do if data is not available?
Sometimes you might find yourself in a situation where you have already
built a performance measure, just to realize later on that you cannot collect
the data. What does “cannot collect the data” mean? You might find out that
some types of data require more investment than your budget allows, or you
might have the budget for it, but the value of the data itself is not worth the
investment. So what should we do? Many people would just give up because
they cannot collect the data. But hold on, do not give up so easily. If you do
not have the data or lack the access, you can try using “surrogate
performance measures.” Surrogate performance measures might not be as
effective as the normal performance measures, but it is better than nothing,
right?
Let me give you an example. Assume that we want to design a customer
satisfaction performance measure by conducting a survey to find the average
satisfaction score. However, we find that each survey will cost quite a lot of
money, and we cannot afford it. So, we should consider a way to measure
other kinds of satisfaction that could substitute for the customer satisfaction
score. We might find other performance measures like the repeated purchase
rate, which may be data we already have. It should be able to reflect customer
satisfaction more or less. This kind of performance measure is what we call
“surrogate performance measure.”
Oh, and one more thing, do not forget external data. Sometimes when we say
we could not collect the data or that data collection requires high investment,
these data might already be collected by external organizations, such as
research institutions, ministries, and other public institutions. Last but not
least, the phrase “unavailable data” is just another obstacle that we have to
overcome.
Source of data
Now let’s get to data collection. From where should we retrieve the data?
Normally, there are four main sources:
1. Electronically stored data
These data may be stored for other purposes - financial data, for instance, are
stored to be used in creating financial reports. Coming across these data is
like finding lost treasure because we do not have to invest anything at all. We
only need to retrieve the data, and can utilize it right away.
2. Data stored on paper
These data may be stored for other purposes as well. However, they might
not have been uploaded to a digital database. For example, we might have the
number of defects stored in monthly production reports. When we want to
use these data for our defect performance measure, we can retrieve them from
monthly reports without having to collect new data.
3. Data that need to be additionally retrieved from a digital platform
If our existing database does not have the data we want, then we have to
collect additional data. My suggestion is, if possible, you should convert the
data into digital format all at once, so that you can skip the punching in
process during which mistakes often arise. An example of this would be
using an employee card scanning system to collect data on absenteeism.
4. Data that need to be additionally retrieved from documents
This is the last source of information. If the first three sources are not
available, collecting data from documents can be your last option. The reason
why I want it to be the last option is because it is time consuming. Your
employees might get so bored of it that they begin to object to the whole
system. But if it really is necessary, then you have to do it – collecting
employee satisfaction scores through a survey, for instance.
It seems like there are two main data collection systems. One is digital
systems, which do not involve any paper documents, and the other is manual
systems, which primarily rely on paper documents. Digital systems are easy
and convenient, but require high investment, especially to build. Manual
systems do not require much investment, but are time consuming and
inconvenient. So organizations have to consider the pros and cons of each
system.
How frequently should data be collected?
This is a very common question. Should we collect data daily, weekly,
monthly, or annually for our performance measures? This question is about
the “frequency” of data collection, which normally depends on two factors:
the cost of data collection, and the changing rate of data.
Regarding the first factor, if each data collection incurs a high cost, then of
course, we will not be able to collect data on a regular basis. Otherwise, our
organization would go bankrupt. For example, if we have to invest one
million dollars each time we collect customer satisfaction data by conducting
a survey, we will not be able to collect the data daily and spend a million
dollars every day. That would be too much to afford, or if we can afford it, it
still would not be worth it.
Regarding the second factor, let’s assume that the cost incurred is not that
high. However, the changing rate of the data is quite low. In this case, we do
not need to collect the data frequently. Let’s say we use online questionnaires
to retrieve employee satisfaction scores. No matter how many times we
collect the data, the cost would not be that high. But still, we would not send
out daily online questionnaires to our employees. Why not? Because
employee satisfaction would not change much over the period of a day.
Hence, collecting data that is unlikely to change frequently would incur
unreasonable costs. Or if it does not incur a high cost, it would still attract
unnecessary annoyance.
On the other hand, if the data is prone to change but we rarely collect it, we
will not be able to see the trend. For example, you would not collect data on
stock prices weekly, right? Because stock prices change every day, or even
by the second. If you collect stock price data on a weekly basis, you will not
be able to see the change in stock prices during the week. Well, it is like
everyone says, “Too much or too little can both be bad.” It is better to be
“just about right.” Now, many of you might still wonder how it’s possible to
know what it means to be “just about right.” I have a suggestion to help you
decide the frequency of data collection for each performance measure.
There is a study that states that we should collect data six times during the
time of change of that performance measure to be able to see the trend
clearly. For example, if we want to see the change in customer satisfaction
over a period of six years, we should collect data annually. However, if we
want to see the change in customer satisfaction over a three-year time period
(because customers change frequently in this industry), then we should
collect data once every six months. And if we participate in a fast-paced
industry, we might want to analyze our customer satisfaction trend over a six-
month time period. This means that we should collect data monthly. I believe
this information should be beneficial for those who plan on collecting data.
So, give it a try.
When should data be collected?
Once we have decided how frequently we are going to collect data, we also
have to decide which day of the week, month, or year we are going to collect
the data. We need to collect accurate data for our performance measure,
which means that the time of data collection is also important. Mistakes
usually occur when selecting the time of data collection. If we choose to
collect data at inappropriate times, we might get incorrect data or data that
does not reflect reality, which means the performance measure will be
incorrect as well. And the rest is history.
Let me give you an example. Assume that we have already decided to collect
customer satisfaction data from customers who use the service of a mall once
a month. Then we ask our employees to distribute customer satisfaction
questionnaires at 11 am on Monday. So, the employees go to the mall to
distribute the questionnaire, and finish at noon. Now, it is time for us to
process customer satisfaction data, analyze it, and make recommendations for
future improvements. Have you noticed the potential mistake? If we only
collect the data on a weekday morning, despite having enough numbers in
our sample, we might only cover a specific group of customers. Plus, they
might not even represent the majority of the mall’s customers. If we collect
data just so we can get it over with, our data would be no different from
garbage, because we could barely use it or could not use it at all.
Who should be the data collector?
This is another common question, because, in many organizations, people try
to evade the responsibility as it means that they have to work more than
usual. Plus such task is very much hated by most people. However, the
answer to this question is not that difficult. Firstly, if the data are already
stored in the database, we would not need anyone to collect them. We simply
have to ask the IT department to retrieve the data from the system, and then
we can process the data right away. However, if such data does not exist and
needs to be collected, the simple solution is to have the ones whose work is
most relevant to that data collect it.
For example, if we have the number of defects designed as a performance
measure, who should be the ones to collect this data? We have to go back and
check to see which department is responsible for creating defects, which
would be none other than the production department. Hence, in this case, the
production department should be the one responsible for collecting the data.
Having the responsibility of data collection does not mean that a particular
department has to be responsible for the results alone. For example, if a
salesperson is responsible for collecting daily sales data, it does not mean that
he has to be held responsible for a decrease in sales. Sometimes the decrease
in sales figures might be a result of poor production quality, bad
advertisement, or many other things. In any case, someone should be
responsible for data collection. Otherwise, you might end up with
“performance measures with no data.”
Cautions of data collection
It takes a lot of effort to acquire data, so we should take good care of the data
collected. We should store it systematically. Even though we might have
already processed and reported the data, we should still not throw them away,
because we will never know when we have to utilize those data again for a
different purpose, or we might even create new performance measures that
require these past data.
The next caution of data collection is about data access restriction. Some data
are sensitive to an organization’s competitiveness such as cost of production.
This is the kind of data that needs access restriction, because the higher the
number of people involved, the higher the risk.
The last caution is computer viruses. I believe everyone has heard of or has
experienced a computer virus before. Our data storage system should be
secure. In addition, we should also back up the data in order to prevent data
losses.
 Chapter 10
Prepare a result report
In previous chapters, I discussed system design, target setting, and data
collection. In this chapter, it is time for presenting the collected data. So let’s
take a look at important factors that we should pay attention to.
Do not fall at the last hurdle.
It is a shame that many organizations with good performance measures and
data collection fail towards the end. “Failure” in this case means that no one
utilizes the data. You can even say that the data report has somewhat become
more of a homework that just needs to be handed in. After that, no one knows
nor cares about what happens to it. It is a pity because reporting is one of the
last processes. Designing a good performance measure and collecting
accurate data require a lot of effort. But if it turns out that we fail at reporting,
these processes would literally be worth nothing. Yes, it would be difficult to
create a perfect report during early days. However, 75% is good enough for
me. We can gradually adjust it toward perfection. Most of the time, bad
reports are not a result of failed computer systems or miscalculations, but of
poor understanding of performance measures or processes. So let’s see how
we can create an interesting report of performance measure results.
Prepare a “simple” report.
The first technique I would like to introduce for effective reports is to make
them as simple as possible. Everyone likes the word “simple” – report
creators like it, and so do users. Doing what everyone likes can lead to
success. Think about it, if it takes us weeks to prepare a report, we would not
want to do it, right? Plus, once the report is sent out to employees of each
department, it would take them weeks to finish reading, and understand
everything in the report. Do you think anyone would ever read such a report?
Therefore, the first rule of thumb is to make everything “simple.” We must
have a user-friendly report system. Let me give you an extreme example. If
we can create a report within five minutes, that would be excellent. This is
where a software or an IT system can be a great helper. It would take only a
few minutes for users to understand the report and put it to good use right
away. This is what is called “simplicity.” Easy plus easy equals effortless. If
this does not succeed, then I do not know what else can.
Prepare a “relevant” report.
Now let’s take a look at another technique that could make a report more
interesting. Use a rifle, not a shotgun to hit the target. What does that have
anything to do with result reports? If you love American films, you would be
familiar with how protagonists use a rifle gun. They would place the gun on
their shoulder, which is why this kind of gun is highly accurate and hardly
misses the target. The short gun is the opposite; it’s hit or miss (but most of
the time, miss). Preparing a report is like firing a gun. If you create a report
like you shoot a rifle, your report will hit the target better. A report that is
meant to be sent to the marketing department, should be designed specifically
for the marketing department with relevant marketing data. Likewise, a report
for the production department should be designed specifically for the
production department. This way, users will receive only relevant data.
However, creating a report like you fire a short gun, trying to shoot
everything in sight (e.g., you prepare one big report that contains the data of
every department, then make multiple copies and distribute to everyone in the
organization) is firing blind, expecting everyone to search for the relevant
data on their own. I would say you are expecting too much. Most people
would end up not reading anything in the report. This could be very
heartbreaking for the people who work on preparing the report.
