Voice of The Customer - Capture and Analysis

Voice of the Customer: Capture and Analysis
McGraw-Hills
Other Books inBiomedical
the Six Sigma
Engineering
Operational
Series
Methods Series
Segregation

Parveen S. Goel, Rajeev Jain, and Praveen Gupta Six Sigma for
HJORTS Population Balances in Biomedical Engineering:
Transactions and Service
through the Distribution of Cell States
PANCHAPAKESAN
Praveen Gupta TheBiomedical
Six SigmaNanotechnology
Performance Handbook
PETERS
Real-Time Biomolecular Simulations
Thomas McCarty, Lorraine Daniels, Michael Bremer, and
SARKODIE-GYAN
Praveen Gupta Neurorehabilitation
The Six Sigma Black
Design, Measurement, and Control
Devices:
BeltEngineering
Handbook
WAITE
Alastair Muir Lean
Biofluid Six Sigma
Mechanics Statistics Systems
in Cardiovascular
Andrew Sleeper Design for Six Sigma
Kai Yang Design for Six Sigma for Service
Voice of the Customer:
Capture and Analysis
Dr. Kai Yang
New York Chicago San Francisco Lisbon London Madrid

Mexico City Milan New Delhi San Juan Seoul
Singapore Sydney Toronto
Copyright 2008 by The McGraw-Hill Companies. All rights reserved. Manufactured in the United
States of America. Except as permitted under the United States Copyright Act of 1976, no part of this
publication may be reproduced or distributed in any form or by any means, or stored in a
database or retrieval system, without the prior written permission of the publisher.
0-07-159341-1
The material in this eBook also appears in the print version of this title: 0-07-146544-8.
All trademarks are trademarks of their respective owners. Rather than put a trademark symbol after
every occurrence of a trademarked name, we use names in an editorial fashion only, and to the
benefit of the trademark owner, with no intention of infringement of the trademark. Where such
designations appear in this book, they have been printed with initial caps.
McGraw-Hill eBooks are available at special quantity discounts to use as premiums and sales
promotions, or for use in corporate training programs. For more information, please contact George
Hoare, Special Sales, at [email protected] or (212) 904-4069.
TERMS OF USE
This is a copyrighted work and The McGraw-Hill Companies, Inc. (McGraw-Hill) and its licensors
reserve all rights in and to the work. Use of this work is subject to these terms. Except as permitted
under the Copyright Act of 1976 and the right to store and retrieve one copy of the work, you may not
decompile, disassemble, reverse engineer, reproduce, modify, create derivative works based upon,
transmit, distribute, disseminate, sell, publish or sublicense the work or any part of it without
McGraw-Hills prior consent. You may use the work for your own noncommercial and personal use;
any other use of the work is strictly prohibited. Your right to use the work may be terminated if you
fail to comply with these terms.
THE WORK IS PROVIDED AS IS. McGRAW-HILL AND ITS LICENSORS MAKE NO

GUARANTEES OR WARRANTIES AS TO THE ACCURACY, ADEQUACY OR COMPLETE-
NESS OF OR RESULTS TO BE OBTAINED FROM USING THE WORK, INCLUDING ANY
INFORMATION THAT CAN BE ACCESSED THROUGH THE WORK VIA HYPERLINK OR
OTHERWISE, AND EXPRESSLY DISCLAIM ANY WARRANTY, EXPRESS OR IMPLIED,
INCLUDING BUT NOT LIMITED TO IMPLIED WARRANTIES OF MERCHANTABILITY OR
FITNESS FOR A PARTICULAR PURPOSE. McGraw-Hill and its licensors do not warrant or
guarantee that the functions contained in the work will meet your requirements or that its operation
will be uninterrupted or error free. Neither McGraw-Hill nor its licensors shall be liable to you or
anyone else for any inaccuracy, error or omission, regardless of cause, in the work or for any damages
resulting therefrom. McGraw-Hill has no responsibility for the content of any information accessed
through the work. Under no circumstances shall McGraw-Hill and/or its licensors be liable for any
indirect, incidental, special, punitive, consequential or similar damages that result from the use of or
inability to use the work, even if any of them has been advised of the possibility of such damages. This
limitation of liability shall apply to any claim or cause whatsoever whether such claim or cause
arises in contract, tort or otherwise.
DOI: 10.1036/0071465448
Professional
Want to learn more?

We hope you enjoy this
McGraw-Hill eBook! If
youd like more information about this book,
its author, or related books and websites,
please click here.
ABOUT THE AUTHOR
Kai Yang, Ph.D., has wide experience in quality and reliability engineering.
The Executive Director of Enterprise Excellence Institute, a renowned quality
engineering organization based in West Bloomfield, Michigan, he is co-author of
the influential Design for Six Sigma: A Roadmap for Product Development. He
is also Professor of Industrial and Manufacturing Engineering at Wayne State
University, Detroit.
Copyright 2008 by The McGraw-Hill Companies. Click here for terms of use.
This page intentionally left blank
For more information about this title, click here
Contents
Chapter 1. Value, Innovation, and the Voice of the Customer 1

1.1 Defining Customer Value 2
1.2 Innovation Roadmap 5
1.3 Voice of the Customer: Mining for the Gold 6
1.4 Overview of This Book 8
Chapter 2. The Product Development Process 9

2.1 Defining Product Cost and Development 9
2.1.1 Product Development Process Flowchart 13
2.2 The Product Development ProcessEnd to End 14
2.2.1 Opportunity Identification and Idea Generation: Stage 0 14
2.2.2 Customer and Business Requirement Study: Stage 1 17
2.2.3 Concept Development: Stage 2 18
2.2.4 Product Design and Prototype: Stage 3 25
2.3 The Nature of Product Development: Information and
Knowledge Creation 33
2.3.1 Axiomatic Design 35
2.3.2 Design as an Information Production Factory 43
2.3.3 Information and Knowledge Mining 45
2.3.4 Information Transformation 53
2.3.5 Information and Knowledge Creation 56
2.3.6 The Ideal Product Development Process 60
2.4 Customer-ValueBased Lean Product Development Process 64
2.4.1 Lean Operation Principles 64
2.4.2 Waste Elimination in Process 65
2.4.3 Value-Stream Mapping 66
2.4.4 One-Piece Flow 68
2.4.5 Pull-Based Production 70
2.4.6 Lean Principles for Product Development 71
2.4.7 Mining the Voice of the Customer to Capture Value 74
2.4.8 Maximizing Technical Competence 75
2.4.9 Front-Loading the Product Development Process 76
2.4.10 Optimizing Information Transformation and Flow 78
2.4.11 Creating a Lean Product 86
vii
viii Contents
Chapter 3. Customer Value and the Voice of the Customer 89

3.1 Customer Value and Its Elements 91
3.1.1 Value and Other Commonly Used Metrics 94
3.1.2 The Versatility and Dynamics of Value 95
3.2 Customer Value Analysis 97
3.2.1 Market-Perceived Quality Profile 98
3.2.2 Market-Perceived Price Profile 102
3.2.3 Customer Value Map 104
3.2.4 Competitive Customer Value Analysis 107
3.3 Customer Value Deployment 109
3.4 Evolution of Customer ValuesBlue Ocean Strategy 111
3.4.1 Formulating a Blue Ocean Strategy 116
3.5 Customer Value and the Voice of the Customer 123
3.6 Capturing the Voice of the Customer 126
3.6.1 Plan for Capturing the Voice of the Customer 127
Chapter 4. Customer Survey Design, Administration, and Analysis 133

4.1 Customer Survey Types 133
4.1.1 Mail-Out Surveys 134
4.1.2 In-Person Interviews 134
4.1.3 Telephone Surveys 135
4.1.4 Other Methods of Gathering Information 135
4.2 Stages of the Customer Survey 136
4.2.1 Stage 1: Establish Goals and Objectives of the Survey 137
4.2.2 Stage 2: Set the Survey Schedule and Budget 138
4.2.3 Stage 3: Establish an Information Base 138
4.2.4 Stage 4: Determine the Population and Sampling Frame 139
4.2.5 Stage 5: Determine Sample Size and Selection Procedure 139
4.2.6 Stage 6: Design the Survey Instrument 139
4.2.7 Stage 7: Pretest the Survey Instrument 140
4.2.8 Stage 8: Select and Train Survey Interviewers 140
4.2.9 Stage 9: Implement the Survey 140
4.2.10 Stage 10: Analyze the Data and Report 140
4.3 Survey Instrument Design 141
4.3.1 Close-Ended Questions 141
4.3.2 Open-Ended Questions 144
4.3.3 The Wording of Survey Questions 144
4.3.4 Order of Questions in Surveys 145
4.3.5 Questionnaire Length 147
4.4 Administering the Survey 147
4.4.1 Administering Mail-Out Surveys 147
4.4.2 Administering Telephone Surveys 148
4.4.3 Administering In-Person Surveys 148
4.5 Survey Sampling Method and Sample Size 148
4.5.1 Population and Sampling Frame 149
4.5.2 Sampling Methods 150
4.5.3 Sample Size Determination 153
4.6 Internet Surveys 157
4.6.1 Drawing People to the Internet-Based Survey 158
Contents ix
4.6.2 Administering a Survey on the Internet 162

4.6.3 Comparing Paper-Based Surveys with Internet Surveys 166
Chapter 5. Proactive Customer Information Gathering

Ethnographic Methods 169
5.1 What Are Ethnographic Methods? 171
5.1.1 Frequently Used Ethnographic Methods 173
5.1.2 Data Recording Methods 174
5.1.3 Types of Ethnographic Research Used in Product Development 175
5.1.4 Key Winning Factors for Ethnographic Methods 177
5.2 Ethnographic Research Project Planning 178
5.2.1 Determining Research Objectives 178
5.2.2 Recruiting Informants 180
5.2.3 Selecting Research and Data Collection Methods 185
5.2.4 Developing the Ethnographic Research Team and
Ground Rules 189
5.3 Ethnographic Project Execution 191
5.3.1 Ethnographic Interviews and Documentation 191
5.3.2 Ethnographic Observations in Shops 196
5.3.3 Ethnographic Observations in Product Usage Processes 198
5.3.4 Ethnographic Studies of Customer Cultures 203
Chapter 6. VOC Data Processing 213

6.1 Types of VOC Data 213
6.2 Analyzing VOC Data 214
6.2.1 Methods of Analyzing VOC Data 214
6.2.2 Affinity DiagramKJ Method 214
6.3 Quantitative VOC Data Analysis 221
6.4 Critical-to-Quality Characteristics (CTQ) 223
Chapter 7. Quality Function Deployment (QFD) 229

