110 International Journal of Electronic Business Management, Vol. 1, No. 3, pp.

110-119 (2003)

KPI WITH DATA FLOW ANALYSIS FOR CPFR:

A CMC CASE STUDY
James T. Lin1*, Phyllis Chang2, Juin-Han Chen1 and Wei-Xiong Xin1
1
Department of Industrial Engineering and Engineering Management
National Tsing-Hua University
Hsinchu (300), Taiwan
2
NiceShipping Technology Corporation
Taipei (100), Taiwan

ABSTRACT
The increasingly global operations and supply chains are significant. And the product life
cycles are decreasingly shorter. Therefore, to rapidly react to global market immediately
and accurately, and to cope with the increase in variability in the supply chains, many
enterprises strive to share internal and external information devotedly. CPFR (Collaborative
planning, forecasting and replenishment) is one new collaborative commerce model to
enable collaborative relationships between buyers and sellers through co-managed
processes and shared information. The early adopters of CPFR are retailers and their
suppliers to increase the visibility to each other and let the suppliers to exactly replenish
goods at right time and accurate quantity. The benefits of developing CPFR for retailers are
lower inventory levels, higher fill rates and sales. And the benefits for suppliers are lower
order cycle time, better customer service and inventory reductions.
CMC is the first one CPFR pilot manufacturer in Taiwan to develop collaboration between
CMC and its trading partner, an overseas buyer with well-known brand. The primary
products of CMC are optical discs, optical recording media, and compact disc writable.
Based on the production environment of CMC, this paper presents the KPI and its data flow
for the CPFR system. Before implementing CPFR, the customer demand trends are
anticipated from sales interviews with customers, the inventory reports of overseas Hub and
factories. And then, develop the master plan according to market demand and historical
sales data of each customer. Finally, manufacture and ship the products according to the
actual orders. As the demand information is not real time and without customers’ promise,
the demand forecast and actual orders are different. Accordingly, unnecessary inventory,
inventory carrying costs and warehousing costs are increasing. After implementing CPFR,
CMC and buyer can collaborate planning, forecasting and replenishment in accordance
with the demand forecast. Besides, in the CPFR process, seller and buyer will resolve
exceptions collaboratively. Therefore, the demand forecast is progressively with
commitment. This paper also analyzes the CPFR related information systems and then to
design the KPI (Key Performance Indicator) information flows. From the KPI information
flows, we will comprehend the influence of each KPI and then how to promote these
performances.

Keywords: Collaborative Planning, Forecast and Replenishment (CPFR), Key Performance

Indicator (KPI), Benefits Analysis

1. INTRODUCTION decreasingly shorter. Therefore, to react to global

market immediately and accurately, and to cope with
the increase in variability in the supply chains, many
The increasingly global operations and supply
enterprises strive to share internal and external
chains are significant. And the product life cycles are
information. Collaborative planning, forecasting and

*
Corresponding author: jtlin@ie.nthu.edu.tw
J. T. Lin et al.: KPI with Data Flow Analysis for CPFR: A CMC Case Study 111

replenishment (CPFR) is one new collaborative transport CD-Rs is by sea.

commerce model to enable collaborative relationships CMC’s main production model was MTO.
between buyers and sellers through co-managed After CMC received the orders, it took one to two
processes and shared information. The early adopters months in the slack season and three to four months
of CPFR are retailers and their suppliers who want to in the rush season to deliver the product to customers.
increase their visibility to each other and let suppliers Because of the high risk of fluctuating prices during
replenish goods at the right time and in accurate the order cycle time (OCT), CMC’s main customers
quantities. The benefits of developing CPFR for frequently ordered small volumes and asked for
retailers are lower inventory levels, higher fill rates, vender managed inventory (VMI) model. Therefore,
and higher sales. The benefits for suppliers are lower CMC had to be responsible to storage and inventory
order cycle times, better customer service, and costs.
reductions in inventory. As shown in Figure 1, CMC cannot directly
CMC is the first CPFR pilot project in Taiwan know the demand of retailers who are the buyer’s
to implement a forecasting collaboration process customers. The buyer collects demand from
between CMC and its overseas buyer. Therefore, this distributors and then gives CMC order forecast one
study will analyze CMC’s supply chain management month in advance. Then CMC plans its production
issues and then introduce the CPFR developing schedule by the order forecast. However, the
process with KPI (Key Performance Indicator) inaccuracy of forecast results in frequent change in
selection and performance evaluation. the production schedule. Moreover, for lack of
information from demand side, CMC cannot forecast
1.1 The Industry Characteristics of Compact Disk demand voluntarily and then lead to high inventory
Manufacturer level, low order fill rate and long order cycle time, …
There are three different types of compact disk etc. Consequently, the important supply chain issues
(CD) products: Read Only, Write Once and Read faced by CMC are to improve accuracy of demand
Only (CD-R), and Rewritable (CD-RW). Currently, forecast, lower inventory level, and shorten order
the main global market trend for compact disk is the cycle time.
Read Only type that is used as storage for software,
music, movies, etc. However, as the prices of the Manufacturer Distributor Retailer