Think of the readers
When preparing a report, you have to ask yourself about who is going to be
reading the report. It is a simple question that people rarely ask themselves. I
have experienced this kind of complaint quite frequently: no one reads result
reports anyway, what should we do? Before jumping right into adjusting the
report, adding more colors and contents, the first question that we should
answer is, “Who will be the readers of this report?” If we can answer this
question, we would know what to do next.
If the report will be sent to the CEO, how much detail do you think you have
to go into? If you have too much detail in the report, do you think the CEO
will read it? What about the report for the CFO? What kind of contents do
you think you should include? It has to be an executive overview instead of
details, right? If the report is meant for head of the department, then you can
add more details because he or she needs the details to solve problems. Or if
you really do not know what that person wants, the easiest way is to ask him
or her. Try this simple technique every time you are sitting in front of a
screen, preparing a report – think of the readers.
Abolish the one-size-fits-all system
The key to creating a good result report system is to decentralize. Do not
stick to the principle of “One Size Fits All,” because it is nearly impossible to
have a form of report that is suitable for every department and every incident.
Creating a standard form of report that does not consider what department the
performance measure is from and the different data needs of departments,
could make the report useless. Meaning, it would be hard to find relevant
data, which would make the report uninteresting.
The public sector often uses this system, because it is easy to create, as they
can simply design a single form for employees to fill in. Sometimes the
employees can only fill a couple of answers in, but have to read through 100
questions to do so. Most of the time, they would fill in “N/A” or “0.” To
make matters worse, those data (that were not needed to be filled in) are
created into tables and graphs. Therfore, the report only contains “N/A”
values. No matter how useful this report could have been, it would still be
dull because relevant data are all obscured by a pool of irrelevant data. So
let’s change how we do things. Let’s prepare a report that contains relevant
data for each particular department. This way, the report would be interesting.
We can even add more details, explanations, and analysis, too. I believe that
most people would be looking forward to read such a report and put it to
good use.
Good news or bad news?
Have you ever noticed that sometimes reports contain good news, but from
time to time they also contain points of failure? Let me give you an example.
Let’s assume that we have two possible ways to report a performance
measurement result: 99% of our products passed quality standards, or 1% of
our products failed quality standards. These two forms are called
conformance reports (99% good quality products) and nonconformance
reports (1% defects). Which one should we use? The answer is that it
depends on the objective of the report. Conformance reports are commonly
used to communicate with people outside the organization, and it portrays the
good side, which would attract customers. Another advantage of this kind of
report is that it sounds positive; it can motivate employees. Everyone loves
good news. However, if reporting is for internal use to make further
improvements, then nonconformance reports are used; 1% defects will be
reported. This kind of report promotes continuous improvement, because if
we say that we are 99% good, most people would perceive that the
performance is already excellent, so there is no need to develop further. So,
try to apply these two forms of reports in your organization.
Using indices in reports
An index is a value that is calculated by taking every performance measure
into account. Let me give you an example. We might have many quality
performance measures. However, to see an overall picture of our
organization’s quality, we need to calculate the index of all performance
measures by adjusting the scores using the same scale. It is not necessary to
give equal importance to each performance measure. We can assign a greater
weight to important performance measures than the less important ones.
Creating an index value can provide an overall direction on a particular
matter. If you cannot picture this, try thinking of the stock market index.
Stock market index was created to act as an indicator of overall market
performance by including every stock price in the market calculations.
However, what you need to be aware of when using the index is that the more
performance measure values included in the index, the lower the variation of
the index would be. So, do not misinterpret the index value. When an index
does not change, it does not always mean that every performance measure
does not alter. When an index is stable, it might be caused by a great increase
in certain performance measures, and a plummet in others. Same goes for the
stock market indices. Stability of the stock market index does not imply that
none of the stocks’ performance changed. It is possible that half of the stock
prices in the market have risen, while the other half have declined.
How to report “confidential” data
There is always a question on whether or not we have to report every single
performance measure to workers. The answer is “we should,” because
otherwise they would not be able to improve. Another top question is, what if
that performance measure is confidential. For example, should we report the
performance measures that can put us at a disadvantage if acquired by our
competitors, such as cost per unit?
Now this question is interesting because there is a risk of competitors
acquiring our data. How can we know if one of our employees has a
girlfriend who works at a competitor company? Or what if one of our
employees switches to a competitor company? Having them sign a contract
might still not be effective because we would never know when our
employees are going to share our confidential data. It would be a little bit too
much to ask them to swear on it. Yet, if we do not report results, performance
measures would be of no use. For example, if we do not report the cost, then
our employees would not know what the cost is, and consequently there
would be no improvement. So, what can we do?
The answer is that we need to report it. However, it is not necessary to report
the actual number. Am I saying that we should report wrong data? Not at all,
but there is a way out. Instead of reporting the actual data, we can report “the
rate of change” of that performance measure. For example, instead of
reporting the cost per unit, we can report the rate of change in cost per unit to
show the percentage of increase or decrease. Or we can utilize the score;
meaning, instead of reporting the cost per unit, we can simply provide the
score. The employees do not need to know how many dollars the score
represents. Rather, they only need to know whether or not the cost has
increased, which should be enough for them to work on improvements.
Another method is to report the index value. We could use a value from any
year as a base value, and set it to 100 - I just made up this number, you would
need to adapt it. If the cost increases by 10% next year, then the index would
be 110. With this method, employees would have enough data for further
improvement while we eliminate the risk of revealing confidential data to
people outside the organization.
How many forms of reporting are there?
When it comes to reporting organizational performance measurement, there
are many forms of reports we can utilize. Let’s start with the most low-tech
and probably the most common one: the paper-based report. This kind of
report does not require high investment, all you need is a computer and a
printer. However, the disadvantage of this method is that it is difficult to
make changes in the future. Let’s say the report’s user needs the data to be in
the form of a table rather than a graph. The user has to contact the report
maker, and the user would never know if he would be able to get in touch
with that person. Plus, we would never know when the report maker would
finish creating the report. Then, there would be one revised version after
another. The user might end up forgetting what data he has asked for in the
meantime.
The next form, paperless reports, do not need to be printed out or bound. The
most common format is a Word document or PDF, which would then be
distributed through email. Users can read the report anywhere as long as they
have a phone or laptop. There is no need to carry the report around, hence it
is more convenient and cost (paper) saving. However, this form of report can
still be difficult to edit.
Another from is a report created through a website. We only have to send a
link to a particular website, together with the login credentials. This kind of
report is getting more popular, because once a report is created, we can
upload it to a website right away. We do not need to type report contents or
send emails to each employee. More importantly, we can adjust the report
automatically, by checking the login history to see which division or
department the users are from. Once we know which department the users are
from, we can choose to display only the data that are relevant to them. In
addition, it would be easier to track download history and to see who never
logs into the system. If the latter is the case, then there is problem.
Another form of performance measurement report – the online report - is a
very good one, because it is both creator and user-friendly. With online
reports, we can simply provide login credentials to users. The users can log
on to the system to see relevant reports. A marketer can see marketing
performance measures, while a financier might see financial performance
measures. What makes it even better is that this form of report allows users to
adjust the reports themselves through filters. Users can select which data they
want displayed on the report, or create their own graphs. Users love this kind
of report, because they can get anything they want in less than 10 seconds.
The creator would be satisfied as well, since he or she does not have to create
new graphs. It is a win-win situation. However, there is one small
disadvantage to this kind of report. It requires investment in software. If you
ask me whether or not it is expensive, I would say it depends on the benefits
you receive from it. If your organization is big and has a lot of data, I would
say the software is worth it.
And here comes the most advanced kind of performance measurement report.
Like online reports, this form requires login credentials from users. When
users log in, they see only the relevant reports, and can adjust the reports
according to their needs. But what’s amazing about this kind of report is that
it offers real-time data – meaning that you can retrieve the data in real-time
and they will always be updated. Let’s assume that we are looking at sales
revenues. Data collection would happen all the time at points of sale. We can
see the graph move all the time, just like the stock market. How’s that? Isn’t
it exciting? I have once joked with my students that it would be pretty
exciting to have a real-time grade report. A real-time board would be set up at
every desk. Whenever the students got bored and paid attention to their phone
instead of listening to me, their grade would gradually reduce. Wouldn’t that
be great?
One company has adopted this system, and found something interesting.
They measured the performance of their call center by allowing the
customers who called in to assess them. Well, it does not sound that
interesting when I put it this way. But here’s the difference. An assessment
system like this is randomized. They did not ask every customer to evaluate
their performance, rather, they only asked some of them. Now a typical
organization would gather this information and analyze it at the end of the
week or month to assess employee performance. But not this company. This
company calculated customer satisfaction right away, and displayed the result
in real-time on their employees’ computer screens, as they responded to
customer questions. What’s even more exciting is that this company used this
performance measure to determine employee bonuses. Can you guess how
the company did that? They finalized their evaluations by the end of every
shift. After every shift, the employees would know whether or not they would
receive a bonus for that day. Customer satisfaction performance measure was
used as a criterion. If the employees could achieve the predefined goal, then
every employee working during that shift would receive a bonus. However, if
they could not achieve it, no one would get a bonus.
Oh, one more thing… The customer satisfaction score displayed on the
employees’ screens an average score that did not belong to anyone in
particular. And that score was used as a bonus criterion. So, how is that? Isn’t
it exciting? However, in the end, this system failed, and the company had to
discontinue it. Before we go any further, let’s think about why this system
might have failed? What could be the reasons behind it? Can you come up
with any? If not, I’ll give you a little more time...
Here’s the answer. Many people might guess that this kind of system would
stress out the employees. Well, that is part of the reason. If anyone gets
measured while they are working, they would stress out. But the main cause
of the failure is that the customers called to complain. But why is that? This
system should benefit the customers because the employees would have to
take good care of customers, right?
Here’s what happened. Many customers called to complain that they had
been experiencing something weird lately. For example, if they made a call
during a certain time, they would get bad responses. Let’s say if they called
around 4 pm, the employees would talk to them impolitely, and did not serve
them well. Sometimes they said that the service was absolutely terrible. But if
they called during other times, the service would be good. They had tried
calling several times, and they still experienced the same thing. What
happened?