7.1 History of QFD 231
7.2 QFD Benefits, Requirements, and Practicalities 231
7.3 QFD Methodology Overview 232
7.3.1 Customer Attributes (WHATs) 234
7.3.2 CTSs (HOWs) 235
7.3.3 Relationship Matrix 235
7.3.4 Importance Ratings 237
7.3.5 Planning Matrix 237
7.3.6 CTS Correlation (HOWs Correlation) 238
7.3.7 Targets and Limits (HOW MUCH) 238
7.3.8 Competitive Benchmarks 239
7.4 Kano Model of Quality 239
7.5 QFD Analysis 240
7.6 Example 7.1 Information System Design 241
7.6.1 Ranking Customer Input 241
7.6.2 Ranking the Functional Requirements 243
7.7 QFD Case Study 1: Global Commercial Process Design 244
7.7.1 QFD Steps 245
7.7.2 The HOWs Importance Calculation 248
7.7.3 Phase I QFD Diagnostics 249
x Contents
7.8 QFD Case Study 2: Yaesu Book Center 251

7.8.1 Determine Customer Attributes (WHATs) 254
7.8.2 Determine Quality Characteristics (HOWs) 255
7.8.3 Assign Degree of Importance to Customer Attributes 255
7.8.4 Determine Operations Items 255
7.8.5 Two-Phase QFD Analysis for Yaesu Book Center 255
Chapter 8. Customer Value Creation by Brand Development 259

8.1 The Anatomy of Brands 262
8.1.1 Peoples Buying Behavior and Brands 262
8.1.2 Brand Identity 264
8.1.3 Brand Equity 275
8.2 Brand Development 277
8.2.1 Key Factors in Brand Development 278
8.2.2 The Brand Development Process 281
8.2.3 Strategic Brand Analysis 283
8.2.4 Brand Strategy Development 287
8.2.5 Brand Implementation 293
8.2.6 Brand Evaluation 293
Chapter 9. Value Engineering 297

9.1 An Overview of Value Engineering 298
9.1.1 Collecting Information and Creating Design Alternatives 298
9.1.2 Evaluating, Planning, Reporting, and Implementing 300
9.1.3 The Job Plan 300
9.2 Information Phase 300
9.2.1 Information Development 301
9.2.2 Function Determination 306
9.2.3 Function Analysis and Evaluation 314
9.3 Creative Phase 328
9.3.1 Brainstorming 328
9.4 Evaluation Phase 329
9.4.1 Relatively Simple Evaluations 330
9.4.2 More Complex Evaluations 331
9.4.3 Selection and Screening Techniques 332
9.5 Planning Phase 335
9.6 Reporting Phase 336
9.7 Implementation Phase 337
9.7.1 Setting a Goal 338
9.7.2 Develop An Implementation Plan 338
9.8 Automobile Dealership Construction (Park 1999) 339
9.9 Engineering Department Organization Analysis (Park 1999) 341
Chapter 10. Customer Value Creation Through Creative Design (TRIZ) 345
10.1 Theory of Inventive Problem Solving (TRIZ) 345
10.1.1 What Is TRIZ? 347
10.2 TRIZ Fundamentals 349
10.2.1 Function Modeling and Analysis 349
10.2.2 Use of Resources 353
Contents xi
10.2.3 Ideality 354

10.2.4 Contradictions 355
10.2.5 Evolution 357
10.3 The TRIZ Problem-Solving Process 360
10.3.1 Problem Definition 360
10.3.2 Problem Classification and Tool Selection 362
10.3.3 Problem-Solution Generation 363
10.3.4 Problem Concept Evaluation 363
10.4 Technical Contradiction Elimination and Inventive Principles 363
Chapter 11. Statistical Basics and Six Sigma Metrics 379

11.1 Six Sigma and Data Analysis 379
11.2 Descriptive Statistics 380
11.2.1 Dot Plot 380
11.2.2 Histogram 380
11.2.3 Box Plot 382
11.2.4 Numerical Descriptive Statistics 382
11.3 Random Variables and Probability Distributions 386
11.3.1 Discrete and Continuous Random Variables 386
11.3.2 Expected Values, Variance, and Standard Deviation 387
11.3.3 Probability Distribution Models 388
11.3.4 Statistical Parameter Estimation 391
11.4 Quality Measures and Six Sigma Metrics 392
11.4.1 Process Performance 392
11.4.2 Process Capability Indices 393
11.4.3 Sigma Quality Level (Without Mean Shift) 395
11.4.4 Sigma Quality Level (With Mean Shift) 396
References 399
Index 403
Chapter
Value, Innovation, and the

1
Voice of the Customer
In todays global economy, several business functions have become global

commodities and subject to stiff price competition: first manufacturing
activities, then IT, and most recently traditional R&D, such as routine
engineering design work. To find success in this competitive reality, your
business needs to take the high ground in value creation. Value is a mea-
sure of how much the customer really appreciates a product or service,
and how much customers are willing to purchase this product or service.
In Figure 1.1, you can see that there are some legendary products or
services, such as Hollywood movies, Intel CPUs, and Microsoft operat-
ing systems, that solidly command the market placethese products
creates enormous profits for their companies.
Figure 1.1 reveals the secrets of how these products and companies
create value. There are two success models for companies to create high
values. One type of company has a commanding lead either in technology
or brand recognition. Their products or brands dominate the market-
place and become industry standards. Examples of this kind of company
include Microsoft, Intel, Cisco, Google, and so on. The driver for value cre-
ation for this type of company is technical or brand dominance, or tech-
nology-driven innovation. The other type of company develops products
or services that capture the heart of customers; examples of this kind of
companies or products include Starbucks and Apples iPod. The driver
for value creation for this type of company is customer-centric innova-
tion. However, these two types of companies or products are few; most
companies are mediocre in both technical or brand dominance and cus-
tomer value position. However, in order to survive in the global economy
of the 21st century, it is wise for a company to excel in at least one of the
above two aspects; that is, either the technical/ brand dominance, or the
2 Chapter One
Value level Value level

Innovative
Technical/brand companies Customer-centric
dominance innovation
Most
Microsoft operation companies Starbucks
systems
iPod Nano
Intel CPUs
Figure 1.1 Value and innovation
customer-centric innovation. In both aspects, thorough understanding

of voice of the customer is very important. The main topic of this book is
about the voice of the customer (VOC). I will discuss thoroughly the role
of voice of the customer in these two success models.
1.1 Defining Customer Value

In the two aforementioned successful models, the real key for success is
that the product or brand brings exceptional value to customers, so that
many customers crave the product and are willing to pay good prices.
In this subsection, the following issues are discussed:
What is value and how do you create it? What is the role of value for long
term commercial success?
Profitability is one of the most important factors for long term com-
mercial success. High profitability is determined by strong sales and
overall low cost in the whole enterprise operation. It is common sense
that:
Business profit = Revenue Cost
In addition:
Revenue = Sales volume Price
Here, price means sustainable price, that is, the price level that
customers are willing to pay. Many researchers (Sheridan 1994, Gale
1994) have found that both sales volume and sustainable price are
mostly determined by customer valuecustomers opinions determine
Value, Innovation, and the Voice of the Customer 3
a products fate. Customers opinions will decide the price level, the size
of the market, and the future trend of this product family. A product that
has high customer value is often featured by increasing market share,
increasing customer enthusiasm towards the product, word of mouth
praise, reasonable price, and a healthy profit margin for the company
that produces it, as well as increasing name recognition.
Sherden (Sherden 1994) and Gale (Gale 1994) provided a good defini-
tion for customer value. In their assessment, customer value is defined
as perceived benefit (benefits) minus perceived cost (liabilities):
Customer Value = Benefits Liabilities
Here the benefits are the factors that increase the customer value,
and the liabilities are the factors that decrease the customer value.
Benefits in this equation include the following categories:
Functional benefits This category includes the actual benefits of
a product or service delivered to customers. In other words, functional
benefits are what the product or service does for customers. For example,
the functional benefits of a food item include providing energy, provid-
ing nutrition, providing taste and so on. For investment services, the
functional benefits include economic benefits and revenue to customers.
The reliability, quality and durability of the product are also part of
functional benefits. For example, a kitchen knife delivers the function of
cutting food. In comparing a high quality knife with a low quality knife,
both of them can cut food, but the high quality knife can cut better, last
longer, so it will deliver more functional benefits.
Psychological benefits In addition to functional benefits, psy-
chological benefits are a very important component for a product or
service. A very obvious example is that a plain color T-shirt will sell at
much lower price than that of a T-shirt with a famous logo. A famous
logo will give customers emotional and self-expression benefits on top
of the regular T-shirts functional benefit, covering body. Brand image
is also a part of psychological benefit. A known good brand brings con-
fidence to customers about products quality, reliability and durabil-
ity. A famous brand may even raise customers social status, such as
a Mercedes Benz car. If a product is the first of its kind, it usually
brings psychological awe to customers, such as the first copy machine. A
first-of-its-kind product will usually command a high price and often
create a brand name. Competitions in the market place also brings
psychological effects into customers value judgments, if there are many
competitors producing the same or similar a product, that usually will
create a perception that this product is a commodity and it doesnt carry
much value.
4 Chapter One
Service and convenience benefits This category includes avail-

ability, which is the ease of accessing the product or service; it also
includes service, which is the ease of getting help in case of product
problems or failure.
Liabilities in this equation include the following categories:
Economic liabilities This category includes all monetary expenses

incurred for owning the product or service. The price that a customer
pays in order to buy the product or service is certainly a major part
of economic liabilities. But, in addition to the price of the product or
service, there are many other indirect ownership costs associated with
owning this product or service, which include:
Acquisition costs: This includes transportation cost, shipping cost,
time and efforts spent to obtain the service
Usage costs: This includes additional cost to use the product or service
in addition to the purchase price, such as installation, training cost,
and so on.
Maintenance costs: This includes repair cost, routine maintenance
costs, regular upgrades and so on.
Ownership costs: This includes financing cost, licensing cost and so on.
Disposal costs: This includes disposal cost of hardware, environmental
regulation compliance cost and so on.
Psychological liabilities This category includes the negative psy-
chological effects of the product or service. An unknown brand name
may make a customer feel uncertain about the dependability of the
product or service. A cheap brand may cause self-esteem liability for
some customers. A poorly performed product or service not only delivers
low functional benefits, but will also make the customer feels bad.
Service and convenience liability This category includes all the
negative effects related to service and convenience. Even if the original
product is very good, lack of service, or poor service will use customer
issues when the product encounters problems. For example, lack of
repair facilities, high cost of spare parts and high repair cost are among
the major concerns for some potential buyers of foreign made cars. Poor
availability, such as long delivery time, is also an example of service and
convenience liability.
This customer value definition is comprehensive, and explains what
types of product or services that customers are willing to pay a premium
price to buy.
There are numerous cases where customers are willing to pay a higher
price to buy a product with better brand name image, because brand
name image is a psychological benefit for the customer. Toyota and
General Motors have a joint venture in California that produces an