CD-R and the CD-R driver drop, the market for

Writable Optical Storage Media is growing rapidly Aggregate Demand
Retailers
Compact disk manufacturers need the program Customer order forecast DC1
to CMC 1 month early
content of Read Only CDs in advance. Therefore, the DC2
Retailers

production model for Read Only CDs is make-toorder Production A company

DC3
Retailers
planning
(MTO). Writable Optical Storage Media can be make- Distribution
planning
Retailers

to-stock (MTS) according to demand forecasts (Zeng,

2002). Information flow
Production Manufacturing Shipment
Accordingly, with regard to Writable Optical scheduling
3-5 Days 2 Days
Planning
2-4 Days
Shipping

14-30 Days
Material flow

Storage Media, CMC and its trading partner, an Figure 1: Product distribution channels of CMC
overseas buyer with a well-known brand, can develop
a CPFR process for collaborative planning, Before developing CPFR, the customer demand
forecasting, and replenishment to reduce the risk of trends are anticipated from sales interviews with
inaccurate demand forecasts. In the CPFR process, customers, the inventory reports of overseas Hub and
seller and buyer will resolve exceptions factories. And then, develop the master plan
collaboratively. Therefore, the demand forecast according to market demand and historical sales data
becomes progressively more accurate and thus of each customer. Finally, manufacture and ship the
stabilizes CMC’s production planning and products according to the actual orders. As the
replenishment. demand information is not real time and without
customers’ promise, the demand forecast and actual
1.2 The Issues of Compact Disk Manufacturer in orders are different. Accordingly, unnecessary
Supply Chain Management inventory, inventory carrying costs, and warehousing
The lumpy demand for compact disks results in costs are increasing.
frequently altering production plans. Accordingly, the By means of developing CPFR, CMC and its
manufacturer can not master order due dates precisely buyer can collaborate on planning, forecasting, and
and this lowers the customer service level. CD-R has replenishment in accordance with the demand
apparent rush and slack seasons and its price forecast. Besides, in the CPFR process, seller and
fluctuates. Moreover, it is not economic to transport buyer will resolve exceptions collaboratively. In the
by air, because of the large quantity, low price, large forecast horizon, the collaborative demand forecast
volume, and heavy weight of CD-R. The main way to
112 International Journal of Electronic Business Management, Vol. 1, No. 3 (2003)