First, let’s look back at the real-time measurement system. This system is
good because it can tell the average satisfaction score right away. But there is
a disadvantage to that… Once the employees know the customer satisfaction
score in real-time, they know whether or not they will get a bonus for that
day. So, when it is near the end of the shift - let’s say the shift ends at 4 pm –
each call center employee knows for certain whether they are receiving a
bonus or not, because the score is displayed in real time. If the customer
satisfaction score is far behind the goal, they would know that they would not
be able to make up for the loss in the next 5 minutes. You can guess what
would happen next… Employees do not have to work hard anymore, so it is
unfortunate for the customers who call during that time. This kind of situation
would happen when the customer satisfaction score is way above the goal
near the end of the shift as well. The employees know for certain that they
would get a bonus for today, so let’s just sit back and relax. This is the reason
why this system had failed. It is kind of unexpected, right?
Manual reports vs. electronic reports
So which one is better: paper-based reports, aka. manual systems, or digital
systems that require software? In order to answer this question, we have to
take a look at three factors:
1. The amount of investment you are willing to make
Manual systems barely need any investment at all except for a computer and
a printer. However, in practice they can prove to be difficult to use. Let’s say
the user needs to adjust the graph or request additional information. This
means that you would have to recreate the graph and table, print, and send
out the revised report. If users rarely make extra requests, then it would not
be very difficult. However, if there are a lot of requests, you might have to
hire a team to deal with it, which means higher cost. If you use a digital
system, of course, you would need to invest in software and system
installations; however, it would make everything much easier in the long run,
and users can work on the revisions themselves.

2. The size of your organization

When organizations are complex, digital systems could be of great help.
However, when it comes to SMEs with small numbers of employees, manual
systems should be more suitable; it is harder to make adjustments in larger
organizations than in smaller ones.
3. The characteristics of the report
If the report is a standard one, meaning it has a definite template, then manual
systems should be fine, because you know for certain that the form of the
report will never change (very common among government reports).
However, if you have to adjust the report according to the users’ needs, then
you would be better off with a digital system.
All in all, regardless of the report’s form, the most important thing is that the
readers have to be able to understand and utilize the report.
How to make reports user-friendly
I have five main suggestions that would help make people read reports:
1) People can digest no more than seven items at a time. Therefore, do
not include more than seven kinds of data in a single graph or table.
2) People usually look at the data at the center of the page first. Hence,
you should place important data in the center of the page.
3) Pictures are easier to understand than texts. Therefore, if possible,
you should utilize pictures.
4) Use simple diagrams rather than pretty ones that look complicated.
5) You can adjust your report presentation to be more interesting by
utilizing other forms of graphs, or using colors.
Try putting these five recommendations into action, and your report might
become more interesting.
Graph lines are also important
The graphs that we usually see contain a line that shows a performance result.
However, I want to introduce two more graph lines that will show
performance measure results, which would make the graphs more useful.
The first line is called the baseline. This line tells you the performance result
during early stages, before work process development took place. This line
would be like the startline of a running race. The other line is the benchmark
or target. This line is the finish line that we want to reach. Adding these two
lines to your graph will help you see the direction of that particular work
process outcome - whether or not you are on the right path. If the line starts
from the baseline and is moving toward the benchmark, then you can sit back
and relax. However, if this is not the case - for example, the performance is
still around the baseline, or is turning away from both the baseline and
benchmark - you have to fix that immediately.
Management cockpit
Nowadays, many companies have changed the layout of their meeting rooms.
They do not just have air condition, TV, or new paint; these ordinary meeting
rooms have become extraordinary. Instead of having cement walls or
beautiful paintings, there are KPI graphs on walls. This kind of room is called
the “management cockpit.” As its name suggests, the cockpit is the room for
the pilots to control an airplane. If you have visited a cockpit before, you
would see that a cockpit has various gauges – altimeter, fuel gauge, direction
radar, and other gauges that pilots have to use to fly safely to the destination.
Why do pilots need these gauges? The simple answer is, because flying an
airplane is very complicated. Even when we drive a car, we still need a
dashboard.
However, I believe we would still manage to drive without a dashboard. On
the other hand, flying an airplane without a dashboard would be very
difficult. If you got on an airplane and do not see it in the cockpit, better
leave the airplane immediately. Pilots are able to control airplanes thanks to
these gauges, especially when flying at night or at high altitudes. Without
these gauges, flying would be very hard. Imagine how much damage would
be caused if the gauges did not work. Same goes for organizations. The more
complex the organization is, the more necessary gauges are for the
management to manage and lead the organization to success.

Chapter 11
Do not use it if it fails a test
We should be done with designing a performance measurement system after
we have designed performance measures, collected data, and reported the
results. But for certitude, we should test the system before launch. Testing
would allow us to determine the cause of potential problems so that we can
fix them before starting to use the system. Let’s see what we need to check.
#1 Test the relevancy to strategy.
As I have mentioned earlier, a good performance measurement system
reflects the organization’s strategy. But how can we know whether our
performance measurement system is aligned with the strategy or not? This is
a very interesting question. You can simply put it on trial. If the predefined
strategy gets translated into action, which results in a certain outcome, that
means the system works. If not, then the system does not work. This answer
is not a simple answer, but rather a convenient one. Well, it’s like we build a
car and when a person asks us, “Is this car safe?”, and we simply reply, “Go
on a test drive. If it does not explode, then it’s safe.” This is not a nice reply
at all.
After we have finished designing an organizational performance
measurement system, we should test it. The first test is to see whether the
system is aligned with the organization’s strategy or not. The best way to test
is to take the system to the executive management team, and discuss with
them. They are the ones who develop the strategy, hence they would know
whether the performance measures match their strategy or not. Most of the
time, designing a performance measurement system takes a long time, and
there are many adjustments along the way. Or sometimes there are changes in
the organization’s strategy. Testing ensures that what we measure covers the
key points in the strategy; no more, no less.
#2 Test the validity of the performance measure.
Having tested strategic relevance, the next step is to test the validity of
performance measures. How can we know whether our performance
measures give us the correct values or not? I think this is a very good
question as well. You can try the following methods to check if your
performance measures give the correct results or not.
1) Take a look at important business processes: production, customer
management, etc.; and see if there is any performance measure that
is controlling these procedures. If not, the number of existing
performance measures is not enough.
2) Ask yourself how effectively the existing performance measures
demonstrate the changes in those procedures. Assume a performance
measure determines the % of defects – let’s say if the number of
defects increases, the value of this performance measure would rise
immediately, which shows that the performance measure works.
However, if your performance measures can only show whether
there are defects or not, meaning that you would not be able to tell if
the number of defects have increased, then that performance
measure does not work, because it can only tell that “there are”
defects.
3) Try checking the correctness of the existing performance measures
by collecting data and calculating the results manually. Do not trust
the IT system or software, before you get to compare the results with
your manually calculated number first.
4) Try plugging in made-up numbers to test the system once again. For
example, if you know that the % of defects is calculated by dividing
number of defects by number of produced goods, then key in 10
 pieces for defects and 100 pieces for produced goods. The accurate
system should return 10%.
5) Try adjusting the numbers from step 4. For example, instead of
typing in 10 pieces, try keying in 50 pieces (with 100 pieces of
produced goods), and see whether the number changes from 10% to
50% or not. You can do this multiple times to validate.
Try adapting these five suggestions to your organization.
#3 Test data accuracy
Another significant component in creating organizational performance
measurement systems is data accuracy. Before using the system, you have to
ensure that your data is accurate. But how can you be sure of that? You have
to test it first. The easiest way to test is to enter all data into the system, and
examine the result of each performance measure. Strange results might
indicate that there is something wrong with the data. Beware of performance
measure results that show no data; most of the time, this is caused by a
system error, not because there is no data to process.
#4 Test the correctness of reports.
How can we know that our performance measure reporting system is
working? The simplest way to do this is to create a report, and take it to its
users. This might take us some time in the case of paper reports for example,
because we would have to send the report out, and wait for users to read it
before we can ask for their opinions. Or if the reports are digital, we have to
allow the users some time to familiarize themselves with the system, before
asking them for feedback.
I recommend creating an open-ended simple questionnaire that allows users
to express their opinions – what should be improved, for instance – so that we
can use these recommendations to improve our report. It does not sound that
difficult at all, right? If you already have this kind of report in place, have
you ever asked the users for their opinions to see if they like the report or
not? Do not forget – a report without readers is nothing but scrap paper.
#5 Test to see if the system will be put to use.
When we build a performance measurement system, we do not only expect
precise performance measures, usable data, and an interesting report of
results; we also hope that the system will be used in a beneficial manner.
Hence, one of the things we should test, is whether the users have utilized the
performance measure system or not. But how can we test that? A simple
answer to this question is to let the users test the system. If it is a digital one,
then give them usernames and passwords. After a week, you can ask them if
they got to use the system. Or if it is a manual system, ask them how they
utilize the data in their job. If we find out that they get to use the data, then
we can rest easy. But if not, we might need to find “the reasons why,”
otherwise the system would become useless.
#6 Test to see if the system has any other uses.
This is the last topic about testing. Do not stop asking questions, even if you
have received positive user feedback. Do they have any other suggestions?
Or even if the users say the system is useless – do not give up, because
what’s useless to one person might be useful to another. Sometimes a system
designer cannot see how the system can be adapted to other kinds of usage.
So try asking this question, and you might find that the system that you have
is more useful than you thought.
Three questions that need to be answered after testing
Once we finish testing, we might receive suggestions, criticisms, or even
complaints on why we did not do this and that. I think this is very normal. It
is better than not having anyone paying any attention at all. In fact, if you
look on the bright side, those criticisms actually stem from good will. If one
does not care, why would he/she give any suggestions at all, right? But what
should we do? Should we follow what they said, or stand by what we think is
right? If we choose to follow what they said, how much should we
compromise? If we stand by what we think is right, will it seem like we listen
to nobody?
The simple answer to this is that we should take those suggestions into
consideration. If the answers to the following three questions are “yes,” then
you should follow the suggestions. However, if one of the answers is “no,”
then you should not follow the suggestions, and go back and explain your
reasons to the users as to why you did not follow their suggestions. The three
questions are:
1) Does the suggestion have a solid rationale?
2) Is the suggestion practically possible?
3) Do we have enough resources, and is it worth the change?
The first question is “Does the suggestion have solid a rationale?” Well, if
you ask this question, everyone who has given the suggestion would say, “Of
course,” right? But the point of this question is to understand whether or not
there are “enough” reasons to change the current performance measurement
system or not. For example, if there is a suggestion that you should not use a
certain performance measure simply because you have never used it before, it
should not stop you from using it.