identical car model. Some of the units have a Toyota brand, and some
have a GM brand; however, the units with the Toyota brand can sell for
a few hundreds dollars more than the identical cars with the GM brand,
because the public perception is that Toyota has the preferable brand
image. Brand name image is an important portion of customer value.
As another example, a neighborhood store will sell an item at higher
prices than the identical item in discount chain stores. This is because
of the perceived convenience in obtaining these items in neighborhood
stores. This is a part of the service and convenience benefit defined in
customer value.
Therefore, to gain business profit, creating products with high customer
value is a must. Many business enterprises often fail to see the multiple
aspects of customer value. They may create a product with tremendous
functionality, but one that is very poor in customer service, accessibility,
and psychological aspects, and therefore, the product will fail.
In a competitive marketplace, success will become more difficult to
achieve. Your competitors can learn from and copy your product, learn
your customer value proposition, and ultimately offer a similar product
at a lower price. There are also disruptive innovations, which change
the whole landscape of the competitive situation. Thus, achieving a
high value position is becoming more and more a moving target. For
example, video renting stores, such as Blockbuster, were everywhere in
the United States and their business was very good. With the emergence
of Internet downloading and Internet-based video rental businesses,
such as Netflix, video rental stores lost a lot of their attractions to cus-
tomers. Blockbuster has suffered huge losses in recent years. Another
example is the sport utility vehicles (SUVs) made by American automo-
bile manufacturers. SUVs helped GM, Ford, and Chrysler a lot in their
revenue and profitability in 1990s and early 2000s. However, with the
drastic improvement of fuel efficiency of Toyota vehicles, such as Prius,
and a surge of gasoline prices in the years 2005 to 2006, Ford and GM
lost a big chunk of the automobile market share due to the consumers
abandonment of SUVs and Trucks.
In any case, the key to value creation is the trend-setting innovation,
either technology-driven innovation or a customer-centric innovation,
because innovation will change the rules of game, and if you do it right,
it will put you on the top of the competition. If you are a business leader,
then you have to learn how to lead innovation. You will need to lead this
kind of innovation many times in your business tenure.
1.2 Innovation Roadmap

What is innovation? What are the key factors for successful innovation?
Based on Amabile et al (1996), All innovation begins with creative ideas . . .
6 Chapter One
We define innovation as the successful implementation of creative ideas

within an organization. In this view, creativity by individuals and teams
is a starting point for innovation; the first is necessary but not sufficient
condition for the second.
Specifically, innovation has two aspects. One is creativity; better cre-
ativity generates out-of-the-ordinary ideas to create first-of-its-kind
products. Another aspect of creativity is to make new ideas into com-
mercial successes.
In developing innovative products or services, there are two kinds of
driving forces. The first is the technology push; the phenomenal devel-
opment of the IT industry from the 1990s up to now started with new
technological breakthroughstake Microsoft and Google for example.
The other driving force is customer or market pull; this innovation is
usually started by discovering a hidden market need. Starbucks and the
iPod are such examples. Starbucks is not simply a regular coffee shop;
its vision is to become someplace other than home and the workplace.
It not only provides coffee, but also provides a casual meeting places,
free electrical outlets, and wireless Internet access.
The driver for technological push innovation is creativity and the abil-
ity to generate new ideas, and connecting this innovation with customer
and market needs. The driver for customer-centric innovation is the
ability to discover a hidden market. You need great vision to discover
some hidden unmet customer need and accurately identify a customer
value proposition, that is, what things customers really crave. In both
technology push and market pull innovative product development sce-
narios, capturing the voices of customers, especially the unarticulated
voice of the customers, is really a key factor. A creative mind may bring
new ideas, but a new idea alone is not enough; the new idea needs to
catch customers hearts. Without knowing the real voice of the customer,
you cannot catch the customers heart, so you cannot succeed.
Another key factor for successful innovations is an effective product
development process. You need an effective product development process
to make creative ideas and customer-centric innovation into products
at low cost and with high quality.
Finally, innovation is not only a product and technology matter.
Business model, brand strategy, services, and so on are all parts of
a grand road map for innovation.
1.3 Voice of the Customer:

Mining for the Gold
In the previous section, I discussed the key factors for value-based inno-
vation. Clearly, capturing the voice of the customer is a very important
factor, whether you are dealing with a technology push-or-market pull

oriented product or service development.
If you had some magic power and were able to discover exactly what
customers are craving, and if you also knew how to produce their
dream product at a low price, then you would be guaranteed to get
rich! Therefore, capturing the exact voice of the customer is like strik-
ing gold.
Of course, this kind of lucky chance is very rare. In the natural world,
you need to mine for gold. You need to explore and search. Sometimes
you find bits and pieces of gold mixed into other minerals, which you
then need to distill and purify to get pure gold. The same is true for the
voice of customer. You need to search for a good source of customer infor-
mation by finding bits and pieces. The real voice of customer may hide
deep in the customers mind, and many customers are humble people.
If you only use traditional customer data collection methods, you may
only get inaccurate and incomplete information. It takes a lot of effort
and sophisticated methods to mine the voice of the customer accurately
and distill this VOC information into valuable inputs for the product
development process.
What strategy and methods for voice of customer capturing and anal-
ysis that we use depends very much on what this VOC information is
used for. The purposes for the voice of customer information capturing
and analysis are the following:
You need the right kinds and sufficient amount of accurate voice of
customer information to provide necessary inputs for all stages and
levels of the product design work. The stages include product design
and manufacturing process design, and the levels refer to the system
level, subsystem level, and component level.
You need a good set of VOC information to learn what are the key
customer value factors for this kind of product; what factors, such as
price and functions or really excite customers; and what the products
customer value position is.
You may need to explore the possibility of shifting your current cus-
tomer value proposition to another one, so you can develop a new
innovative product that is different from competitors. In this case,
you need to capture enough VOC information for decision making.
We may want to capture the right kinds and a sufficient amount of voice
of customer information to provide necessary inputs for product improve-
ments in limited scope, instead of a complete new product design.
This book will show you how to capture and analyze the voice of the
customer in all these four scenarios.
8 Chapter One
1.4 Overview of This Book

Since VOC capturing and analysis is a very important part of the product
development process, we will not cover the process of VOC capturing and
analysis in isolation. Chapter 2 discusses all aspects of product devel-
opment processes. Ill cover some state-of-the-art product development
theories and advanced product development processes. During this dis-
cussion, the roles of VOC capturing and analysis in the product develop-
ment process are thoroughly discussed. After reading this chapter, you
should have a good idea of how voice of customer information can help
your product development process, and what kinds and what amount of
VOC information are really needed for this.
Chapter 3 discusses issues related to the customer value. Because the
customer value position determines the market position of the product,
you need to know about the relationship between VOC and customer
value, and how to obtain key customer value information from VOC.
I will also discuss how a customer value position can be modified to
create a new product market position; this strategy is called the blue
ocean strategy (Kim and Mauborgne 2005). Chapter 3 also discusses
how to link VOC information to product design specifics in a very clear
and exact manner.
Chapter 4 discusses conventional VOC capturing methods, including
customer surveys, interviews, and Internet surveys, in great detail.
Chapter 5 discusses an anthropology-based VOC capturing method, the
ethnographic method. Because this VOC capturing method can capture
hidden, unarticulated VOC information much better than conventional
VOC capturing methods, it is becoming very popular and many customer-
centric innovation practices are heavily relying on the ethnographic
method. In this chapter, I give very detailed descriptions of the method
and provide several examples and case studies.
Chapter 6 explains how to process raw voice of customer data and
transform them into clearly defined customer data. Chapter 7 discusses
the method of quality function deployment (QFD). The QFD method is
a systematic method to transform VOC data into product functional
requirements, and then design specifications. This method serves as the
interface between VOC data and the product design process.
Customer value creation and improvement is related to other meth-
odologies, such as brand development (Chapter 8), value engineering
(Chapter 9), and the theory of inventive problem solving (Chapter 10).
These methodologies are thoroughly discussed in the relevant chapters.
The last chapter, Chapter 11, provides some necessary background in
statistics.
Chapter
The Product Development

2
Process
This book is about how to capture and analyze the voice of the customer.
However, the major purpose of capturing and analyzing the voice of the
customer is to provide vital input for the product development process.
In order to derive the best approach to capturing and analyzing the voice
of the customer, you need to know how the product development process
really works. This chapter will provide you with that information. Ill
introduce the key stages in production development and then discuss
the first four stages in detail. Ill then discuss the nature of the product
development process from an information perspective. Youll learn about
the development and evolution of the theories and best practices for the
product development process. Youll discover the leading framework for
the product development process, called the lean product development
process. Finally, Ill discuss the relationship between the voice of the
customer and the product development process.
2.1 Defining Product Cost and Development

A product is anything that can be offered to a market that might sat-
isfy a want or need. A product is one of two types: tangible (physical) or
intangible (nonphysical). Tangible products are what most people think
products are; examples of tangible products include bicycles, laptop
computers, printer paper, cars, and airplanes. Intangible products are
related to service industries, for example, vacation packages, insurance
policies, and medical treatment.
Every product is sold in the market for a price. For good products, the
customers are willing to pay higher prices. The prices that customers are
willing to pay depend on a supply-and-demand relationship. How much
10 Chapter Two
the customers will demand a certain product depends on the value of the
product. Based on the research of Sherden (1994) and Gale (1994), the
value of the product can be expressed as the following equation:
Value = Benets Liabilities

Because the customers are really buying the benefits offered by a
product, not just the physical entity of the product, in a real marketplace,
the product itself, or generic product, could be bundled with everything
that is needed to deliver the benefits to customers. This bundle is called
the whole product. The whole product typically augments the generic
product with training and support, manuals, cables, additional software,
online help, warranty, installation instructions, professional services, and
so on. The concept of the whole product is illustrated in Figure 2.1.
As this analysis shows, customers demand products purely for their
benefits. The liabilities, such as the prices they have to pay to buy the
products, are necessary evils that customers have to live with. For a
product, if the gap between benefits and liabilities is large, then custom-
ers will perceive that the benefits they get from the product are much
larger than the liabilities they have to accept, and the product will sell
quickly and easily. Besides the benefits offered by the product, market
competition can also greatly affect the selling of the product. The supply-
and-demand equation is very different in a competitive environment,
with one supplier versus multiple suppliers. Competition will increase
the supply of the same or similar products, which will give customers
more choices. Competitors can also offer products with the same or better
values for their products, by offering more benefits, having lower liabili-
ties (mostly lower prices), or both. The product with a better value will
take more market share. When competition in the marketplace becomes
fierce, one common tactic is to reduce the selling price of the product in
order to improve the products value position in the marketplace and
Procedures Installation
Generic
product
Support Training
Additional
software
Figure 2.1 The Whole Product
The Product Development Process 11
Products
Core Operation
Impetus Concept
Design Production Sales/service
Ideation development
BUSINESS PROCESSES
BUSINESS MANAGEMENT
SUPPLIER MANAGEMENT
INFORMATION TECHNOLOGY
Figure 2.2 Business Operation of Product Producing Companies
keep the market share. However, there is a limit to how much you can
lower the pricethe limit is the cost of providing the product, because
the selling price has to be higher than cost in order to make a profit.
Figure 2.2 illustrates how a typical product is produced. There is
always a core operation, which consists of the product development
and the actual production of the product. There are also other business
operations, however, such as marketing, finance, personnel, and so on.
From Figure 2.2, you can see that the cost of providing a product has
the following components:
Cost of product development