has some degree of commitment. And in the frozen CPFR and to provide a checklist for each step to
period, the collaborative demand forecast has more ensure all critical items have been completed. The
degree of commitment. Therefore, the demand CPFR roadmap is divided into five steps as follows
forecast becomes progressively more accurate and [6].
CMC can manufacture products according to the
demand forecast in the frozen period. Consequently, Step 1: Evaluate current conditions
developing CPFR process will stabilize CMC in The first step is to evaluate the role and the
production planning and replenishment and will needs of the company in the supply chain. The next is
reduce the risk of CMC to produce unrequited to define the company’s CPFR objectives. This
products. allows the company to recognize the current state and
where to change to implement CPFR. And then, the
1.3 Research Objectives company should select trading partners to develop
However, how to evaluate the key performance CPFR by evaluating if the trading partners have the
of developing CPFR? This research will analyze appropriate commitment and experience with CPFR.
CPFR related information systems and then design The CPFR capacity assessment tool is used in this
the KPI (Key Performance Indicator) calculator assessment process to evaluate the developing CPFR
information flows. From the KPI calculator ability of trading partners.
information flows, we will comprehend the influence
of each KPI and then how to promote these Step 2: Define scope and objectives
performances. After selecting CPFR trading partners, the
company makes the following preparations.
2. LITERATURE REVIEW 1. Gain commitment from trading partners
2. Assign team members and establish their roles
3. Select products and locations that will be
2.1 The Principle of CPFR included in the process
A set of CPFR guidelines are supported and 4. Decide which parts of the nine-step CPFR
published by VICS [3]. Trading partners share their process to test
plans for future events, and then deal with exceptions 5. Establish key performance metrics to measure
collaboratively. By means of sharing demand forecast the initiative’s success
information, CPFR provides the process to link
business plans and ensure supply chain Step 3: Prepare for collaboration
synchronization. Thus, CPFR offers trading partners In step 3, the company studies the details of the
more accurate demand forecasts for supply chain CPFR business process and identifies the technology
inventory and cost reductions. Therefore, the and additional resources required to support
principle of CPFR is to facilitate trading partners developing CPFR. Sales and replenishment team
sharing information based on reliance to enhance the members develop rules for managing exceptions and
information visibility and accuracy of forecast. The changes. Collaboration technology team members
buyer and seller collaborate a mutual forecast scheme install and configure the information systems (such as
for consumer demand. By means of this scheme, the purchased, developed, or simple spreadsheets and
buyer and seller can share responsibility for forecast e-mails, etc.) used to support collaboration between
risk due to the commitment. For example, the retailer partner teams. At the end of this step, collaboration is
will continually adjust sales forecast depending on ready to begin.
the point-of-sale information and then enhance the
accuracy of order forecast. Step 4: Execute: Performing the pilot
In step 4, the sales and replenishment
2.2 The Roadmap of CPFR collaboration teams begin to exchange forecasts with
each other, and the collaboration technology team
In pilot run step of developing CPFR,
gains experience managing the environment. The
companies can start with several partners and
initial results and participant comments should be
products to focus on a particular process area like
summarized and reviewed by the trading partners.
collaborative order forecasting before expanding into
The project team should identify future technology
the entire CPFR process. This pilot run step will help
and software modifications that will enhance the
companies understand the benefits of CPFR, the
process.
requirement of document changes for the future
expansion, and the strategies for overcoming
Step 5: Assess results and identify improvements
obstacles.
In step 5, the team and their managements
VICS [3] proposes a roadmap of CPFR is
review the team’s progress, report results to their
proposed by VICS to assist companies in developing
respective organizations, and prepare for broader
J. T. Lin et al.: KPI with Data Flow Analysis for CPFR: A CMC Case Study 113

CPFR rollout. There are a number of ways to expand

CPFR, as follows:
1. Expand to other CPFR processes
2. Add Stock-Keeping-Units (SKUs)
3. Increase the level of detail
4. Automate the process
5. Add trading partners
6. Integrate the results
Figure 2: The KPIs of CPFR [1]
2.3 The Benefits of Developing CPFR Process
The benefits of CPFR mentioned in VICS’ The agreement on KPIs by both trading
CPFR guide [1] include improving the efficiency of partners is essential for successfully tracking the
companies, improving the cost structure, integrating progress of the collaboration and in order to measure
the supply chain closely, and acquiring better the benefits achieved through CPFR. Moreover, these
competitiveness. Barratt and Oliverira [2] suggest definitions for KPIs leave enough scope to adjust
additional benefits such as reduced cost, more them according to specific situations such as different
predictable order cycles, faster inventory turns, more forecast periods or frozen time periods.
frequent deliveries, information formatted to facilitate
usage, increased customer service, etc. VICS 3. CMC CASE STUDY
classifies the benefits of CPFR according to the
following four types: inventory, revenue, process
This project is an IT Applications Promotion
efficiency, and transportation.
(ITAP) D project supported by the Department of
Furthermore, some discussions in the literature
Industrial Technology (DoIT), Ministry of Economic
suggest benefits of CPFR in category management
Affairs ROC. The objectives of this project are to
and efficient customer response. Holmstrom et al. [4]
develop CPFR and shipment tracking. The related
illustrate that combining CPFR with category
information systems of this project are SPE
management will facilitate the replenishment process,
(Shipment Planning & Execution), SV (Shipment
achieve more accurate forecasts, lower inventory
Visibility) and IV (Inventory Visibility). This section
costs, and manufacturing efficiency.
describes the demand forecasting and planning
For compact disk manufacturers, reduced order
process before and after CMC developing CPFR
cycle time is the principle benefit of developing
process.
CPFR. In addition, the customer service level can be
promoted resulting from the collaborative processes
3.1 CMC’s Demand Forecasting and Planning
of trading partners. Moreover, detecting and resolving
Process before Developing CPFR
the exceptions will need the manufacturer to
Before developing CPFR, CMC’s demand
understand customer demand trends and product life
planning mainly depended on demand forecasts from
cycles.
its trading partners. CMC based its production
capacity on these demand forecasts. However, the
2.4 Key Performance Indicators of CPFR
trading partners were not responsible for these
KPIs are used to measure the overall
demand forecasts. CMC could not implement its
performance of the relationship between
short-term production schedule until customers
manufacturer and retailer. Not all of these KPIs are
confirmed actual demand. Accordingly, before
required by CPFR projects to measure the
developing CPFR, CMC’s production plan frequently
performances. Only the KPIs most relevant to the
changed. CMC’s demand planning of CMC before
current relationship of trading partners should be
developing CPFR is described as follows.
selected in order to facilitate implementation of
CPFR. Besides, the responsibilities for measuring
Step 1: The customers gave six-month forecasts to
each KPI should be assigned according to the relative
CMC twice a year. These forecasts were aggregated
to each partner’s business processes and ability to
by sales of CMC and given to production planners to
measure the results. ECR (Efficient Consumer
plan long-term production capacity. Production
Response) Europe classified the key performance
planners then generated an aggregate production plan
indicators (KPIs) of CPFR into the ten categories
to sales about customer demand.
listed in Figure 2 [1].
Step 2: The customers provided short-term order
forecasts to CMC salespeople every month. These
order forecasts are given to production planners to
allocate capacity and adjust the production schedules
that are given to sales to meet customers demand.
114 International Journal of Electronic Business Management, Vol. 1, No. 3 (2003)