The second question: “Is the suggestion practically possible?” Of course,
many people would say that it is possible. But sometimes they think their
suggestions are possible due to a lack of information. A suggestion on
comparing our cost with the competitor’s sounds reasonable. However, it is
nearly impossible for us to know the competitor’s cost, so there is still a
problem with this suggestion.
The last question: “Do we have enough resources and is it worth the
change?” Sometimes suggestions are rational and are possible, but there is a
lack in budget or other resources. For example, conducting a survey on
customer satisfaction instead of just measuring the repurchase rate alone
sounds like a rational and possible suggestion; however, there is a budget
constraint. Hence, we could not conduct the survey.
If you find yourself stuck on any of the three questions, please explain to the
one who gave suggestions that you have already carefully considered the
suggestion, so that he or she could understand the reasons of rejection.
However, if your answers are all “yes,” meaning that the suggestion is
rational, possible, is good value for money, and demands an appropriate
amount of resources, then you should make the adjustment. Do not let your
ego seize the development of your performance measurement system. Do not
forget, Ego = 1/Knowledge.
 Chapter 12
Implement a performance measurement system
Having created a performance measurement system, it is now time to
implement it. So let’s see what the key points of this process are.
Designing is hard, but using is harder.
Anyone who has been involved with organizational performance
measurement should agree with the statement above. A performance
measurement system that has not been put to use is nothing but a stack of
blank papers. The first question is, how will the system be put to use? In
general, there are two methods to choose from: apply to the whole
organization at the same time, or apply to one division after another. If your
organization is small, the first choice should be fine. However, if your
organization is large, I would recommend starting with one division first, aka.
a pilot program. An advantage of launching a pilot program is that you can
learn about the system first, because you have no experience in it. You can
limit the damage it could cause, should the system have any errors. You can
utilize what you have learned from the pilot program to improve the system,
and apply the revised version to other divisions of the organization.
Another benefit is that if you do not have a sufficient amount of human
resources, a pilot program would allow you to delegate your human resources
and time efficiently in order to deal with potential problems. Imagine starting
to apply the system to the whole organization at once. Employees would ask
so many questions that we might not be able to answer or deal with them on
time, which might then lead to failure. Lastly, focusing on only one division
could increase the success rate. And once the system has proven to be
successful, it can be set as an example for other divisions, which would make
organization-wide system adoption easier.
Where should we start implementing the system?
Having discussed pilot programs for performance measurement systems, let’s
now talk about where we should start putting the system to use. This is
another interesting question. Normally, we would start where problems occur
the most. Why is that? Starting where the problems occur most often, would
make it easier to identify problems and see the results clearly, after the
problems have been dealt with. Similar to building a software to remove
computer viruses. If we start off by scanning computers that have no viruses,
the results would show that there is no virus. However, users would doubt
whether there really is no virus or the software simply does not work. But if
we scan a computer with many viruses, the software would be able to detect
the viruses so that we can deal with them. Wouldn’t that look more
impressive?
Same goes for performance measurement. We should identify the department
with most problems first, in order for our performance measurement system
to detect problems and recommend solutions. Once the problems have been
solved, other departments would be willing to try out the system. Another
thing is that if we start with a department with many problems, there is more
for us to gain than to lose. If there are already many problems, solving them
should make things better – or simply put, it would be difficult for us to make
things any worse. However, starting with an excellent department would be
more risky, because we would have to make what’s already excellent better -
that can be very difficult. And if things turn out good, employees might still
doubt whether the performance measurement system really helped make
things better or not, because the department was already great before the
system.
The last question is, which department has the most problems? Sometimes
you can answer this question easily by experience. However, if you do not
know the answer, you would have to take a look at the work processes. The
departments with many problems are usually the ones in the latter part of
processes. It does not necessarily mean that those departments are the ones
who create problems – sometimes the problems are accumulated from
previous processes. In the next section, I’ll show you what kinds of concerns
you have to address before putting a performance measurement system to
use.
Point of concern #1: Why should we have a performance measurement
system?
“Why should we have a performance measurement system?” – this question
frequently arises when we implement a performance measurement system in
an organization. There are two answers to this question:
 1) Performance measurement tells you the current status of your
organization, whether it is good or bad.
2) Once you know the result, you would know which area you need to
improve.
You lose almost nothing by measuring performance. Although some people
might say that performance measurement is a waste of time, money, and
labor, considering what you could gain from it, most would agree that it is
worth it. Organizational performance measurement is like getting a health
checkup. We have to pay money to get our health checked, but in return, we
get to know how we should adjust our daily routine. I think that is really
worth the money.
Point of concern #2: How can we use a performance measurement
system?
If an organization’s employees have no previous experience with a
performance measurement system but one day their organization makes an
announcement, “We are going to use KPI from now on,” I believe many
employees would be confused and nervous. They would be concerned about
how the KPI is going to affect them. In fact, organizational KPI application is
not as frightening as you might think. Instead, what you should be more
worried about is using it the wrong way. Utilizing a performance
measurement system would affect employees in the following procedures:
1) Employees play a part in the necessary data collection.
2) Collected data are transformed into performance measures.
3) The results will be analyzed to identify problems or opportunities in
order to improve work processes.
4) Employees develop their work processes, and look forward to seeing
the future results of performance measures.
And that actually covers everything. If you did great on everything, your
organization would only flourish. And when your organization flourishes,
your employees would receive rewards, such as bonuses or salary raises. A
win-win situation.
Point of concern #3: Would we get punished if a performance
measurement system is in place?
Before answering this question, I would like to ask you if you’ve ever gotten
your health checked before. Why did you get your health checked? Didn’t
you have to pay for it? Many would simply answer, they did it to know their
health condition, so it is worth the money. If they know that they are healthy,
they would feel at ease and realize that they are living their lives correctly.
And no one would complain, “What a pity. I want to be sick because I’ve
already paid for the checkup.” Also, if there is something wrong with your
health, then you would be able to deal with it right away. For example, if you
know that you have high cholesterol (LDL), you could get on with a healthy
diet and exercise to recover your health, right? But what does this have to do
with organizational performance measurement?
I’ve always compared organizational performance measurement to health
checkups. We measure to improve, not to harm people in our organization.
I always emphasize this point. We have to make everyone understand that
performance measurement is a tool to help management and employees work
together to solve problems. It is not a tool for management to fire employees
- this mindset is very important. If employees’ mindsets suggest otherwise,
then performance measurement will fail.
Point of concern #4: Why does it have to be me?
“Why do you have to measure ‘my’ work results?” “Don’t you trust me
anymore?” These questions are frequently asked when a performance
measurement system is about to be applied in the organization. First things
first. Have you noticed the word “me” in the question? Once there is “me,”
there is “them,” right? Questions stemming from this kind of a mindset marks
the beginning of a conflict. Hence, it is vital to make everyone understand the
rationale before starting to use a performance measurement system. There
should be no “me” or “them.” Performance measurement measures the
“organization.” The main objective is to utilize the results for organizational
development, not to point fingers at and punish someone. Therefore, this
whole thing is not about trust at all. Rather, performance measurement will
point out problems for everyone to see and fix.
Point of concern #5: Why do we have to measure when everything is
already good as is?
“Our organization has been established for 10 years. We didn’t need any
performance measurement. We didn't have any hardship. The profit is good.
Why do we have to measure performance now?” This is probably another
question that employees might ask themselves, or even say it aloud once they
know there will be performance measurement, because they do not feel
comfortable with it. In fact, I think this is a very good question, because it is
an opportunity to register a new mindset in your organization. This is how I
would like to answer this question. The fact that your organization has
survived, and remained profitable for 10 years without performance
measurement is a good thing. However, not facing any hardship without
performance measurement for 10 years does not necessarily mean that you
can survive another 10 years without hardship, right?
Let’s say we meet a 40-year old person who has been recommended to get
his health checked, and he says, “I’ve already lived for 40 years without any
health checkups, and I’m doing just fine. Why do I need a health checkup
now? Why do I have to waste my money?” Not having had any health
checkups for 40 years, and not having any illnesses does not guarantee that
the person will be able to live another 40 years in good health, right? This is
the exact same thing – if we never check our organization’s health through
performance measurement yet our organization remains profitable, it does not
guarantee the organization’s survival in the future. In contrast, it could be
very risky not to measure, since we might have accumulated many problems.
Despite the fact that we are still profitable, it might mean that the problems
are still hidden, the economy is good, the problems are tolerable, or
customers are still fine with waiting for late delivery to arrive. However,
when the economy turns sour, our organization might be the first to fall. If we
were to start using performance measurement at that time, it would be too
late. It is like giving a gravely ill person a health checkup – it is just too late.
Sometimes we might have heard, “We manage to generate profit without
performance measurement. Why do we have to change now?” The above
statement sounds reasonable – if something is already good, why change it?
But the important question is, how do you know that “it’s already good”? If
you do not measure, then you have to use your own judgment. If you use
your own judgment, you are putting your organization in a risky position.
Some people might say, “Just look at the annual profit over the years! How
can you say it’s not good? Yes, making profit is good, but the profit could be
a by-product of positive external factors, such as a good economy or an
expanding market that allows you to sell anything. It could turn out that these
good things are hiding problems. We might have too much inventory,
frequently broken machines, or lazy employees, but they might not have
affected our organization so far because we might have been able to generate
profit regardless.
These problems are like rocks under the water. If the water level is high,
boats would get through very easily. However, if the water level is low, the
rocks would surface, making it difficult for boats to float. Same goes for
organizations. Without performance measurement, we would not know if
there is a problem or not. If we wait until the positive external factors
disappear – poor economy and shrinking market – before identifying the
source of the accumulated problems, it might be too late. Even if you might
believe that your kind of industry would never go bad, still, with performance
measurement, you might be able to generate a much higher profit than
without performance measurement, because performance measurement tells
you what the problems are, so that you can deal with them. Wouldn’t it be
better if you could get rid of the rocks under the water before they could
create any damage to the organization? Performance measurement never
harms anyone - only poor performance measurements do.
Point of concern #6: Why do they have to start with our department?
Will we get to know the result?
There are two other common questions. “Why do they have to start with our
department?” Most of the time, this question arises when we choose a certain
department for a pilot program. The employees are curious as to why it has to
be them. This question can be answered easily. You have to explain to your
employees that, in the end, performance measurement would be applied to
the whole organization, but you choose to start with this department first,
because you can see the success opportunity in this department and this
department can be a leading example for other departments in the
organization.