Cost of production
Cost of running supporting operations
Besides these three cost components, the success of the product devel-
opment and production system also depends on other factors, such as
the quality and value of the product, the time to market, and so on. The
economic model for a product development and production system is
illustrated in Figure 2.3.
The ultimate goal for any product development and production system
is to make a profit, and the profit is the output of the system. There are
three important inputs in the system:
Cost The total cost includes the product development cost, produc-
tion cost, and the cost of running supporting operations. Lower costs
lead to higher prot.
12 Chapter Two
Cost
Product Development Profit
Product Performance And Production
System
Time
Figure 2.3 The Economic Model of Product Development and

Production System
Product Performance Product performance depends on how much

benet the product delivers to the customers, and how well it does so.
Product performance further depends on the following factors:
How well the product captures the hearts and minds of customers
How well the technology and engineering work is conducted in the
design process to deliver the performance that the customer wants
The quality and reliability of the product
The delivery and service of the whole product
Time The product should be introduced to the market with the right
timing. In many cases this time factor depends on how quickly and ex-
ibly the company can introduce a new product into the marketplace. If
the market demands a product, and you are the rst one to deliver this
product, you will have a dominant position in the marketplace.
Based on this analysis, you can see that the total cost, product perfor-
mance, and time are three dominant factors in the product development
process. We can derive the following specific performance metrics for
product development processes:
Product development lead time Most companies consider their
product development lead time to be exceptionally important for
determining the performance of their product development activi-
ties. Product development lead time is particularly important because
this metric determines the speed with which new products can be
introduced into the marketplace. Companies that have high speed
in product development can introduce new products more often and
adapt more quickly to changes in customer tastes. This ultimately
translates into a larger market share for the company. Lead time is
usually measured in months, and can range from fractions of a month
to tens of months, depending on the complexity of the product and the
skill of a companys development staff.
Efficiency In attempting to reduce product development lead times,
however, few companies can afford to ignore the efciency of their prod-
uct development. In product development, efciency is the cost of man-
power and other resources required for the product development.
Robustness In addition, the robustness (including quality, reliabil-

ity, and exibilityhow well the product does what it was meant to
do) of the design is particularly important for evaluating any product
Life-cycle cost Life-cycle costs, including development costs; pro-
duction costs; sales and distribution costs; service, support, and war-
ranty costs; and disposal costs may be included in computing the
life-cycle cost for a product. Some companies even include the costs
due to pollution during the production and use of the product as part
of the holistic analysis of the life-cycle cost. Product development has a
particular vested interest in keeping the life-cycle cost for any product
as low as possible.
On a longer time scale, product development lead time, efficiency,

robustness, and life-cycle costs will contribute a great deal to the level
of customer satisfaction, market share, and revenues that the company
will have. These will in turn translate into profitability and will influ-
ence the organizations long-term business viability.
2.1.1 Product Development Process

Flowchart
Product development is a complicated process. There are many subpro-
cesses, such as capturing what customers really want, creating product
concept designs and detailed product designs, and designing the manu-
facturing process. These subprocesses focus on issues such as the func-
tionality of the product, as well as aesthetics and ease of manufacture.
Other tasks in product development may include building and testing
of prototypes using various pieces of test equipment, and analysis of
candidate designs. Common goals in product design and development
processes include shortening the time required to design new products,
introducing new designs more frequently, producing designs that are
more innovative while meeting the customers needs, and reducing the
cost required to design new products.
In general, product development involves the activities that are
used to
1. Determine that a new product is required to serve some need

2. Conceive of a concept for the product based on the wants and needs
of customers
3. Develop all the technical specications for the product
4. Devise a production process
5. Validate both the design and the production process
14 Chapter Two
Figure 2.4 illustrates a simple process flowchart for the product devel-
opment process. In this process, there are eight stages:
Stage 0 Impetus/ideation
Stage 1 Customer and business requirements study
Stage 2 Concept development
Stage 3 Product/service design/prototyping
Stage 4 Manufacturing process preparation/product launch
Stage 5 Production
Stage 6 Product/service consumption
Stage 7 Disposal
The earlier stages of the cycle are often called upstream, and the later
stages are often called downstream. Since this book deals with voice of the
customer, and the voice of the customer is mostly related to the first four
stages of the product life cycle. This chapter will cover the first four stages.
There is quite a bit of other literature that covers the last four stages.
2.2 The Product Development

ProcessEnd to End
Many companies manage their product development process by using
flowcharts similar to that in Figure 2.4. Managing product development
by stages is also called the stage-gate process. For different companies,
and different kind of products, the number of stages, and the name of
each stage, might be somewhat different. Some product development is
simply a modification of existing product. However, in even the simplest
product development process, there are several distinct stages. For dif-
ferent stages, the objectives and the type of work performed in each
stage will be distinctively different. In this section, I will discuss the
first four stages in a typical product development process in detail. It is
crucial for all the people involved in the product development process,
be they engineers, marketing people, or managers, to understand this
process in order to do their job well.
2.2.1 Opportunity Identification and

Idea Generation: Stage 0
Opportunity identification and idea generation is the very first stage
of the product development process. It is Stage 0 in the flowchart
illustrated in Figure 2.4. This stage may also be called the fuzzy front
end, because this stage is at the front end of the product development
process, and it is not clear what exactly will be accomplished in this
stage. Typically, an opportunity for a new or improved product is
Stage 0: Impetus/Ideation
New technology, new ideas, competition lead to new product/service possibilities
Several product/service options are developed for those possibilities
Stage 1: Customer and business requirements study

Identification of customer needs and wants
Translation of voice of the customer into functional and measurable product/service requirements
Business feasibility study
Stage 2: Concept Development

High-level concept: general purpose, market position, value proposition
Product definition: Base-level functional requirement
Design concept generation, evaluation, and selection
System/architect/organization design
Modeling, simulation, initial design on computer or paper
Stage 3: Product/Service Design/Prototyping

Generate exact detailed functional requirements
Develop actual implementation to satisfy functional requirements, i.e. design parameters
Build prototypes
Conduct manufacturing system design
Conduct design validation
Stage 4: Manufacturing Process Preparation/Product Launch

Finalize manufacturing process design
Conduct process testing, adjustment, and validation
Conduct manufacturing process installation
Stage 5: Production
Process operation, control, and adjustment
Supplier/parts management
Stage 6: Product/Service Consumption

After-sale service
Stage 7: Disposal
Figure 2.4 A typical product/service life cycle

16 Chapter Two
identified by marketing people or upper management, and then the

product development team will explore this opportunity. However,
the outcomes of this stage can be very diverse; sometimes, the team
may find this opportunity is worth pursuing, so it develops a follow-
up product development plan. Sometimes the team may find it is not
feasible to develop an appropriate product at this time. This stage is
also called the discovery stage. Whatever name people may call this
stage, this is where you design the DNA for your product and plant
the seeds. This stage might be the highest leverage point in product
development because it will define what the product will be.
The fuzzy front end is the messy getting started period of the new
product development process. The front end is where the organization
formulates a concept of the product to be developed and decides whether
to invest resources in the further development of an idea. It includes all
activities, from the search for new opportunities, to the formation of a
germ of an idea, to the development of a precise concept. The fuzzy front
end ends when an organization begins development.
Although the fuzzy front end may not be an expensive part of product
development, it can consume 50 percent of development time (Smith
and Reinertsen 1998), and it is where major commitments are typically
made involving time, money, and the products nature. Consequently,
this phase should be considered as an essential part of development
rather than something that happens before development, and its cycle
time should be included in the total development cycle time.
There are three different elements in this fuzzy front end:
1. Opportunity identication. In this element, large or incremental
business and technological changes are identied in a more or less
structured way. As discussed in Chapter 1, there are two major
driving forces for new product development: the market pull and
technology push. The market pull simply means that there are new
market trends, new voices of customers that lead to new product
opportunities, and so on. Technology push means that new discover-
ies and inventions in science and technology may lead to new product
opportunities. Overall, in opportunity identication, there are two
voices that need to be heard: the voice of the customer and the voice
of technology. There are many methods in capturing these two voices,
such as voice-of-the-customer (VOC) capturing, which is the main
theme of this book. The methods of capturing the voice of technol-
ogy include technology roadmaps, technology trees, and the theory
of inventive problem solving (TRIZ).
2. Opportunity analysis. The second element is the opportunity analysis.
It is done to translate the identied opportunities into the business-
and technology-specic context of the company. In this element, there
are also two kind of analyses: one is VOC analysis, and the other
is technology analysis. Chapter 6 of this book will discuss voice-of-

the-customer analysis in detail. The technology analysis can be sup-
ported by technology road maps, technology trees, and the theory of
inventive problem solving.
3. Idea generation and handling. The third element is the generation
and handling of ideas. The most important objective of Stage 0 is to
generate viable product ideas that can be further developed into great
products to t the market opportunities. Until the 1990s, most com-
panies were satised with regular product extensions and hunches as
new ideas. In todays global economy, with much greater global com-
petition, the ability to generate breakthrough innovations becomes
so important that this light bulb glow or hit-and-miss type of idea
generation approach is no longer adequate. A well-organized idea
generation and handling system becomes necessary. A good idea gen-
eration and handling system should have the following features:
You need to generate many ideas, because the majority of ini-
tial ideas will be discarded due to some kind of infeasibilities.
Generating many ideas will ensure that some good ideas will sur-
vive at the end.
You need to develop ideas from many different sources, from dif-
ferent people, some ideas from your own research and develop-
ment people, some ideas from your customers or suppliers, some
ideas from university researchers, and so on. Comparing, combin-
ing, improving, and cross-pollinating ideas from these sources will
often generate some very strong, unique, mature new ideas.
All incoming ideas need to be screened, analyzed, reviewed,
grouped, improved, and nally sorted by a responsible person (usu-
ally a product development manager) or team. Some obviously
good ideas will be sorted out and can be sent to the next product
development stage. The ideas that cant pass the screening and
sorting are not necessarily bad ideas; they might be good ideas,
but the time is not ripe to fully develop them. These ideas should
be put in an idea vault and periodically retrieved and reviewed
for their viabilities. After each idea gets reviewed, the feedback
should be sent to the person who generated the idea, so he or she
can further improve his or her ideas. Other people in the company
should be able to get access to this idea vault.
2.2.2 Customer and Business Requirement