CMC and its trading partner input each order forecast

Step 3: The customers placed orders according to real in the CPFR system.
demand. The quantities of actual orders and order
forecasts might be different. Thus, the production Step 4: The second week of each month is the time to
planners verified if CMC could fulfill these orders generate exceptional reports according to exception
and then reply to sales order delivery dates and criteria. According to the collaborative arrangement,
quantities. Finally, sales confirmed order delivery the exceptions are mainly resolved first by CMC as
dates and quantities with their customers and then follows.
input customer order related data into the Enterprise 1. When the order forecast values of CMC and its
Resource Planning (ERP) system. customer are both outside of the limits set by the
The previous demand forecasting and planning exception criteria and the customer’s values are
process produced variations between order forecasts larger than CMC’s values, CMC will first verify
and actual orders, which resulted in frequent if it can adjust more capacity to fulfill customer
alterations of capacity and production planning. demand. If it is feasible, CMC will increase the
order forecast to solve exception. Otherwise,
3.2 CMC’s Collaborative Demand Forecasting CMC and its customer will generate new order
and Planning Process after Developing CPFR forecast collaboratively.
By developing a CPFR process with one 2. When the order forecast values of CMC and the
trading partner, an overseas buyer with a well-known customer are both outside of limits set by
brand, CMC transformed its demand forecasting and exception criteria and CMC’s values are larger
planning process to solve the problem of unstable than its customer’s values, CMC will decrease
capacity and production planning. The demand the order forecast to solve the exception.
forecasts are collaboratively forecasted with CMC’s
trading partners to improve the degree of accuracy. Step 5: In the frozen time period, if the corresponding
This process stabilizes CMC’s capacity and order forecasts are in agreement, the order forecasts
production planning resulting from increasing are frozen to turn into actual orders. Otherwise,
accuracy of forecast and commitment from CPFR repeat step 3 and 4 iteratively to collaborate order
process. Because CMC’s CPFR process mainly forecasts.
focuses on order forecasts, sales forecasts are not
included in this process. CMC’s collaborative Step 6: On the fifteenth of each month, order
demand forecasting and planning process after forecasts are frozen and its customer starts to place
developing CPFR is described as follows. actual order according to frozen order forecasts of
that month.
Step 1: At the beginning of each year, CMC and its
trading partner collaborate on the following activities. Step 7: At the end of each month, generate KPI
1. Select the product items to be forecast reports.
collaboratively.
2. Decide the time bucket of forecast will be one 3.3 The Benefits of Transforming the Demand
month. Forecasting and Planning Process
3. Decide the forecast frozen time period will be The main differences of the CMC demand
one month. forecasting and planning process between before and
4. Specify exception criteria for handling planning after developing CPFR are the ways to generate order
variances of demand forecasts between CMC and forecasts. Before developing CPFR, the order
the trading partner. forecasts were generated by CMC’s customer all
alone and then were given to CMC for reference.
Step 2: Each season, CMC and its trading partner However, after developing CPFR, the order forecasts
establish a joint business plan to identify planned are collaboratively generated by CMC and its
events such as openings, closings, promotions, customer, considering related information from both
holidays, advertisements, etc. These planned events sides such as historical sales data, future market
are the reference for that season’s demand forecasts. trends, CMC’s capacity, etc. Accordingly, the order
forecasts after developing CPFR are more accurate
Step 3: The first week of each month is the time for and result in CMC converting its production model.
both CMC and its trading partner to input order Before developing CPFR, CMC had two types
forecast. CMC’s customer forecasts order volumes of production models. In the slack season, on account
according to market information, inventories of of surplus capacity, CMC adopted Make-To-Stock
distribution centers, etc. CMC forecasts order (MTS) production model which results in a higher
volumes of the customer from customer long-term inventory level. But in rush season, on account of
demand planning and constrained capacity, etc. Then slack capacity, CMC adopted Build-To-Order (BTO)
J. T. Lin et al.: KPI with Data Flow Analysis for CPFR: A CMC Case Study 115