The other frequently asked question is, “Will we get to know the results of
the performance measures?” It is even easier to answer this question – “Of
course.” If the people who get measured do not get to know the results, then
why are we measuring at all? After all, the people who can improve the work
processes are these same people. If these people do not know the results, then
performance measurement will not create any benefits.
When should we start using it?
Once the system is ready, there will be a question like “Eh, when should we
officially start using it?” Monday, Tuesday, Wednesday, Thursday, Friday, or
maybe Saturday or Sunday? I would like to skip the part about the auspicious
day to launch a system in this book. In fact, prior to saying which day we
should start using the system, we should define what “start using” really
means first. When we apply a performance measurement system in an
organization, in general, we would refer to the “first day of data collection.”
In fact, you can start on any day. But if possible, it should match the cycle of
that particular performance measure’s result report – that would be ideal.
Let me give you an example. If a defect performance measure is collected
monthly, then the first day of data collection should be the first day of the
month. Once we reach the end of the month, we would get a report for the
whole month. However, if we start on the 15th, we would only get results for
half of the month. In addition, we would have to spend more time to explain
why we only have a half-month report for this month, and that for next month
it is going to be a full-month report. This would add to the confusion,
especially during the early days of launching a new system. We would make
things more difficult than they have to be. Another thing we should do is to
communicate the launch date of the system. If most of the employees do not
know anything about the system, it would only make things more
complicated.
Training, training, training
Training should be available during the early stages of performance
measurement implementation. You cannot simply arrange a training session
just to get it over with, as it requires good preparation. The level of success in
training depends on the level of effort in preparation. The more you prepare,
the more likely the training is to yield the best result. In the early stages of
performance measurement system adoption, there will be complications and
questions. The best way to prevent further confusion is to train people.
Training sessions can be arranged using different methods – lectures and
system trials, for instance. Normally, the training instructor is on the system
design team, because he or she knows the system the best. However, if the
system designer does not have good presentation or teaching skills, I would
like to recommend that you hire someone with good basic teaching skills,
and inform him or her with the details of the system. Also, the instructor
should be a specialist in this field. These would ensure the success of the
performance measurement system training.
Why is it so complicated? Why are there so many questions?
During the early days of performance measurement system application, there
will be questions and complaints about why the system is complicated and
why they have a lot of questions. I always reply, “It’s good that they have
questions,” because if no one asks questions and keeps quiet, it is usually a
bad sign showing that no one cares about the system. Therefore, it is good to
receive many questions, because it means that people are actually paying
attention. Furthermore, these questions will help you identify which areas
users are still unclear about, so that you can clarify it for them. If you want
users to have questions, you have to provide contact channels for them. My
suggestion is to provide as many contact channels as possible - phone, email,
chat, or other channels - because the more channels available, the more
convenient it is for employees to ask questions.
If there is a question, there is an answer. During the early stages, it might be
necessary to hold a daily meeting for the team. For example, you can have a
discussion session for asking and answering questions. Or if there are any
suggestions, the team could decide together whether to adjust the
measurement or not. Although you might not end up following the
suggestions in the end, you still have to reply to the people who asked the
questions or gave suggestions to let them know the reasons behind your
decision. Or if you do follow the suggestions, in addition to responding to
them, you have to let every user know that there will be a change. This
method should guarantee success than any other method available.
How can we know if the system succeeds or not?
How can we know if our performance measurement system “succeeds”? We
would know if our employees utilize the performance measurement results in
their jobs. In order for that to happen, performance measurement results have
to reflect the true performance outcome. We have to invest money and labor
for all these to come true. There should be a training session to establish
mutual understanding, answer questions, or even fix problems. The more we
do, the more we gain; the less we do, the less we gain. Well, what if it does
not succeed? We have to take a step back and reevaluate everything to see
what’s wrong. Typical system failures are due to the following four reasons:
1) Management never really pays attention to the system.
This is one of the top reasons, and is the hardest one to solve. Many
managers use performance measurement only because other companies
are using it. They lose interest in performance measurement after a
while, and start seeking other kinds of tools. No system would survive
with a mindset like this.
2) Employees do not understand the information that they receive.
Many organizations have good and beneficial performance measures.
However, due to ineffective communication, employees cannot
comprehend the performance measures. What are they reflecting? For
what can employees use them? In the end, employees do not utilize the
data, which is a shame.
3) Employees receive irrelevant data.
Sometimes they understand the result of performance measures, but they
do not use it because it is not relevant to their jobs. You have to beware
of this. Do not send them the data just to get it over with – you have to
ensure that the data is relevant to their jobs.
4) Employees receive incorrect data.
If the performance measure is good and relevant, but the data is always
incorrect, how can employees utilize it?
These are the four common problems. If you find out that your employees do
not use the performance measurement system, you should go back and take a
look at the reasons why, so that you can fix the problems accordingly.
Signs of acceptance
After implementation, would you like to know how you can tell if the system
is widely accepted or not? The following incidents demonstrate that your
employees have already adopted the performance measurement system into
their jobs:
1) How-to questions about the performance measurement system have
declined.
2) Criticisms regarding the performance measurement have ceased.
3) Employees have requested adjustments to the report format, such as
changing the graph format or asking for additional data.
4) When the report is late, users ask when they will receive it.
Getting everyone to utilize the performance measurement system in their jobs
does not necessarily mean that this system will always be perfect. As a
person who is responsible for organizational performance measurement, you
always have to ask yourself, “Is the data users receive relevant to their jobs?”
and “Is this system still user-friendly?” These questions should allow you to
adjust the system to align with the tasks, and make it sustainably beneficial.
 Chapter 13
Put the data from performance measure to good use
There are many organizations that design good performance measurement
systems that is very well-received by their employees. However, it is a shame
that these organizations do not receive much benefit from the performance
measurement system. In fact, it is not completely useless, because they have
analyzed why certain performance measures have increased, while others
decreased. Still, the benefits that they receive is less than what they should
be. This is like buying a convertible car but only driving it to the market and
back. That’s a shame. Now let’s see what benefits you can receive from
analyzing and evaluating a performance measurement system.
Do not rush to analyzing the result.
When we receive the data from a performance measure, the first thing we
normally do is to check if the value has increased or decreased, and why.
Well, that does sound like the thing we should do, right? No! Because that is
jumping to the explanation part, you have to consider the fact that changes in
the value of performance measure are common. Think about the probability,
it is more likely that the value changes than remain static (because there is
only one possibility of the value being stable.)
Do you want to bet with me on whether the stock market index value will be
different on Monday and Friday of this coming week? You probably
wouldn’t bet at all. Because everyone knows that it is nearly impossible for
any stock market index value to remain stable throughout the week. What I
am trying to say is that the value of a performance measure either increases or
decreases. There is a possibility that it might remain stable, but that could
hardly ever happen. What does this have to do with anything? Before you
start analyzing why the performance measure has changed, you should
consider the following characteristics:
1) The change is usual. There is nothing special.
For example, a daily increase or decrease of 5 points in the value of the stock
market index is nothing special, because it would be difficult for the stock
market index value to be totally stable. If there is such a change, do not waste
your time explaining why. Sometimes I notice that some analysts trying to
explain every time there is a change in the stock market index value. For
example, they would say that the stock market index has decreased by 5
points because of poor economic conditions. And the next day, when the
stock market index increases by 5 points, they just say that it indicates a
change in the direction of the economy - there are only good news. Well,
does the economy change every day?
2) There is a change due to a special incident.
This change is abnormal. For example, the stock market index has increased
or decreased by more than 20%. This is this kind of change that you should
take the time to identify the source of, because you normally do not observe
the stock market values increase or decrease by 20% every day.
Try following this suggestion when you have to analyze your organizational
performance measures, and you will be able to manage time and labor
efficiently by choosing to analyze only important matters.
Do not panic because of what you just saw.
Oh my god, our sales declined 30% this quarter! Yes, this might be bad if...
There’s an “if”? The sales have declined by 30%, how can it be good? Of
course, it can. Performance measures will show you things of concern if they
are “abnormal.” Let’s say sales had always increased by 10% every quarter,
but this quarter, sales declined by 30%. Something must have happened. As
the person to analyze the performance measure, you have to identify the
cause. What if it is normal? For example, you are selling a New Year’s card,
and your sales at the beginning of the year have dropped dramatically when
compared to the end of last year. Then you do not have to worry about it,
even if the sales dropped by 30%. If you take a look at past data, and
compare them to the same quarter of previous year, you would see that it is
normal.
Therefore, before you jump right into analyzing the results, I want you to see
if the change is “abnormal” or not. If it is “abnormal,” then start analyzing in
detail. If it is normal, then do not panic, and do not spend time in analyzing
the result - it would be a waste of time.
Correlation is also important.
The following conversation happens in a department that works in analyzing
organizational performance.
Employee 1: "We’re doing good this quarter. Customer satisfaction has
improved."
Employee 2: "Yes, yes. But actually other performance measures have
improved as well."
Employee 1: "True. Employee satisfaction has also improved."
Once this conversation is over, the two employees go back to work. One of
them starts to write an analysis on the increase in customer satisfaction, while
the other writes about the improvement in employee satisfaction. In fact,
there is nothing wrong with that. Each person is doing their job. But what’s
missing is the correlation between the two performance measures if we
analyze them separately.
We do not need to use any advanced statistics. We only have to plot the
customer satisfaction score on the same graph as the employee satisfaction
score. Sometimes, we will get to see a certain pattern. For example, every
time employee satisfaction increases, customer satisfaction increases as well,
or vice versa. Once you see the pattern, performance measure analysis would
be more useful. So instead of analyzing the two performance measures
separately, try analyzing them together – it could change the way you run
your business.
Analyze from many angles.
The following event might have happened at your office.
Boss: “Eh, why do we get so many complaints from customers? Go take a
look at which product model customers complain about the most.”
Subordinate: “Sure thing, boss.”
The subordinate comes back after 10 minutes.
Subordinate: “Model 323.”
Boss: “We should arrange a meeting tomorrow to brainstorm on how we can
improve this model.”
Subordinate: “Great.”
After that, everyone goes back to their desks to work. A week later, a meeting
is held, and model 323 has been improved. Have you noticed anything
strange? If you only skim through it, everything seems to be normal. But if
you look into it in detail, the solution was already there when the boss asked
his first question.
“Go take a look at ‘which product model’ customers complain about the
most.”
The question contains an important assumption: the product model is the
cause of the complaints. Of course, there will always be a product with the
highest number of complaints. So, the company will end up improving that
product model.