Study: Stage 1
In Stage 1, the information captured in the VOC study by using meth-
ods such as surveys, focus groups, and ethnographic research is further
analyzed to gain deeper understanding of its meaning.
18 Chapter Two
The raw voice of customers is often vague, confusing, and not well
defined. With the help of some methods such as the affinity diagram
method (KJ method), this raw voice-of-the-customer data will be trans-
lated into a new set of relatively well-defined, quantitative quality metrics.
The affinity diagram method will be thoroughly discussed in Chapter 6.
Usually, people in the industry call these metrics top-level specifications.
In Six Sigma practice; these top-level specifications are called Critical-
to-Quality characteristics (CTQ). The translation of the raw voice of the
customer data into CTQs will help the product development group to
establish a clear picture of exactly what kind of product customers really
want. This CTQ information can be further used to estimate the cost of
developing and producing such a product; thus it will help to build the
business feasibility study. The methods for translating the raw voice of
the customer to CTQs will be discussed in detail in Chapter 6.
Another key analysis of the raw voice-of-the-customer data is to con-
duct customer value analysis. A customer value analysis identifies key
customer value factors, such as key performance characteristics and
price, and it evaluates the competitive comparison of your product with
your competitors product in all these customer value factors. Customer
value analysis is very useful in determining how well your product can
compete in the marketplace, and identifying the key areas to improve
to increase your competitiveness. Customer value analysis is discussed
thoroughly in Chapter 3.
During this stage, the business case will be developed based on
estimates of the total available market, customer needs, investment
requirements, competition analysis, and project uncertainty. Based on
your business case, you will make the decision of launching the product
development project.
2.2.3 Concept Development: Stage 2

Concept development (Stage 2 in Figure 2.4) is an important stage in
product development. It is a process in which the concept of a new prod-
uct, or product platform, is developed gradually, from fuzzy to specific;
from a higher level, such as the system level, to the lower level of the
product architecture, such as subsystems and components. The concept
development stage comes before prototype building and physical testing
(which is Stage 3).
Concept development is an extremely important stage; the cost
incurred in this stage is lower than later stages, because you havent
started expensive prototype building, validation testing, manufactur-
ing tooling tests, and so on. However, the impact of this stage is huge,
much like conceiving a new baby. If the new baby has perfect DNA, the
future for the new baby will look great; if the baby has DNA flaws, it is
nearly impossible to make it up at a later stage. A study by Fredrikson
(1994) concluded that the design decisions made during the early stages
of the design life cycle have the largest impact on total cost and quality
of the system. It is often claimed that up to 80 percent of the total cost
is committed in the concept development phase (Fredrikson 1994). In
my experience, at least 80 percent of the design quality is also commit-
ted in the early phases, as depicted in Figure 2.5. In the early design
stage, the cost committed is low, but the impact on product design is
high; in the later design stage, the cost committed is very high (after
buying production machineries, facilities, and so on), but at this later
phase, design changes for corrective actions can only be achieved at high
costs, including customer dissatisfaction, warranty, marketing promo-
tions, and in many cases under the scrutiny of the government (such
as recall costs).
In the concept design stage, where few final design decisions have
been made, there is a lot of freedom to choose and try out different
design concepts. Different concepts could lead to vastly different prod-
uct performances, quality, and cost. So it is important to front-load this
concept design stage, that is, put a lot more effort into this stage.
The concept design stage can be further divided into the following
steps:
High-level concept development
Product denition and base-level functional requirement
Design concept generation, evaluation, and selection
System/architect/organization design
Modeling, simulation, initial design on computer or paper
Cost
Cost vs. Impact
Potential is negative
Potential is positive (Impact < Cost)
(Impact > Cost)
Impact
Time
Design Produce/Build Deliver Service Support
Figure 2.5 Effect of Design Phases on Product Life Cycle Cost
20 Chapter Two
There are several methods that can be used to support this stage,
which Ill discuss here.
2.2.3.1 User Innovation Sometimes the best product concept ideas

come from customers. User innovation refers to innovations developed
by consumers and end users, rather than manufacturers. Eric von
Hippel of MIT discovered that most products and services are actu-
ally developed by users, who then give ideas to manufacturers. This is
because products are developed to meet the widest possible need; when
individual users face problems that the majority of consumers do not,
they have no choice but to develop their own modifications to exist-
ing products, or entirely new products, to solve their issues. In using
this method, you should focus on lead users. Lead users are users who
are at the leading edge of the applications of the product, or they are
ahead of their time. For example, some high-end product research and
development engineers use computer-aided design software to analyze
some extremely challenging problems; they really use the software to
its limits, so they often supplement the existing software with their
own subroutines, driver programs, and so on. On the other hand, most
regular users, such as draft engineers, just use this software as draft-
ing software; their application of the software is far from reaching the
softwares functional limits. Lead users usually have pretty good ideas
about the weaknesses of the current product and what additional func-
tions should be added to make the product more perfect.
2.2.3.2 Theory of Inventive Problem Solving TRIZ (Teoriya Resheniya

Izobreatatelskikh Zadatch) is the Russian acronym for Theory of
Inventive Problem Solving (TIPS). TRIZ has developed over 1,500
person-years of research and study over a significant portion of the
world, including the most successful solutions of problems from science,
mathematics, and engineering, and systematic analysis of successful
patents from around the world, as well as the study of the internal,
psychological aspects of human creativity.
Dr. Genrich S. Altshuller, the creator of TRIZ, started an investigation
of invention and creativity in 1946. Based on the fact that almost all
human inventions are documented in the form of patents, Dr. Altshuller
theorized that studying patents could lead to uncovering the secrets
of inventions and inventors. Dr. Altshuller initially reviewed 200,000
patent abstracts; in subsequent years, many millions of patents were
further studied by many TRIZ researchers. Based on the extensive stud-
ies of those patents, these are the major initial findings of TRIZ:
Though there are millions of patents and inventions, all innovations

are based on a very small number of inventive principles and strat-
egies, and those inventive principles and strategies can be taught
and learned. By learning and using these principles and strategies,

people can shortcut the invention process.
The bottlenecks for breakthroughs in technology or product are
often caused by contradictiona contradiction is usually caused by
an improvement to one technical attribute of a system that leads to
deterioration of other technical attributes. For example, as a container
becomes stronger, it becomes heavier, and faster automobile accelera-
tion reduces fuel efciency. Inventions are often related to overcoming
contradictions; many contradictions can be resolved by using inven-
tive principles.
By studying the history of inventions, TRIZ developed sophisticated
prediction methods for technology evolution trends, that is, predicting
what will be next in a given branch of technology. This TRIZ technical
prediction tool can be used to guide a company to steps ahead of their
competitors.
Over the next 40 years (19501990), TRIZ developed into a system of

philosophy, a problem-solving process, and methods. TRIZ has five key
philosophical elements:
Ideality The ultimate criterion for system excellence; this crite-

rion is the maximization of the benets provided by the system, and
minimization of the harmful effects and costs associated with the
system.
Functionality The fundamental building block of system analysis;
it builds models about how a system works, and how it creates benet,
harm, and costs.
Resources Maximum utilization of resources is one of the keys to
achieve maximum ideality.
Contradictions Contradiction is a common inhibitor for increas-
ing functionality; removing contradiction usually greatly increases
the functionality and raises the system to a totally new performance
level.
Evolution The evolutionary trend of the development of technologi-
cal systems is highly predictable, and can be used to guide further
development.
The TRIZ problem-solving process is a four-step process, consisting

of problem definition, problem classification and tool selection, solution
generation, and evaluation.
The TRIZ methods include inventive principles, separation principles,
system simplification, and system functional analysis and improvement
methods, as well as a technical prediction method.
22 Chapter Two
After the breakup of the Soviet Union in the 1990s, TRIZ became
known to the rest of the world. The bulk of TRIZ research, development,
and implementation also moved to the West. Every year, TRIZ research-
ers will study new patents to update the TRIZ knowledge base. However,
TRIZ researchers found that while many millions of patents are being
filed, the number of new inventive principles is growing at a much,
much slower speed. This fact actually proves TRIZ is a valuable, stable
knowledge base that is not going to be obsolete quickly. The application
of TRIZ is also on the risebased on the TRIZ Journal, 105 companies
are currently using TRIZ, including Dupont, Mobil, 3M, Honeywell,
Ford, Toyota, BMW, Samsung, Pfizer, and Lockheed Martin. TRIZ has
helped these companies to develop many breakthrough inventions and
greatly improved their competitive positions. Noticeably, Samsung is
an exemplar company for using TRIZ. According to Fortune magazine,
TRIZ has helped Samsung to greatly improve its research and develop-
ment capabilities and aided Samsung in surpassing Sony as a leading
company in consumer electronics.
Overall, TRIZ is a systematic innovation method that is based on the
distilled fundamental knowledge of millions of successful patents and
business case studies. Learning TRIZ will make product development
team members significantly more creative, and shorten the process of
innovation. Successful TRIZ applications can greatly improve their com-
petitive advantage.
In the concept development stage, TRIZ is a powerful tool that can
be used to generate creative product ideas. TRIZ will be discussed in
detail in Chapter 10.
2.2.3.3 Value Engineering From the customers point of view, a prod-

ucts value is the functions and benefits it delivers. From the manufac-
turers point of view, the product consists of a bundle of parts and inter-
faces, and these parts and interfaces cost money. Value engineering is a
systematic method that can trace relevant parts and interfaces to each
products functions and benefits, and calculate what is the lowest cost
to deliver each function and benefit. From this information, the product
development team can find ways to improve the product design by de-
livering the same functions and benefits with the lowest possible cost.
Value engineering is very useful to improve product concepts. Value
engineering is thoroughly discussed in Chapter 9.
2.2.3.4 Quality Function Deployment Quality function deployment (QFD)

is a flexible and comprehensive group decision-making technique used in
product or service development. QFD can strongly help an organization
focus on the critical characteristics of a new or existing product or service

from the separate viewpoints of the customer market segments, company,
or technology development needs. The results of the technique yield
transparent and visible graphs and matrices that can be reused for future
product developments.
QFD transforms customer needs (the voice of the customer) into engi-
neering characteristics of a product or service, prioritizing each product
characteristic while simultaneously setting development targets for
product or service development.
The most important tool in the quality function deployment is the
House of Quality (HOQ). The House of Quality is a graphic tool for defin-
ing the relationship between customer desires and the products capa-
bilities. It utilizes a planning matrix to relate customer wants to how a
firm (that produces the products) is going to meet those wants. The HOQ
provides and organizes the information that the product team needs to
refine the product concept. It looks like a house with a correlation matrix
as its roof, customer wants versus product features as the main part,
competitor evaluation as the porch, and so on. Figure 2.6 shows what a
house of quality looks like. The QFD method will be discussed in detail
in Chapter 7.
4
Correlation
Matrix (Roof)
2 Technical
Specifications
5 Customer
1 Customer Market
Requirements & 3 Evaluations
Relationship
Importance
Matrix
Ratings
6 Target Goals
for Tech. Spec.
7 Technical
Competitive
Evaluations
8 Tech. Spec.
Imp. Ratings
Figure 2.6 House of Quality