production model which results in a long order cycle facilitate operational and informational efficiency
time. (See figure 3 as illustration.) between business to business and then to fulfill
After developing CPFR and converting the customers demand on time. Collaborative commerce
demand forecast process, CMC can start planning model is one way to integrate business processes of
production schedule and transportation in the frozen both sides. CMC utilizes “CPFR Workplace” to
time period of order forecasts. Because of the higher improve efficiency and reduce inventory in supply
accuracy of order forecasts, this change of production chain. CMC “CPFR Workplace” provides
model leads to shorten order cycle times and collaborative mechanism of one business network
improves customer service levels. This means that framework. The buyer and seller can operate
CMC can adopt a more flexible production strategy to collaboratively by means of sharing information and
acquire better efficiency resulting from the higher joint business process. Moreover, the trading partners
accuracy of order forecasts. can exchange and integrate information to react
marketplace change rapidly.
Slack Season PUSH PULL The functions of CMC’s CPFR Workplace
system (See figure 5) are described as follows.
Rush Season PULL

raw
manufacture piece packing finish good shipping finish good
material (sku) (by sea) (sku)
Port near DC
of customers

Figure 3: CMC production model before developing

CPFR

PUSH PULL

raw manufacture piece packing finish good shipping finish good

material (sku) (by sea) (sku)
Port near DC
of customers Figure 5: Picture of CMC CPFR workplace
Figure 4: CMC production model after developing
CPFR Working groups: Manage the roles of buyer and seller
by defining collaborative team members and their
The CPFR process facilitates the accuracy of responsibilities.
order forecasts. However, to specifically evaluate the
benefits of developing CPFR, the improvement of Collaboration arrangement: Arrange collaboration
related KPIs should be measured. Therefore, this such as product items, areas, exception criteria, and
study will analyze the CPFR related information conditions of collaboration.
systems and then to design the KPI calculator
information flows. From the KPI calculator Joint business plan: Establish joint business plan
information flows, we will comprehend the influence collaboratively to identify event calendar (openings,
of each KPI and how to promote these performances. closings, promotions, holidays, advertisements, etc.).

Demand collaboration: Integrate demand planning

4. DESIGN KPI DATA FLOW and exchange forecasts of trading partners. This
CMC AS ILLUSTRATION function provides mechanism to collaboratively
forecast and then to generate common demand
When developing CPFR, trading partners need forecasts.
other information systems to support the CPFR
process. Therefore, KPI with data flow for CPFR Order collaboration: The demand forecast, on hand
should be designed according to the information inventory and shipping information are integrated to
system framework of both sides. This study takes collaboratively generate and exchange order forecast.
CMC as a case study, in the light of information This function includes the frozen mechanism of order
system framework of CMC and its trading partner, to forecast and order management.
design KPI data flow.
Exception list： Monitor and automatically trigger the
4.1 CMC CPFR System Framework and exceptions and then to give notices in order to make
Functions trading partners to control exceptions.
The objective of supply chain integration is to
116 International Journal of Electronic Business Management, Vol. 1, No. 3 (2003)

fulfill rate indicator.