I used this example to illustrate that once you have received a performance
measurement result, the first thing you should do is to consider many angles.
You should not ask only which product model has been complained about the
most. You should try to think of other possible causes as well – which factory
produces most of the product with highest number of complaints, or which
dealer has sold most of the product with highest number of complaints?
Analyzing performance measurement results from various angles could
change your perspective on problems. In the end, you will be able to fix the
problem at its root cause.
Do not believe in the correlation right away.
In one of the previous examples, I suggested that you plot two performance
measures on a single graph to see the correlation. In this section, I want to
show you that the correlation might not always be the causation. Why is that?
A correlation between performance measure A and performance measure B
could be due to various reasons.
Firstly, the correlation could be purely coincidental, especially if you did not
collect enough samples. For example, if you plot the height of 2-3 employees
and their salaries, you might find out that the taller the employee, the higher
their salary, which is very unlikely. This kind of situation happens from time
to time; they plot so many graphs that they come across a strange correlation.
For example, they found a correlation between the budget of science, space,
and technology in the USA, and the rate of suicide by hanging or suffocation.
There was a correlation between the number of deaths from drowning in a
pool and the number of movies Nicholas Cage has starred in. There are so
many other examples. But these should be enough to prove that correlations
do not necessarily imply causal relationships.
Secondly, there is another important factor to consider in analyzing
correlations. However, we have not yet included that factor in our analysis.
Let me give you an example. Have you ever noticed that whenever a police is
out directing traffic, there will always be a traffic jam? There is even a belief
that “police creates the traffic jam.” I believe many people have complained
about something like this before, but do not jump to conclusions. Have you
ever noticed that on Sunday morning, when we rarely get to see the police,
there is no traffic? Well, then of course, it must be “the police that cause the
traffic jam.”
If someone counts the number of police seen on the road, and the time it
takes to commute to one’s destination, and plots a graph, we might get to see
a positive correlation between the two – the more police you see, the worse
the traffic. Now, can we draw the conclusion that the police cause traffic
jams? Not yet. According to the findings (assuming enough data were
collected), we might say that the correlation between the police and the traffic
jams is no coincident. However, it is another thing to claim that the police
“cause” traffic jams; this is something we need to be careful of. The
correlation between the two might not be causal. Rather, the correlation
might be caused by a third variable.
Now, it is getting more confusing. The traffic jam might be caused by an
accident; the police might be directing the traffic because of the accident. The
worse the accident, the more police will be out directing traffic to resolve the
situation; and the worse the accident, the worse the traffic jam. But without
the third variable, we could have concluded that the police cause traffic jams.
In fact, if the police hadn’t done their jobs, the traffic might have been even
worse. In addition, the traffic might be a result of other causes such as rain
and flood – and that is why more police might have to come out and work. As
a matter of fact, it might be “the traffic jam that causes the police to come out
and work,”
Therefore, you cannot jump to the conclusion that one performance measure
has an effect on another simply because there is a correlation between the
two. We have to do something extra in order to prove the logic behind the
correlation. The easiest thing you can notice is that the cause must always
come before the effect. Brand awareness performance measure has to
increase before sales performance measure – that “could be considered” as
logical. However, the other way around would be the end of all. If sales were
to increase before brand awareness, we would not be able to conclude that
brand awareness increases sales, as the effect cannot precede the cause.
However, an increase in brand awareness that is followed by an increase in
sales still does not guarantee the causation. That is why I used the phrase
“could be considered.” In order to validate, we have to “make an
experiment.” An experiment would allow us to hold other variables constant,
while changing only one variable that we are interested in, to see if the
change is significant or not.
If we want to see whether advertisements could increase brand awareness or
not, we have to keep other variables constant - do not give a discount or hire
more salespersons, for instance. Then, see if the increased brand awareness
could help generate more sales or not. If you are really going to conduct an
experiment, you have to design it well, because controlling other variables is
not an easy task. However, it is not impossible. Now, I do not want to turn
this book into a research-based book. So, I just want to conclude by saying
that before jumping to conclusions after seeing a correlation, you have to
consider other possible variables that might be the cause first.
Can you hear what KPI is telling you?
Reading this title, many of you might be concerned that something might be
wrong with me. Am I too obsessed with performance measurement that I am
actually talking to it? Not at all. You can relax now. I am just fine. What I
mean is that if we do not carefully analyze the result of a performance
measure, we might miss a lot of things. Let’s take a look. Assume that the
percentage of defect performance measure is as follows:
Round 1: 10%
Round 2: 5%
Round 3: 3%
Round 4: 2.9%
Round 5: 2.85%
If we simply look through it, we might think that our performance is
improving. Moreover, if we only take a look at the values round by round, we
would only be able to conclude that the percentage of defects is gradually
decreasing. However, if we carefully analyze these numbers, we would see
that these performance measure values are trying to tell us something. We
can see that in the first rounds, the amount of decrease is greater. Later on
(round 3 onwards), the numbers are still decreasing but at a much slower rate,
even though same amount of resources or effort is spent.
Why is that? The process that we are trying to improve is reaching its
improvement limit. The process will not improve any further if we keep using
the same method. We need to change the process or come up with new ideas
to see dramatic improvements. This is where many people make mistakes.
They invest their resources and effort on things that are reaching their limit of
improvement. More importantly, the result of the performance measure might
have already told us that, but we just might not have listened. So, go back and
take a look at your performance measure values to see if they are indicating
anything like this.
Do not analyze results separately.
Sometimes when you analyze each performance measure separately, you do
not get anything out of it. Let me give you an example. Let’s say we have
two performance measures: the number of tasks employees can finish per
week and the task error rate (in percentage). Analyzing them separately might
lead to the following conclusions.
The number of tasks employees can finish per week has gradually increased
since the first week of the year until the 10th week. We might conclude that
our employees have worked very hard, which is why the number of tasks
done has continuously increased. The second one, the task error rate, has
remained stable until the 6th week when the error rate has started to gradually
increase. We might conclude that after the 6th week, the error rate started to
increase because our employees began to work negligently. Also, it does not
seem like the rate will decrease anytime soon.
After this analysis, each department would fix their own problem
independently. However, if we analyze performance measure number 1 and 2
together, we might be able to get something useful out of them. During the
first weeks (weeks 1-5), our employees were able to work more. However,
the error rate was low and stable, which means that the employees still did
not work at full capacity. But the important point is about the 6th week when
employees still managed to finish more work. However, working more
always comes with more errors. Or simply put, employees could finish more
work because they reduced the quality of their work. We can imply further
that the true capability of our employees lies upto the 5th week, because
despite the fact that they were able to finish more work after the 5th week,
they did it by compromising quality. This is just an example. Try analyzing
two performance measures together, and you might come across something
even more interesting than siloed results.
Analyze performance measure to identify capacity.
“Can I hire more people?” Many of you might have heard this kind of
complaint before. Most of the time, these complaints have valid reasons.
They really do not have enough people, and need assistance. However, most
of the time, this is not the case. The person who complains the most might
not be the busiest one. Sometimes the one who says nothing is the busiest
one, needing help the most. This person is just too busy that he or she has no
time to complain. What I am trying to say is that sometimes performance
measures could prove something. For example, you can use a performance
measure to test if your department is working at its fullest capacity or not.
Testing something like this with machines would be very easy, because
machines never slack off or complain. With or without an inspector, they
would still work at their highest capacity. Human work capacity, on the other
hand, is more complicated, because people could work fast sometimes and
work slowly at other times – sometimes they are on fire, sometimes
dispirited.
One of the methods to help estimate such outcome is through an appropriate
performance measure. Let’s assume we want to estimate the capacity of a
certain department that mainly uses human labor. We can start off by listing
all of the tasks this department is responsible for. Then, we can estimate the
amount of time needed to finish each task. For example, if our employees
have to contact clients, we would have to estimate how much time they
would need to spend on a client. Let’s say it is approximately 10 minutes
each. After that, we could collect data for a week (5 business days). For
example, if we know that the department contacted 600 customers this week,
we could calculate the amount of time this department spent working, which
is 600 x 10 = 6,000 minutes. Then, we could find the total amount of time.
Let’s say there are five employees in total. Each of them work 40 hours a
week. Hence, the total time would equal 5 x 40 x 60 = 12,000 minutes.
For the last part, we have to find the percentage the employees spent
working, which equals 6,000/12,000 = 50%. We can roughly see that this
department works only half of the time they have. If this department asks for
more people, we would need to ask them why, given such percentage.
Therefore, performance measures could be another good source of
information when it comes to finding out work capacity.
If you change your perspective in analysis, you might see things
differently.
What I am going to write about for this topic is a simple technique that will
make you look at performance measurement results differently. First things
first, each department has their own performance measures. Why is that so?
Most organizations have an organizational structure that is divided into
various departments; the performance measures are correspondingly divided.
So, what is wrong with dividing our organization into different departments?
I have to say that in fact there are some advantages for doing things this way.
We would get to see the performance of the department we work for.
However when you look through the eyes of each department without
involving other departments, this thing called “sub-optimum” can happen.
Sub-optimum is an attempt to make our department’s performance
measurement result look best, without caring about other departments.
Let me give you an example. If our department gets measured in the amount
of time we spend serving a customer, of course, management would want us
to work hard to reduce this number. What could happen is that when a
customer who has complicated problems comes asking for assistance and
although we know exactly how to solve the problems, instead of trying to
immediately solve the problems, we would suggest the customer to contact
another department as we also know that solving the problems would take a
lot of time, which would make our performance measurement results look
bad. Our performance measurement result might end up looking good for our
department, but this is definitely bad for the company.
Therefore, in addition to analyzing the data from the perspective of each
department, we should look at the process. In the case above, we should also
analyze the total time spent from when the customer walks in until the
problem is solved, so that we will be able to see the results from another
perspective. This will definitely help the organization as a whole.
Analyze the result like you are taking a photo.
There is a particularly useful camera function, which I believe many people
have already tried before: zooming. Sometimes we zoom in, sometimes zoom
out. No one always zooms in to get close-up photos. And likewise, no one
always zooms out to capture only tiny humans with the surrounding scenery.
Same goes for performance measurement analysis – we have to both zoom in
and out. Zooming in means to analyze each performance measure in detail,
because sometimes when we only look at the index, we might not get
anything out of it. Just like simple mathematics, when we add up the values
of many performance measures to create an index, the rate of change of that
particular index would gradually decrease. Or simply put, it would remain
stable. However, a stable index does not imply that every performance
measure is stable; some might increase while others decrease. By zooming in,
we would be able to see a much clearer picture.