24 Chapter Two
2.2.3.5 Pugh Concept Selection Pugh concept selection is a design

evaluation for several design alternatives. Pugh concept selection uses
a scoring matrix called the Pugh matrix. The Pugh matrix is related
to the QFD method and is a form of prioritization matrix. By using
the Pugh concept selection method, every design alternative will get a
score, so the best design alternative can be selected.
2.2.3.6 Set-Based Design and Modularity Modular design is a design

practice in which a product is broken into smaller subsystems. The
subsystems are connected via standard interfaces. In this case, the sub-
systems become decoupled, that is, the designing of one subsystem is
not dependent upon other subsystems. Therefore, the design work for
each subsystem can be conducted in parallel. Under modular design,
set-based design practice can be used. Set-based design means that
instead of generating one design concept for each subsystem, multiple
design concepts are generated for each subsystem by different teams.
For example, a bicycle can be divided into the following subsystems,
as illustrated by Figure 2.7: frame, drive, wheels, brakes, and suspen-
sion. A possible scenario for set-based design is that three different
design concepts could be developed for each subsystem. Because these
subsystems are designed to fit standard interfaces, the possibilities add
up as follows:
3 frames 3 drives 3 wheel sets 3 brakes 3 suspensions
= 243 combinations
This huge amount of design information will give you sufficient
choices to end up with an extremely solid final design.
Figure 2.7 Subsystems of a Modular Design

Set-based design should only be used in the concept development

stage, where the design cost is low and before you build prototypes and
conduct tests. Set-based design is an important tool in lean product
development, which will be discussed in detail later in this chapter.
2.2.4 Product Design and Prototype: Stage 3

After the concept development stage, the product design needs to be
refined. The specific product design should be developed for each subsys-
tem and components. Design parameters of every level should be devel-
oped, analyzed, verified, and tested, if necessary. In this stage (Stage 3),
physical prototypes will be built and tests will be conducted.
The product design and prototype stage can be further divided into
the following steps:
1. Generate exact detailed functional requirements

2. Develop actual implementation to satisfy functional requirements,
that is, design parameters
3. Build prototypes
4. Conduct manufacturing system design
5. Conduct design validation
The product design and prototype stage is also a very important stage.
However, the freedom to change the design will be rather limited, because
the design concept, product architecture, and technologies selection for
the product design would have been finalized at the concept develop-
ment stage. For example, in an automotive design project, the choice of
using a traditional internal combustion engine or a hybrid engine would
have been firmly finalized in or before the concept development stage.
If you make a decision to use a traditional internal combustion engine,
then in the product design stage, you have to stick with this choice,
even if gas prices are suddenly raised to $5 per gallon. If you change
the design at this point, everything else will be affected and you would
essentially have to redesign the whole thing. In the product design and
prototype stage, you still can change many design parameters, without
changing the product concept, technology, and architecture.
There are several methods that can be used to support this stage,
specifically Design of Experiments (DOE), Taguchi method and robust
design, and computer simulation models.
2.2.4.1 Design of Experiments Design of Experiments (DOE) is also

called statistically designed experiments. The purpose of the experiment
and data analysis is to find the cause-and-effect relationship between
the output and experimental factors in a process. The process model of
DOE is illustrated in Figure 2.8.
26 Chapter Two
Controllable Factors
x1 x2 ..... xn
...
Input Output y
Process Response
...
z1 z2 ..... zp Figure 2.8 A Process Model of DOE
Uncontrollable factors
In any DOE project, you deliberately change the experimental fac-

tors and observe their effect on the output. The data obtained in the
experiment will be used to fit empirical models relating output (y) with
experimental factors. In product design practice, y is often used to rep-
resent a product performance metric. x1, x2, , xn are often relevant
product design parameters or process factors. For example, in a chemical
process, the performance metric, y, could be the yield of the process, and
the key process factors, such as temperature and pressure, could be x1,
and x2. Mathematically, you are trying to find the following functional
relationship:
y = f(x1, x2, , xn) + (2-1)

where is experimental error, or experimental variation. The exis-
tence of means that there may not be an exact functional relationship
between y and (x1, x2, , xn). This is because
Uncontrollable factors (z1, z2, , zp) will inuence the response y,

but are not accounted for in equation (2-1).
There are experimental and measurement errors on both y and
(x1, x2, , xn) in the experiment.
A DOE project requires many steps, as described in the following

sections.
Step 1: DOE Project Definition This is the first step, but certainly
not a trivial one. You need to identify the objective of the project and
find the scope of the problem. For example, in a product design, you
need to identify what you want to accomplish. Do you want to reduce
defects? Do you want to improve the current products performance?
What is the performance? What is the project scope? Do you work on a
subsystem, or a component?
Step 2: Selection of Response Variable (Output) After defining the

project, you need to select the response variable, y. In selecting the
response variable, the experimenter should determine if this response
variable could provide useful information about the process under
study. Usually, the response variable is a key performance measure of
the process. You also want y to be
A continuous variable, which would make data analysis much easier
and more meaningful
A variable that can be easily and accurately measured
Step 3: Choice of Factors, Levels, and Ranges Actually, steps 2 and 3

can be done simultaneously. It is desirable to identify all the impor-
tant factors that have significant influence on the response variable.
Sometimes, the choice of factors is quite obvious, but there are cases
where a few very important factors are hidden.
There are two kinds of factors: the continuous factor and the discrete
factor. A continuous factor can be expressed by continuous real numbers.
For example, weight, speed, and price are continuous factors. A discrete
factor is also called a category variable, or attributes. For example, types
of machines, types of seed, and types of operating system are discrete
factors.
In a DOE project, each experimental factor will be changed at least
once. That is, each factor will have at least two settings. Otherwise, that
factor will not be a variable, but rather a fixed factor in the experiment.
The numbers of settings of a factor in the experiment are called levels.
The choice of number of levels in the experiment also depends on time
and cost considerations. The more levels you have in experimental fac-
tors, the more information you will get from the experiment, but there
will be more experimental runs, leading to higher cost and longer time
to finish the experiment.
Step 4: Select an Experimental Design The type of experimental

design you select will depend on the number of factors, the number of
levels in each factor, and the total number of experimental runs that
you can afford to run.
The most frequently used experimental designs are full factorial
designs and fractional factorial designs. If the number of factors and
levels are given, then a full factorial experiment will need more experi-
mental runs, thus becoming more costly, but it will also provide more
information about the process under study. The fractional factorial will
need a smaller number of runs, thus costing less, but will also provide
less information about the process.
28 Chapter Two
I will discuss how to choose a good experimental design in subsequent

sections.
Step 5: Perform the Experiment When running the experiment, you

must pay attention to the following:
Check performance of gauges/measurement devices rst.
Check that all planned runs are feasible.
Watch out for process drifts and shifts during the run.
Avoid unplanned changes (such as swapping operators at the halfway
point).
Allow some time (and backup material) for unexpected events.
Obtain buy-in from all parties involved.
Preserve all the raw data.
Record everything that happens.
Reset equipment to its original state after the experiment.
Step 6: Analysis of DOE Data Statistical methods will be used in

data analysis. From the analysis of experimental data, you can obtain
the following results:
Identication of signicant and insignicant effects and interac-

tions. (However, it is also possible that none of the experimental fac-
tors will be found to be signicant, in which case the experiment is
inconclusive. This situation may indicate that you may have missed
important factors in the experiment.)
DOE data analysis, which can identify signicant and insignicant
factors by using Analysis of Variance (ANOVA)
Ranking of relative importance of factor effects and interactions
Empirical mathematical model of response versus experimental
factors
Identication of best factor-level settings and optimal output per-
formance level
Step 7: Conclusions and Recommendations Once the data analysis

is completed, the experimenter can draw practical conclusions about
the project. If the data analysis provides enough information, you
might be able to recommend some changes to the process to improve
its performance. Sometimes, the data analysis cannot provide enough
information, in which case you may have to do more experiments.
When the analysis of the experiment is complete, you must verify

whether the conclusions are valid or not. These are called confirmation
runs.
2.2.4.2 Taguchi Method and Robust Design The Taguchi method is a

comprehensive quality strategy that builds robustness into a product
or process during its design stage. The Taguchi method is a combination
of sound engineering design principles and Taguchis version of DOE,
which is called an orthogonal array experiment.
In regular DOE, usually only controllable factors are studied. In the
Taguchi method, both controllable factors and noise factors are studied.
Noise factors are also called uncontrollable factors; they are factors that
are not under the manufacturers control, such as usage environment
(temperature, humidity, vibration, and so on), customers misuse,
incoming material, and component variation. Clearly, both controllable
factors (design parameters or process factors) and noise factors will
affect the product performance (y). In the Taguchi method, a special
type of experimental design, called inner-outer arrays, is used to study
both control factors and noise factors.
In regular DOE, the common objective is to optimize y; in the
Taguchi method, the objective is to achieve robustness. For example,
in a traditional DOE study on a chemical process, you may try to find
optimal control-factor levels, that is, the best temperature and best
pressure to maximize the yield. Under these best temperature and
pressure levels, the average yield may be high, but it might be very
sensitive to other noise factors, such as incoming materials. When
incoming material properties are changing because of batch-to-batch
material variations, your chemical quality and effective yield may
fluctuate a lot. In this case, the optimal settings are not robust. The
Taguchi method looks for controllable factor-level settings that can
deliver consistent performance even under the influence of significant
noise factor variation. In the chemical process example, the Taguchi
method tries to find the best temperature and pressure levels that
not only deliver reasonably high yield, but also deliver stable quality
and yield under noise factor variation, such as variation of incoming
material.
Robust design is a very important approach in the product design
stage to ensure that the product has consistent quality and reliability.
2.2.4.3 Computer Simulation Models Increasingly, choices of computer

simulation models are available for many applications. There are two
categories of computer simulation models. They are mechanism-based
simulation models and Monte Carlo simulation models.
30 Chapter Two
Mechanism-Based Simulation Models Mechanism-based simulation