KPI & report: Managers can analyze the key IV (Inventory Visibility): The IV system is to
performances and download reports in accordance visualize the global inventory of one company in
with each warehouse or each product item. order to control the inventory in different places and
trading partners. Therefore, the SV system can reduce
4.2 CPFR Related Information Systems supply chain inventory level and fulfill customer
When trading partners developing CPFR, in demand.
addition to CPFR system, they need other
information systems to support CPFR process. Take
4.3 Select KPIs according to the Developing
CMC as illustration, figure 6 shows the information
CPFR Process and Information Sharing
system framework of CMC and its trading partner.
CMC and its trading partner use KPIs to evaluate
On the customer of CMC side, the retailers provide
the performances of developing CPFR process.
sales data from POS (point of sales) systems. Then,
Different trading partners play various roles in supply
DRP (Distribute Requirement Planning) system plans
chain. Therefore, each partner expects to improve
the future events, sales forecasts, and order forecasts.
different performances. Following are the original ten
And then, the ERP system generates actual orders.
KPIs selected by CMC and its trading partner
On the other hand, the ERP system of CMC acquires
collaboratively.
collaborative agreement of order forecasts from
1. Sales Forecast Accuracy
CPFR system to plan production. Moreover, when
2. Order Forecast Accuracy
customer places actual orders, CMC schedules
3. DC Inventory Turn
production plan and confirms the order production
4. Retail Inventory Turn
time. Then, according to the production time of actual
5. Manufacturer Order Fill Rate
orders from ERP system, the SPE (Shipment
6. Delivery Punctually (on-time)
Planning & Execution) system plans the shipping
7. Delivery Punctually (in-full)
schedule of these actual orders. Following that, SPE
8. Order to manufacture cycle time
system transfers the shipping schedule to SV
9. DC in-stock Service Level
(shipping visibility) system, and then the LSP
10. Retail in-stock Service Level
(logistic service party) updates the shipping states.
However, because sales forecast generation, the
The following are the describing of SPE, SV and IV
third to fifth steps of CPFR process, is not included
systems of CMC.
the CPFR process developed by CMC and its trading
partner, they don’t share the related information such
CMC Buyer Retailer
as sales forecast, inventory information, actual sales .
Event Calendar Accordingly, several KPIs (e.g., sales forecast
Sales Forecast
Order Forecast CPFR DRP POS accuracy) cannot be calculated for lack of the related
SAP/R3 Order Forecast
information. Finally, CMC and its trading partner
Order
Order
Sales/Inventory
Information
select four KPIs (See table 1) to evaluate developing
Shipment LSP
Shipment
Visibility
Sales
Information CPFR process performances.
Planning& Shipment Update status ERP
Execution
Seller Inventory Inventory Buyer Inventory
4.4 Design of KPI Data Flow
Visibility KPIs are used to measure and track the overall
Inventory performance of the relationship between
system:
HUB/DC role: manufacturer and retailer. In the case of CMC and the
Figure 6: Information system framework trading partner, the seller CMC expects to acquire
more accurate order forecasts to early plan and
SPE (Shipment Planning & Execution): The SPE schedule the production and stabilize production. Its
system is designed for manufacturing exporters to trading partner hopes to reduce the order cycle time
visualize shipping planning and operation process. and improve delivery accuracy. Thus, after
This system provides shipping information collaborative agreement, both sides select four KPIs,
management, exportation documents management, order forecast accuracy, finished goods production
exception management, freight rate analysis, etc. lead-time, on time delivery, and order fill rate, to
SV (Shipment Visibility): The SV system is designed evaluate the benefits of developing CPFR. The
to manage goods in transit and to visualize the global frequency and timeframe for each KPI is monthly.
goods of one company in order to dynamically Table 1 shows the definitions and calculating
control the volumes of goods in different LSP and formulas of these four KPIs.
trading partners. Information visibility reduces “CPFR Workplace” system will automatically
uncertainty of the supply chain and can evaluate the calculate the values of KPIs periodically. However,
performance of LSP serving customers with order
J. T. Lin et al.: KPI with Data Flow Analysis for CPFR: A CMC Case Study 117