Management needs to zoom out as well. Zooming out means to look at the
overall picture instead of analyzing each performance measurement value.
There is a saying, “not seeing the forest for the trees.” It explains everything.
To zoom out is to look at the past in order to predict the future of the whole
organization. So, try to adapt my suggestions to your organization, and do not
forget to zoom in and zoom out.
Look at both opportunities and threats.
Let me start off by telling you about two incidents.
Incident 1
Boss: “Why didn’t we reach our sales goal? We’re still 50% short.”
Subordinate: “Because right now, the quality of our product is lower, we’ve
received many customer complaints.”
Boss: “If that’s so, let’s find a solution with the production team.”
Incident 2
Boss: “We did great this time. Look, our sales are 50% higher than the goal.”
Subordinate: “You should treat us to a meal then.”
Boss: “Of course. Tell me when, and we’ll go celebrate.”
These two incidents might seem ordinary at first sight. However, if you
consider the details, you would see that people are only worried when there is
a problem, but never realize potentially missed opportunities. In the first
incident, the company is 50% below the goal. The company might begin to
worry and try to find a solution, which is a good thing that could lead to
improvements. Whereas in the second incident, the company overachieved
the goal by 50%, so they celebrated their success without knowing why the
sales have exceeded the goal so much.
Surpassing the goal by 50% is something to be proud of for sure, but we
should pay careful attention to the rationale behind it in terms of strategic
planning. In this case, the company’s overachievement might simultaneously
indicate that the company might have set the bar too low. Therefore, the
resources that were delegated to achieve the goal was also too low, despite
vast opportunities in the market. If the company were able to set a more
accurate goal, they would have delegated more resources that would have
improved the outcome. Instead of finding the reasons behind bad
performance measurement results, asking why the outcome is highly different
from the goal might give you some useful insights.
Analyzing a pair is better than analyzing an individual.
Sometimes analyzing two performance measures together could give you
interesting data. Here’s an example. Let’s say the value of the quality index
has increased, while the value of productivity index has decreased
(productivity = output/input). In this case, if we analyze the two performance
measures separately, we would not get much useful information out of our
analysis. We might try to find supporting evidence for what we saw. For
example, why has quality improved? Because of this and that… Why has
productivity fallen? Because of this problem and that problem - or something
along these lines.
However, if we analyze both performance measures together, we might find
out that the reason why quality has improved while productivity has fallen is
not because we have fixed the problem at its root cause, but rather because
we have been working on the same tasks repeatedly. We might try to use
better quality control. And once we are able to identify the problem, we
would work on the problem over and over again, until the quality problem is
eliminated, making quality performance measures look better (product return
rate would be reduced, for instance). However, we have added so much input
(such as labor) that it causes productivity (output/input) to decline.
What should we do if we find that quality performance measure has
improved and the quality problem has been dealt with, but still productivity
has not increased at all? Many people have faced this problem before, and
many might wonder why such a thing has happened? If the quality is better
and the problem is eliminated, why doesn’t productivity improve? The reason
is that once the problem has been solved, we would have leftover resources.
We might not have utilized those resources to the organization’s advantage.
Let me give you an example. In the past, there was a problem with our
process, which was why we had to redo everything all the time. It used to
take us 8 hours to manufacture one product, but we solved the problem, and it
only takes us 6 hours now. Is that better? Yes, it is. But now the question is
what did we do with the extra 2 hours? If we don’t do anything, productivity
will not improve.
See how analyzing two performance measures together can make things
much more interesting? Now let’s assume that we find out that both quality
and productivity have improved. What does it mean? It means we are on the
right track. However, it does not mean we could not improve further. We
have to continue to identify if there is anything we have not fully utilized in
production (after solving the quality problem.)
However, discovering that both quality and productivity performance
measures are low, is clearly a bad thing because it shows that we really have
problems. But if we have decent performance measures, we should be able to
identify the source of the problem, and solve it before it could get any worse.
In summary, analyzing two (or more) performance measures at the same time
will help you get a clearer picture – it is obviously better than analyzing each
performance measure individually – “two is better than one.”

Analyze the coherence between strategy, operation, and profit.

Performance measurement results usually reflect the connection between
each performance measure. Generally speaking, if our organization has a
strong strategy and good operations, we should be able to generate more
profit. Conversely, if our strategy and operations were poor, our profit would
be low – we might even incur loss.
Let me give you an example. If we have invented a very interesting product
(good strategy) and the cost per unit is very low (excellent operation), our
profit would increase. However, if we manufacture a product that customers
do not like (poor strategy) and the cost per unit is low (excellent operation),
we should be able to generate some profit, but it might not be much. If we
manage to produce a very interesting product (good strategy), but the cost per
unit is very high (poor operation), the profit will not be that higher either.
Lastly, we manufacture a product that customers do not like (poor strategy)
and the cost per unit is high (poor operation), you should be able to guess
what would happen to the profit. We might even incur loss. What I am trying
to say is that even though strategy and operations might lead to certain
financial results, it might take some time. You should keep this in mind when
reading performance measurement results.
Analyze to prioritize.
When we collect data from our performance measures, apart from the results,
we would also receive another important data, which is the “priority” of those
performance measures. Many of you might have some experience in
measuring customer satisfaction. And what you are likely to find out is that
the customers are satisfied with many aspects, and there are still many other
things that customers are dissatisfied with. And this is where a problem
usually arises – sometimes there are many dissatisfied customers, while we
have limited amount of resources. So, which problem should be resolved
first? Some people might say, “Choose the one with the lowest score.” But
here’s what is interesting: The thing that customers are most dissatisfied with
might not be the thing that the customers give the highest priority to.
If we measure the level of importance of each topic, we would be able to
analyze and make decision easier. There are four possible scenarios:
1) High satisfaction; customers give high priority.
This means we have done an excellent job. Our strategy would be to
maintain this level of satisfaction.
2) High satisfaction; customers give low priority.
Many might perceive this as totally fine because we have done well
anyway. However, this is where you are wrong. We might have spent too
many resources in the aspect that customers do not give priority to. What
we should do is to reallocate our resources to improve things with higher
priority.
3) Low satisfaction; customers give low priority.
Many people might panic when they see low customer satisfaction
scores. However, if the customers do not give importance to that
particular aspect, you could relax a little. It means that we could solve
this problem when we have surplus resources, because customers do not
give high importance to this aspect. This is where we can prioritize
which problem should be solved first and which problem should be
solved later.
4) Low satisfaction; customers give high priority.
We need to work on it immediately by using our resources to solve it or
reallocating our resources from the less important aspect to fix this
problem first.
This is another example, which proves that result analysis can really be
beneficial.
Compare to your plan.
We frequently forget comparing our performance results to the plan we had
made. Some might ask why it is necessary to do such a thing because we
would not be able to fix it anyway. What’s done is done, the results are here,
why should we compare them with anything at all?
There are two main benefits of comparing results with the predefined plan:
1) It would make us seek for the reasons why we could not follow the plan
or why we had overachieved. Identifying the source of problems or
opportunities help us solve problems or recognize opportunities better in
the future.
2) We could plan ahead even better. It is true that no plan could be 100%
accurate; but even if the plan fails, we can learn from our mistakes,
which would help us improve.
Do not forget to review performance measurement results and compare them
to your plan.
Many heads are better than one.
With this headline, I am talking about another technique that I would like to
introduce for when you want to analyze and interpret performance
measurement results. If an analysis is made by one person, usually, the
weakness is that the interpretation is from a single perspective. If an
accountant writes the analysis, we would only get the accountant’s
perspective; if an engineer does the analysis, we would only get the
engineer’s perspective. And do not argue on whose perspective is better,
because the quarrel would never end.
My suggestion is to analyze the results as a team. Each person can express
their opinions, and we could be quite certain that the analysis is accurate
when there is consensus. However, if each team member sees things
differently, it could still be useful because these perspectives could serve as a
list of explanations, giving you fresh perspectives. An analysis from a team
with members from different departments sharing their perspectives could
make you see the overall picture better than an analysis performed by a single
department could. In addition, the former kind of analysis is more interesting
and useful. I would like to ask for only one thing – do not try to compete with
each other, because in the end, everyone might end up losing.
 Chapter 14
Make your performance measurement system sustainable
Have you ever seen anyone who uses a hammer to saw wood? And have you
ever seen anyone who uses a saw to hammer a nail?
What would happen if so? Same goes for performance measurement. If you
have it but never utilize it, or use it the wrong way, apart from not creating
anything good, it can cause damage. If you want performance measurement
to promote sustainable benefits for your organization, what do you have to
do? You can get started by reading this chapter.
#1 Consistency is the key.
Performance measurement is like having a girlfriend. If you want her to stay
with you for a long time, you have to take good care of her – not only during
the early stages of the relationship (the so-called promotion period), only to
abandon her after a while. Many organizations do just that. During the early
days, they would collect data, analyze the results, and utilize the information
to improve their work processes. But after a while, they begin to get bored of
it. They measure consistently at first, then it starts to become an on-and-off
thing. And later on, they stop measuring altogether.
If such is your plan, then please do not take on performance management at
all, because you have to invest labor, time, and money in building a
performance measurement system. Using it for only 3-4 months is quite a
shame. If you think performance measurement is good and important, then
you have to consistently give it attention. The world is very fast-paced
nowadays – analyzing performance measurement results once a year is not
adequate. Especially, operations should be measured weekly. However, if the
measures are meant for the company executives, then once a month or once a
quarter should be enough. Do not forget that “consistency” is the key. Do not
let your hard-earned performance measurement be forgotten. It would be
such a shame.
#2 Good communication is a must.
If you want your organizational performance measurement system to be
sustainable, its reports have to be able to grab attention. Here are some
interesting techniques that you could use:
1) Only report the data that is relevant to the readers.
Do not send less or more than what the readers need. Too few information is,
of course, not useful. But the more common mistake is reporting too much
information, because people simply report everything that they have. Allow
me to compare this to driving a car. Can you imagine what would happen if
you had to monitor 200 gauges? You would never look at any of them,
because there are just too many.
2) Arrange the contents logically.
This part is also important. Creating a good storytelling for what you want to
present – explanations for each section, followed by charts and analyses, for
instance – would make the report easy to read.
3) Ensure that the readers could comprehend the data.