models are usually commercial computer packages that are designed for
specific fields of applications, such as mechanical, electrical, and elec-
tronics. The computer packages follow established scientific principles
or laws to model the relationships for specific classes of applications.
Mechanical components and system design can be created and ana-
lyzed using computer-aided design (CAD) software. Analysis capability
includes geometrical dimensions and tolerances (GD&T), three-dimen-
sional views and animation, and so on. CAD software is usually the
starting point for many computer-aided engineering (CAE) analyses,
such as stress and vibration analysis.
Finite element analysis (FEA) is a computer-based technique that
can evaluate the many key mechanical relationships, such as the rela-
tionships among force, deformation, material property, dimension, and
structure strength, in a design. In FEA, the original continuous 3-D
geometrical shapes are subdivided into many small units, called finite
element meshes. The precise differential equation forms of basic relation-
ships in mechanics are approximated by a large-scale linear equation
system. FEA methods can study mechanical stress patterns in static
or dynamic loaded structures, strain, response to vibration, heat flow,
and fluid flow.
The use of FEA to study the fluid flow is called computational fluid
dynamics (CFD). CFD can be used to analyze air flow, fluid flow, and so
on; thus, it can be used to analyze aerospace design, automobile engine
design, ventilation systems, and others.
Electrical and electronic circuits can be analyzed by using electrical
design automation (EDA) software that is based on the mechanism
models in electrical engineering, such as Ohms law, logical circuit
models, and so on. In EDA software, component parameters, input
electrical signals, such as power, waveform, and voltage, and circuit
diagrams, are inputted into the program. The program can perform
many different analyses, such as evaluation of circuit performances,
sensitivity analysis, and tolerance analysis.
Electromagnetic effects can be analyzed by using a special kind of
FEA software that is based on Maxwells law.
There are also numerous mechanism-based computer simulation
models for other applications, such as chemical engineering, financial
operations, and economics.
Monte Carlo Simulation Model Monte Carlo simulation is a tech-

nique that simulates large numbers of random events. For example,
transactions in a bank branch office are random events. A random
number of customers enter the office at random arrival times, and
each one will make a random type of transaction, such as deposit,
withdrawal, or loan, with a random amount of money. A Monte Carlo

simulation model can simulate such events. A Monte Carlo simulation
starts with generating a large number of random numbers that fol-
low prespecified probability distributions. For example, the customer
arrival time to the bank branch can be assumed to be a Poisson ar-
rival process, in which the key parameter, that is the mean inter-ar-
rival time, can be set by the analyst or estimated with old data. After
you set all the parameters for the probability models, such as mean
and variance of the transaction amount, average transaction time,
and system interaction relationships, the Monte Carlo simulation can
generate virtual bank transaction processes for a large number of
virtual customer arrivals. The simulation model can then provide key
system performance statistics in the form of a chart and or histogram.
In the bank example, these key performance measures could include
customer waiting time, percentage of time that a clerk is idle, daily
transaction amount, and so on.
A Monte Carlo simulation is very helpful in analyzing service pro-
cesses, factory flows, banks, hospitals, and so on.
2.2.4.4 Prototypes Prototypes are trial models for the product. Ulrich
and Eppinger (2000) defined a prototype as an approximation of the
product along one or more dimensions of interest. In this context, a
prototype can be a drawing, a computer model, a plastic model, or a
fully functional prototype fabricated at the pilot plant.
Ulrich and Eppinger also further defined two types of prototype: ana-
lytical or physical. An analytical prototype represents the product in a
mathematical or computational form. Many aspects of the product can
be analyzed in this form. For example, a finite element model (FEM)
can be used to analyze a mechanical part with respect to force stress,
deformation, and so on. A Monte Carlo simulation model can be used
to simulate service flow, waiting time, and patient process rate in a
clinic. A physical prototype is a real look-alike prototype made of either
substitute materials or actual materials designed for the product. For
example, a prototype of an automobile manifold may be made of easy-
to-form plastic instead of the aluminum that the finished product would
be made of. In this case, researchers are only interested in studying the
geometrical shape aspect of the manifold, not the heat resistance and
strength aspects of the manifold.
A prototype can be also focused or comprehensive (Ulrich and Eppinger
2000). A focused prototype only represents a part of, or a subset of, the
products functions or attributes. A comprehensive prototype represents
most of the product functions and attributes, and some prototypes rep-
resent all product functions and attributes.
32 Chapter Two
All analytical prototypes are focused prototypes, at least at the current

technological level, because it is impossible to build a virtual model that
can represent all of a products functions and attributes. For example, an
FEM model can only represent the mechanical aspect of the products
characteristics; it cannot represent chemical properties such as corro-
sion. Physical prototypes can be very comprehensive; a preproduction
prototype has all of the products required functions and attributes.
Physical prototypes can also be very focused; a plastic prototype may
only be able to represent the geometrical aspect of a part, not material
properties or functions.
There are four commonly used physical prototypes: experimental pro-
totype, alpha prototype, beta prototype, and preproduction prototype.
Experimental prototypes are very focused physical prototypes; they
are designed and made to test or analyze a very well-defined subset of
functions and attributes. For example, a plastic prototype of an auto-
mobile manifold is built so that the engineering team can study the
geometrical aspects of manifold functions.
Alpha prototypes are used in product functional performance valida-
tion. An experimental prototype made in the lab, or a concept car model,
is one such example. Usually an alpha prototype can deliver all the
intended functions of a product. The materials and components used in
alpha prototypes are similar to those that will be used in real produc-
tion. But they are made in a prototype process, not by the manufacturing
process to be used in mass production. The rapid prototyping technique
is often used to make alpha prototypes because of its speed. Rapid pro-
totyping is the automatic construction of physical objects using solid
free-form fabrication. The first techniques for rapid prototyping became
available in the 1980s and were used to produce models and prototype
parts. Today, they are used for a much wider range of applications and
are even used to manufacture production-quality parts in relatively
small numbers.
Beta prototypes are used to analyze reliability requirements valida-
tion, usage requirements validation, product specification validation,
and so on. They may also be used to test and debug the manufacturing
process. The parts in beta prototypes are usually made with the actual
production process or supplied by the intended part suppliers, but usu-
ally they are not produced at an intended mass production facility. For
example, an automobile door panel beta prototype might be made by
the same machinery as that of the assembly plant, but it is made by a
selected group of validation engineers and technicians, not by hourly
workers.
Preproduction prototypes are the first batch of products made by the
mass production process. At this point in time the mass production
process is not operating at full capacity. These prototypes are usually
Physical Beta
Prototype
Alpha Real
Prototype Product
Experimental
Prototype
Focused Comprehensive
Improbable
FEM
Region
Model
Monte-Carlo Clinic
Simulation Model
Analytical
Figure 2.9 Types and Classications of Prototypes
used to verify production process capability, as well as test and debug

the mass production process.
Figure 2.9 illustrates the types and classifications of prototypes.
So far, I have discussed Stage 0 to Stage 3 of the product life-cycle
process. These stages cover the product development process, which is
also called the upstream of the product life cycle. The remaining four
stages cover the production, usage, and disposal of the product, which are
called the downstream of the product life cycle. Since the main objective
of this book is capturing and analyzing the voice of the customer, which
is primarily related to the product development process, I will not discuss
the last four stages of the product life cycle in detail in this chapter.
2.3 The Nature of Product Development:

Information and Knowledge Creation
In the last two sections, we discussed the common features of product
development processes. There are many companies that develop and
manufacture products, but the results of their product development
processes are vastly different. Even within the same industry, the prod-
uct development process of the best company can be vastly better than
that of its peers. Recently, the product development process used by
34 Chapter Two
Toyota (Morgan and Liker 2005) received a lot of attention. Compared

with other automotive manufacturers, Toyota not only has a highly
efficient lean production process, but also Toyotas key product develop-
ment performance metrics, such as product development lead time (how
long it takes from the beginning of design of a new vehicle to launch
into production), the quality of design (how competitive the designed
vehicle is in the marketplace), and total development cost, are much
better than its competitors.
These facts lead to several questions. What makes a product
development process a better process? Can people learn, understand,
and develop a better product development process? What is the nature
of product development? Is there a scientific basis of product design and
development? If yes, can we use this scientific basis to guide product
development processes? I will try to answer these questions in this part
of the chapter.
Research on the product design and product development process
gradually became popular in recent years. Research in the product
design arena started in Europe. Primarily, the Germans developed some
design guidelines that kept improving at a consistent pace. A huge body
of research has been published in German on design practice. Most of
these efforts are listed in Hubka (1980) and Phal and Beitz (1988). The
German design schools share common observations. For example, a good
design entity can be judged by its obedience to some design principles;
the design practice should be decomposed to consecutive phases; and
methods for concept selection are required. Besides the Germans, the
Russians also developed an empirical inventive theory which promises
to solve difficult and seemingly impossible engineering problems, the so-
called TRIZ or Theory of Inventive Problem Solving (TIPS) (Altshuller
1988, 1990; Rantanen 1988; Tsourikov 1993).
In the United States, over the last three decades, there have been pro-
gressive research efforts in the product design area, particularly in the
field of mechanical design. Dixon (1966) and Penny (1970) are among
the pioneers in this research. Ullman (1992) stated that the activities of
design research were accelerating along the following trends: artificial
intelligence computer-based models, design synthesis (configuration),
cognitive modeling, and design methodologies. Suh (1990) developed the
axiomatic design approach; in this approach; the design practice is mod-
eled as mappings from one domain to another domain. Several design
principles are derived to characterize and benchmark the design practice.
This approach enables evaluation of design solutions based on different
principles. McCord and Eppinger (1993) and Pimmler and Eppinger (1994)
developed a theoretical framework to optimize the sequence of product
development projects. The stage-gate system was developed by Cooper
(1990) to manage the product development process. In the stage-gate

system, the stages are the product development stages that I illustrated
in the preceding section; the gate means a rigorous evaluation process at
the end of each stage, where the product development team tries to make
sure that all necessary pieces of work in this stage are performed satisfac-
torily, and unnecessary projects are discontinued. In Reinertsens excellent
book, Managing Design Factories (1997), the product development process
is treated as an information generation process; how to generate informa-
tion effectively becomes an important goal for a good product development
process. To improve the product development lead time, queue theory is
used to guide the rules for design project sequencing. Huthwaite (2004)
proposed his lean design approach, which emphasizes customer values
and design complexity reduction. In recent years, Toyotas product develop-
ment system has been getting a lot of attention (Morgan and Liker 2006,
Kennedy 2003).
In the literature I just cited, the generic product development process
is studied from different perspectives. In this section, I am proposing a
comprehensive theoretical framework to model and analyze the prod-
uct development process. I call this framework the information and
knowledge-based product development theory. Before discussing this
theory fully, I will first discuss axiomatic design theory and a part of
Reinertsens work as necessary foundations to my theory.
2.3.1 Axiomatic Design