the following for each KPI should be specified. each month, the CPFR system diagnoses if exceptions
1. The data source (where does the data reside?) exist between the order forecasts of both sides. If
2. Level of measure (e.g., Stock Keeping Unit exceptions exist, the CPFR system automatically
(SKU), category, department, consumer) notifies both sides to solve the exceptions. In the third
3. Specific SKUs to be included in the CPFR week of each month, the forecasts should be frozen
process are frozen by CPFR system and the buyer starts to
4. Agreement on the definition of each measure and place actual orders according to the frozen forecasts.
the method of calculation The seller’s ERP system receives the actual orders
5. Agreement on the frequency and timeframes for and starts to plan the production schedule. And then,
each measure (e.g., monthly, weekly, daily, the production schedule is transferred to the SPE
hourly) system. In the CMC case, the CPFR system calculates
order forecast accuracy once a month by comparing
Table 1: KPI definitions and calculating formulas the actual order volumes from the SPE system with
Calculating the order forecasts from CPFR itself system. For
KPI Definition
formula example, at the beginning of September, the order
Order Order Forecast value 1-(|actual order value - forecasts of October are rolling into the frozen time
forecast in a defined time Order Forecast value| /
accuracy before actual order actual order period, and these frozen order forecasts can be
divided by the actual value)*100% acquired from the CPFR system. (See table 2.) When
order value the buyer places the actual orders in October, these
Finished Number of days it Seller ship goods
can be acquired from the SPE system. (See table 3. )
goods takes from buyer date-Buyer Delivery
production deliver order to seller order date Accordingly, the order forecast accuracy of each item
lead-time ship goods can be calculated as follows.
On time Number of orders (Number of orders 1. Item1: 1-(|550-600| / 550)x100% = 91%
delivery delivered on time delivered on time / the
divided by the total total number of 2. Item2: 1-(|280-400| / 280)x100% = 68.4%
number of orders orders)*100% 3. Item : 1-(|450-400| / 450)x100% = 89%
Order fill Quantity of goods Quantity of goods 4. Item4: 1-(|400-400| / 400)x100% = 100%
rate shipped on time shipped on time / the
divided by the total total quantity of
quantity of orders orders)*100% Table 2: Order forecasts in CPFR system
during a defined Order Oct. Nov. Dec. Jan. Feb. Mar.
period of time forecast
Item1 600 450 300
Item2 400 400 400
This study takes CMC as case study to Item3 400 400 300
illustrate how to design KPI calculator by considering Item4 400 400 400
when, which partner and what information system
does the related data reside. Table 3: Order related data in SPE system
Order Item Order Order release Planned shipping
numb Quantity date date
er
CPFR SPE P001 Item1 550 2003/10/1 2003/10/9
P002 Item2 280 2003/10/1 2003/10/10
P003 Item3 450 2003/10/1 2003/10/15
Quantity of Quantity of
P004 Item4 400 2003/10/1 2003/10/22
Order Forecast Order

(2) Finished Goods Production Lead-Time, on Time

Delivery and Order Fill Rate
KPI Order Forecast Accuracy Figures 8, 9 and 10 are the data processes of
Calculation = 1-(|actual order value in
SPE system-Order Forecast
calculating finished goods production lead-time, on
value in CPFR system|/ time delivery, and order fill rates. After the buyer
actual order value in SPE places actual orders, the seller inputs actual order-
Order Forecast
system)X100%
Accuracy related data into the ERP system and then to plan
production schedule. The outcomes of production
Figure 7: The data process of calculating schedule planning are transited to the SPE system.
order forecast accuracy The SPE system plans the shipping schedule
according to order due dates and then transits the
(1) Order Forecast Accuracy planning outcomes to the SV system. Finally, the
Figure 7 illustrates the data process of logistic service party (LSP) updates the actual
calculating order forecast accuracy. At the beginning shipping states in SV system. In CMC case, the CPFR
of a month, the buyer inputs order forecasts of two system gets the related data from the SV system to
month later, and then the seller replies the acceptable calculate KPIs of finished goods production lead-time,
volumes in the CPFR system. In the second week of
118 International Journal of Electronic Business Management, Vol. 1, No. 3 (2003)