Try to avoid unnecessary jargon If they have to be included, provide their
definitions, because if the readers could not understand them, the whole
report would have no value at all.
4) Make it as brief as possible.
Most management do not have much spare time. Therefore, good reports
should be brief and to the point, but not too brief that the report is incomplete.
You just have to ensure that the report is not dragging on. You can start off
by reading the report, and see which parts you can remove without losing
useful information. If you choose to keep too much information, management
might not read your report at all.
Do you have any organizational performance measurement reports to show
me? Which kind of report is good? And which kind of report is bad? There
are many bad examples out there. Where can we find good examples? I
always say that good reports contain two keywords: “simple” and “relevant.”
If your report has these two qualities, I guarantee that everyone would want
to read it. But where’s an example of a good report? I would say it is one that
you read every day – it is Facebook. For me, Facebook is the world’s best
report system. And where exactly is my proof? Well, there are billions of
people around the world who read contents on Facebook. If you do not call
that the best, I do not know what else to call the best.
Let’s take a look at “simple” first. Facebook itself is very simple. You do not
have to learn how to use it – you can simply sign up, which is very easy, and
you can start posting or reading other people’s posts. There is nothing
complicated about it, right? Moreover, your friends’ posts are also very easy
to understand. For example, if the caption says, “Holiday, day off,” with a
beach photo, you do not need any further explanation to elaborate on what it
means. I think it is very easy to comprehend.
The next factor is “relevancy.” Facebook beats other websites because of this
particular aspect. Imagine all of us entering the same website; we would get
the same content. No matter how hard that website is trying to categorize its
contents or trying to cover every topic, it would not be able to tailor a specific
message to each individual. However, for Facebook, each person’s wall
contains different contents. My wall is different from yours. And you will
only receive relevant and interesting information, because that information
belongs to your friends. My wall contains information about my friends, and
your wall contains information about your friends. Despite the fact that we
are using the same website, we would still receive different information. Isn’t
that awesome? If a report is good, then there is a higher chance for the
performance measurement system to be utilized.
#3 Prioritize, prioritize, prioritize.
Setting priorities is vital for organizational performance measurement. The
hardest part of management’s job is to utilize the limited amount of resources
to solve an unlimited amount of problems. Therefore, performance
measurement could act as a tool to help management set priorities as to which
problems deserve to be solved first and which problems should be solved
later, given a limited amount of resources. Using performance measurement,
we would know what is important and what is not. It would allow us to
allocate the limited amount of resources to fix the most critical problems.
We could divide priorities into short-term and long-term ones. For example,
if we are not able not deliver goods to customers in time - no one would
argue that this is not important – in the short-run, we have to find those
particular goods from somewhere else and deliver them to customers. But in
the long-run, this would not be good for the business. We should consider
increasing our production capacity in order to support the rising demand. If
the analysis from your organizational performance measurement report can
reflect the order of priorities, I believe the readers would find it useful. And
eventually, the performance measurement would be beneficial to the
organization sustainably.
#4 Feedback is important.
Do you want to know your exam scores? I believe that everyone would say,
“Yes.” However, some students of mine said that they do not want to know
their scores for my class. I would laugh and reply, “Deep down, you want to
know the score. You just don’t want your friends to know how much you
scored, am I right?” Performance measurement is like taking an exam. People
who get measured would want to know the results. If you do not want them
to know the results, then why measure at all? Without the results, people who
get measured would never know how they can improve – they would
continue to do what they are doing, which does no good for the company.
Thus, valuable feedback should be specific, meaning that you should tell each
department how well they perform – in which aspects they do well and which
aspects could be improved. Simply stating that, overall, all departments have
done a good or bad job would not be useful because departments would not
know which areas they need to improve. If they have done a good job, they
should receive recognition. You should not complain when they perform
poorly, while saying nothing when they do well. More importantly, when
they do not perform well, you should not only punish them and find a person
to blame. You should be removing the organization’s problems, not the
employees.
Feedback should be constructive, meaning that people must receive rewards
and recognition to continue to perform well. However, this does not
necessarily mean that you should compliment your employees when they
perform poorly. Rather, what you should do is to tell them why they
performed poorly and how they can improve. Do not make complaints that
would only hurt employees’ feelings, and do nothing to help them improve.
We often see this happen.
“You did really bad. Look at the result. I haven’t seen anyone who performed
this bad before.”
“You have been working here for so long, is this the best you can do?”
This is the kind of feedback that does no good. Feedback is vital. If you could
provide good feedback, I believe that the chances are high that your
organizational performance measurement system will be sustainable.
#5 Rewards and success
A performance measurement system would succeed and be beneficial to the
organization sustainably once we add another positive force. When
employees achieve something that makes the organization reach or surpass its
goal, they would expect to receive rewards. Many people might think rewards
only refer to monetary rewards. Various organizations, especially in the
public sector, would say that this is the difficult part because of budget
constraints. In fact, money is a kind of reward, however, it is not the only
kind of reward. Another type of reward that is no less effective than money is
recognition.
Most of the time, employees want recognition more than money. Or in some
cases, money can demotivate them. Let’s take a look at the following
example. Assume that I just finished teaching and am about to drive home. It
is raining heavily, and I see a group of my students waiting for a taxi, so I
decide to stop my car and ask them where they are headed. Once I learn that
they are headed the same way as I am, I tell them to hop in for a ride. When I
drop them off, they hand me 10 dollars, saying that it is for the ride. Next
time, I would not stop to pick them up. You can see from this example that I
would prefer a “thank you” more than money. By giving me money, the
students would be discouraging me from giving them a ride again. Rewards
are another thing that would make organizational performance measurement
sustainable.
Cautions of organizational performance measurement
We frequently hear many suggestions regarding how to fully utilize
organizational performance measurement. In this postscript, I would like to
introduce the cautions – if you do not want your organizational performance
measurement to fail, what shouldn’t you do?
1) Do not use performance measurement to control everything.

It is true that one of the pros of performance measurement is to create some

kind of control system to make employees follow the organization’s goals.
But do not forget that employees are human, not machines. Everyone wants a
certain level of freedom. Not being able to achieve it because of performance
measures would make their jobs boring. Working under various performance
measures could unnecessarily cause employees stress. In the end, employees
would not want to work at all.
2) Do not tie performance measurement to punishment.
We want to eliminate problems from your organization, not employees. This
mindset is very important, because we want performance measurement to be
a tool to help employees eliminate problems, rather than being a tool to harm
them. Tying performance measurement to punishment would do more harm
than good. Although punishment is another kind of driving force, it is rather
considered as a negative force. Generally, positive force is more popular. For
example, when the result of performance measures show a sign of
improvement and the organization can generate more profit, employees
should receive rewards.
These two cautions are very significant. Be careful not to let these happen to
your organization.
 Postscript
I hope this book helps you to understand how vital performance measurement
is. More importantly, I hope you learn how to design and put performance
measurement system to good use. The idea of performance measurement is
becoming more and more important. This book might be one of the starting
points that helps ensure performance measurement succeeds in organizations.
By “succeed,” I mean the thing that we planned to measure improves. For
example, if we have a customer satisfaction performance measure, then
customer satisfaction has to improve. If we have a defect performance
measure, the number of defects should decrease.
Enjoy performance measurement!
 Bibliography
Sajjanit, C. and Rompho, N. (2019) Measuring customer-oriented product
returns service performance, International Journal of Logistics
Management, 30(3), pp. 772-796.
Truktrong, S. and Rompho, N. (2019). Comparing the key success factors
affecting change operations between state-owned enterprises and
private organisations. The 26th European Operations
Manangement Association, 17-19 June 2019, Helsinki, Finland.
Truktrong, S. and Rompho, N. (2019) Willingness to Change in State-
owned Enterprises and Private Organizations, Change
Management, 18(2), pp. 23-36.
Rompho, N. (2018) Operational performance measures for startups,
Measuring Business Excellence, 22(1), pp.31-41.
Morita, M., Iwai, C., Rompho, N., and Phadoongsitthi, M. (2017)
Comparison of Group Decision Making in Japan, Thailand,
Vietnam, and Russia Using a Business Game, Simulation &
Gaming, 48(6), pp. 791-813.
Rompho, N. (2017) HC and financial performance with two HRM
strategies, International Journal of Productivity and Performance
Management, 66(4), pp. 459-478.
Phadoongsitthi, M., Rompho, N., Iwai, C., and Morita, M. (2017) Effects
of national culture on group decision making: a comparative study
between Thailand and other Asian countries, International Journal
of Economics and Business Research, 13(2), pp. 110-133.
Amornpashara, N. Rompho, N., and Phadoongsitthi, M. (2015) A study of
the relationship between using Instagram and purchase intention,
Journal of Global Business Advancement, 8(3), pp. 354-370.
Boon-itt S. and Rompho, N. (2012) Measuring Service Quality
Dimensions: An Empirical Analysis of Thai Hotel Industry,
International Journal of Business Administration, 3(5), pp. 52-63.
Rompho, N. (2012) An experiment in the usefulness of a strategy map,
 Measuring Business Excellence, 16(2), pp. 55-69.
Rompho, N. and Boon-itt, S. (2012) Measuring the success of a performance
measurement system in Thai firms, International Journal of
Productivity and Performance Management, 61(5), pp. 548-562.
Rompho, N. (2004) Building the Balanced Scorecard for the
University Case Study: The University in Thailand. In A.
Neely, M. Kennerley, and A. Walters (Eds.). Performance
Measurement and Management: Public and Private, 28-30
July 2004, Edinburgh. Stirling: Centre for Business
Performance, Cranfield School of Management.
About the author
Professor Dr. Nopadol Rompho
Professor Dr. Nopadol Rompho graduated with a Bachelor of Engineering,
majoring in Chemical Engineering (second class honors), from
Chulalongkorn University, Thailand; a Master of Science in Chemical
Engineering from Oregon State University, USA; an MBA from Thammasat
University, Thailand; and a Doctorate of Philosophy (Management) from
University of Glasgow, United Kingdom.
Professor Nopadol has work experience in both government and private
sectors. He worked as a consultant for many large organizations, and has
continuously produced outstanding academic works and research about
organizational performance measurement, operations management, and
quantitative analysis. He is a founder of ZeeZcore that offers the OKRs
software as a service.
Currently, he is a full-time operations management professor at the
Thammasat Business School in Thailand, a member of an international
editorial board, and a member of many other boards in academic societies in
Thailand.
He can be contacted via email nopadol@tbs.tu.ac.th or LinkedIn: Nopadol
Rompho