on time delivery and order fill rates. KPI data flows are designed to calculate and
Table 4 is the order data in the SV system and evaluate performance improvement. Moreover, each
the following are the calculating of each KPI. KPI can exhibit related issues. For example, the order
1. Finished goods production lead-time of item 1 is forecast accuracy of item 1 is 91%, but of item 2 is
40 days, of item 2 is 40 days, of item 3 is 43 days 68%. This shows that the forecast method for item 1
and of item 4 is 50 days. can achieve its goal, but the forecast method of item
2. Ratio of on time delivery = 3 / 4 = 75% 2 should be modified by considering more factors.
3. Order fill rate = (550+280+400+400)/ KPI data flow analysis can be utilized to discover the
(550+280+450+400) = 97% causes that produce worse KPI values. Therefore, in
the initial stage, the buyer and seller have to select
SV appropriate KPIs cautiously to support performance
improvement.
Date of Date of
delivery order ship goods
Table 4: The order data in SV system
Order Planned Actual Order
Order Shipment Actual receipt Order Due
numbe shipping shipping release
Quantity Quantity date Date
r date date date
KPI Production lead-time P001 550 550 2003/10/9 2003/10/9 2003/11/10 2003/11/9 2003/10/1
=Date of ship goods in SV system- P002 280 280 2003/10/10 2003/10/9 2003/11/10 2003/11/10 2003/10/1
Calculation Date of delivery order in SV system P003 450 400 2003/10/15 2003/10/13 2003/11/13 2003/11/15 2003/10/1
P004 400 400 2003/10/22 2003/10/18 2003/11/20 2003/11/22 2003/10/1

Production
lead-time
5. CONCLUSION
Figure 8: The data process of calculating finished Many successful CPFR cases have been
goods production lead-time presented internationally. CMC project is an IT
Applications Promotion (ITAP) D Project supported
by the Department of Industrial Technology (DoIT),
SV Ministry of Economic Affairs ROC. This project has
demonstrated the nine steps of CPFR process to
Numbers of collaborate between partners with commitment. More
訂單準時
order Delivery
訂單運送
Total numbers
送達次數 of orders
總次數
accurate forecast information could be shared by
on time
collaborative partners. Therefore, CMC can perform
proper capacity planning in advance to reduce
KPI Delivery on time inventory levels and order cycle time, and to improve
=(Numbers of order Delivery on time
計算
Calculation in SV system/the total numbers of the order fill rate.
order in SV system)x100% CPFR process transfers CMC demand planning
process. CMC and its trading partner collaborate to
Delivery
on time
predict the forecast of product demand, and then to
facilitate the accuracy of demand forecast. Therefore,
Figure 9: The data process of calculating on time CMC can produce products according to demand
delivery forecast in order to reduce order cycle time and
inventory levels. Moreover, for buyer, accuracy of
demand forecast can reduce the sales loss resulting
from out of stock. However, to evaluate the benefits
SV of developing CPFR concretely, partners should
select appropriate KPI and then calculate the value of
each KPI.
Quantity of Total Quantity Besides CPFR system, when developing CPFR,
goods shipped of orders trading partners need other information systems to
support CPFR process. Therefore, KPI with data flow
for CPFR should be designed according to
KPI Order Fill Rate
=(Quantity of goods shipped in information system frameworks of both sides. This
Calculation SV system/the total Quantity of study takes CMC as case study, in the light of
orders in SV system)x100%
information system frameworks of CMC and its
trading partner, to design KPI data flow. Four KPI,
Order
Fill Rate order forecast accuracy, finished goods production

Figure 10: The data process of calculating order fill rate

J. T. Lin et al.: KPI with Data Flow Analysis for CPFR: A CMC Case Study 119

lead-time, on time delivery and order fill rate are as Association.

illustration. Design of KPI data flow can evaluate the 4. Holmstrom J., Framling, K., Kaipia R. and
benefits of developing CPFR concretely and then to Saranen, J, 2002, “Collaborative planning
understand the importance of each KPI to different forecasting and replenishment: new solutions
roles in supply chain. Moreover, KPI data flow can needed for mass collaboration,” Supply Chain
indicate the performance improvement to facilitate Management: An International Journal, pp.
efficiency in supply chain. 136-145.
5. Lu, S. N. and Zoe, K. L., 2002, Supply chain
management book, Taipei: Shang Zhou.
REFERENCES 6. Roadmap to CPFR, 2002, Voluntary Inter
industry Commerce Standards Association.
1. A Guide to CPFR Implementation, 2001, ECR 7. Wei, Z. Q., 2001, CPFR-Wal-Mart and Sara
Europe project team. Lee: The Demonstration Project of
2. Barratt M., and Oliveira, A., 2001, collaborative commerce model, Department of
“Experiences of collaborative planning Commerce in Ministry of Economic Affairs
initiatives,” International Journal of Physical R.O.C.
Distribution & Logistics Management, pp. 8. Zeng, S. H., 2002, Stride forward new
266-289. generation of compact disk industry, Industrial
3. Collaborative Planning, Forecasting and Economics & Knowledge Center of Industrial
Replenishment (CPFR®) Vision 2.0, 2002, Technology Research Institute.
Voluntary Interindustry Commerce Standards