Motivated by the absence of scientific design principles (Suh 19842001)
tried to find a scientific basis for good design practice. He did this by
looking into many successful and unsuccessful design case studies, and
attempted to determine whether there are common features for good
design practices. At first, Suh found that there were 12 such common
features; later he found that these 12 common features of good design
can be further reduced to just two. Suh called these two common fea-
tures for good design practices axioms, so his scientific design principles
are based on these two axioms. An axiom is any sentence, proposition,
statement, or rule that forms the basis of a formal system. As an equally
important result, he modeled the design processes as mappings from
one domain to another domain. This is a very important observation,
and it helped to give deep insight into the design process. In this sub-
section, I will first discuss the mappings and domains, and then I will
discuss the two design axioms.
2.3.1.1 Design as a Mapping Process If we look into what designers do

all the time, you would find that they always treat design as a mapping
process, as illustrated by Figure 2.10.
36 Chapter Two
What we How we
want to want to
achieve achieve it
Mapping
Figure 2.10 Design as a Mapping Process
Domain 1 Domain 2
For example, if what you want to achieve is move people from one
place to another on the ground without consuming fossil fuel power,
then the design engineers will have to figure out how we want to
achieve it. The design engineers may give you a design solution of
bicycle. A product design project can be divided into many tasks from
given whats, find hows. Each of these tasks can be seen as a mapping
from one domain to another domain, as illustrated in Figure 2.10.
In Suhs point of view (2001), the design world consists of four
domains: the customer domain, the functional domain, the physical
domain, and the process domain. The customer domain is characterized
by the needs that customers want the product to fulfill. In the func-
tional domain, the customer domain elements are translated into func-
tional requirements. All products are developed to provide functions,
that is, the job the product must do to satisfy customers. Functional
requirements are specific requirements for functions. For example, in
automotive design, in the customer domain, a customer may say that he
wants a faster car. This is one customer attribute; the engineer needs
to figure out what the customer means by faster car. Finally, by com-
paring the performance of the competitors product, the engineer may
figure out that faster car may mean ability to accelerate from a speed
of 0 km/hour to a speed of 100 km/hour within 10 secondsthis is a
functional requirement. Clearly, a product may have many functional
requirements. Specifically, functional requirements (FRs) are a mini-
mum set of independent requirements that completely characterize the
functional needs of the design solution in the functional domain.
The physical domain is also called the design parameter domain.
Design parameters (DPs) are the elements of the design solution in the
physical domain that are chosen to satisfy the specified FRs. Design
parameters (DPs) mean what specific physical design can be used to
deliver all required functions. For example, if a functional requirement
for a car is to accelerate from 0 km/hour to 100 km/hour in 10 seconds, the
design parameter for that would include necessary powertrain design
specifications, including engine design, transmission design, and so on.
The process domain deals with process variables. The process domain
Y=f(x)
FRs
CNs . DPs PVs
. . . .
. . . .
. . .
Y=f(x)
Physical Mapping Process Mapping
Customer Functional Physical Process
Domain Domain Domain Domain
Figure 2.11 Four Domains in the Design Process
refers to the relevant manufacturing process specifications that can

build the production facility to produce the designed product. Process
variables (PVs) are the elements in the process domain that characterize
the process that satisfies the specified DPs. Figure 2.11 illustrates the
four domains in the design process.
Table 2.1 shows the specific contents of the four domains for different
industries.
2.3.1.2 The Independence Axiom The design process involves three

mappings between four domains (Figure 2.11). The first mapping in-
volves the mapping between customer attributes and the functional
requirements (FRs). This mapping can be performed by the means of
Quality Function Deployment (discussed in Chapter 7). The physical
TABLE 2.1 Four Design Domains for Different Industries

Domain (CN) Domain (FR) Domain (DP) Domain (PV)
Manufacturing Attributes that Functional Design Process
customers desire requirements parameters that variables that
specied for satisfy FRs fulll DPs
the product
Materials Desired Required Microstructures Processes
performances properties
Software Attributes desired Output Inputs, Machine codes,
in software specications algorithms, compilers,
of the program methods modules
codes
Organization Customer Function of the Programs, ofce, People and
satisfaction organization activities other resources
Business ROI Business goals Business Human/
structures nancial
resources
38 Chapter Two
mapping involves the functional requirements domain and the design

parameter domain (DPs). The process mapping involves the design
parameter domain and the process variable domain. Out of the three
mappings, the first mapping involves translating the raw voice of the
customer to CTQs, and translating CTQs to functional requirements.
The last two mappings, that is, the physical mapping and process map-
ping, are very challenging and important. The equation y = f(x) is used
to reflect the relationship between the domain, array y, and array x,
in the related mapping, where the array {y}mx1 is the vector of require-
ments with m components, {x}px1 is the vector of design parameters with
p components, and A is the sensitivity matrix representing the physical
mapping with Aji = y j / xi . In the process mapping, matrix B repre-
sents the process mapping between the DPs and the PVs. The overall
mapping is matrix C = AB, the product of both matrices. When matrix
A is a square diagonal matrix, that is, m = p and Aji 0 when i = j and
0 elsewhere, the design is called uncoupled; i.e., each y can be adjusted
or changed independent of the other y. An uncoupled design is a one-to-
one mapping and is represented as follows:
y1 A11 0 . 0 x1
. 0 A 22 . .
. = . . 0 .
(2-2)

ym 0 . 0 Amm xm
In the decoupled design case, matrix A is a lower/upper triangle
matrix; that is, the maximum number of nonzero sensitivity coefficients
equals p(p-1)/2 and Aij 0 for i = 1, j and i = 1,p. A decoupled design is
represented as follows:
A 0 . 0
y1 11 x1
. A21 A 22 0 . . (2-3)
. = . . .
0 .

ym A Am 2 . Amm xm
m1
The decoupled design may be treated as an uncoupled design when
the xs are adjusted in some sequence conveyed by the matrix. Uncoupled
and decoupled design entities possess conceptual robustness; that is, the
xs can be changed to affect other requirements in order to fit customer
attributes. A coupled design results when the matrix has the number of
requirements, m, greater than the number of xs, p, or when the phys-
ics is binding to a great extent, so off-diagonal sensitivity elements are
nonzeros. The coupled design may be uncoupled or decoupled by smartly
adding m-p extra xs to the problem formulation. A coupled design is
represented as follows:
x
y1 A11 A12 A1 p 1
.
.
. A21 A 22 . . (2-4)
. = .
A( m 1) p

.

ym Am1 . Am( p 1) Amp
x p
After these mathematical preparations, I can now present the first
axiom of the axiomatic design:
Maintain the independence of FRs: In an acceptable design, the DPs and
the FRs are related in such a way that a specific DP can be adjusted to
satisfy its corresponding FR without affecting other FRs.
Mathematically, the first axiom states that if a design matrix linking

functional requirements and design parameters, A, is a diagonal matrix,
then this physical mapping is uncoupled, or completely independent. In
this case, you can adjust one design parameter value to match the corre-
sponding functional requirement, independently. If the design matrix A
is a lower or upper triangular matrix, such as the one in equation (2.3),
then this physical mapping is decoupled. In this case, you can adjust the
design parameters in a fixed sequence to match corresponding functional
requirements one by one. If the design matrix is a full matrix, then the
physical mapping is coupled. In this case, you cannot easily adjust design
parameters to match the corresponding functional requirements.
Axiom 1 of axiomatic design basically states that, out of three kinds
of design mappings, uncoupled, decoupled, and coupled design, uncou-
pled design is an ideal design. The decoupled design is not as good as
uncoupled design, but it is better than coupled design, so it is acceptable.
Coupled design is undesirable. Figure 2.12 illustrates these three kinds
of design matrices. The same things can also be said for the process map-
ping, that is, the mapping from physical domain to process domain.
We will use Example 2.1 to further explain the concept of axiomatic
design.
Example 2.1 Water Faucet Design Water faucets are used in kitch-
ens or bathrooms to control the water temperature and flow rates.
From an axiomatic design point of view, the faucet has two important
functional requirements, FR1 = control the flow of water, FR2 = control
X 0 0 X X X X X X
0 X 0 0 X X X X X
0 0 X 0 0 X X X X
Figure 2.12 Three Kinds of Design Matrices
uncoupled decoupled coupled
design design design
40 Chapter Two
the water temperature. There are two popular designs for faucets used
in kitchens or bathrooms, as illustrated in Figure 2.13.
One design is called the two-handed design. In this design, the two
design parameters are: DP1 = valve angle for the hot water valve, and
DP2 = valve angle for the cold water valve. Clearly, when you adjust
DP1, that is, the hot water valve, it will affect both the flow rate (FR1)
and water temperature (FR2). The same will be true for cold water valve
(DP2). So whenever you adjust these two valves, you will simultane-
ously change both the flow rate (FR1) and water temperature (FR2).
This design definitely violates Axiom 1. From the design matrix shown
in Figure 2.13, this design is a coupled design. You need to do a lot of
adjustments to get both flow rate and temperature right.
The other design is called the one-handed design; in this design the two
design parameters are: DP1 = valve angle for mixed water (done by adjust-
ing the vertical angle of the valve), DP2 = mix ratio of cold/hot water (done
by adjusting the horizontal angle of the valve). So flow rate (FR1) can be
Functional Requirements Design Parameters

FR1: Control the flow of water DP1: Angle of valve 1, 1
FR2: Control the temperature DP2: Angle of valve 2, 2
of water
Hot water Cold water Hot water Cold water
1 2
2
Coupled Design Uncoupled Design

(DPs create conflicting functions) (DPs maintain independence of functions)
Control Flow 1 Control Flow 0 1

= =
Control Temperature 2 Control Temperature 0 2
Two-handed Faucet One-handed faucet

Figure 2.13 Water Faucet Design
adjusted by DP1 alone, and water temperature (FR2) can be adjusted by

DP2 alone. This design is an uncoupled design; its design matrix is shown
in Figure 2.13. This design satisfies Axiom 1.
Implications and Benefits of the Independence Axiom in Product

Development Practice The independence axiom will definitely favor
modular design practice, because the design parameters in different
modules are most likely to be independent of each other. In addition,
the independence axiom will also encourage maintaining parametric
independence of design parameters within each module. In applying
axiomatic design, there is a common misconception stated as follows:
We cannot use the independence axiom, because it means I have to
make many physical parts, so each part can behave independently of
the others. The truth is, a design parameter is not same as a physical
part. In a single physical part, you could have many design parameters.
Figure 2.14 illustrates a two-headed wrench that has different wrench
units on the same physical part. Both wrench units are different, inde-
pendently adjustable design parameters.
There are many benefits of applying the independent axiom:
1. Parallel developments and testing. As discussed earlier, the indepen-

dence axiom favors modular design practice. If a system is developed
based on modular design practice, it will enable parallel development
and testing of different modules, or subsystems. Modular design and
testing is illustrated in Figure 2.15. Suppose you are trying to design
and build a personal computer. A personal computer can be subdi-
vided into many subsystems, such as CPU, memory, and so on. If you
have a coupled design, that is, there are a lot of dependencies among
subsystems, then changing the design of a subsystem will affect the
design of other systems. In this case, you have to design one subsys-
tem at a time. Only after you freeze the design of one subsystem can
you start the next. This will give you a lengthy and sequential design
process. If you apply the independence axiom and adopt a modu-
lar design practice, you can develop several subsystems in parallel,
which can speed up the product development process greatly.
2. Easier engineering change. Engineering change is a situation where
the goals, objectives, and functional requirements of a product will be
forced to change due to a variety of causes, such as market changes,
government regulation changes, and so on. Clearly, an independent
Figure 2.14 A Two-Headed Wrench

42 Chapter Two
CPU Coupled System, sequential design
Memory
System Memory
CPU
System
I/O Board
I/O Board
Control
Software User
Define Control
Interface
Architecture Software
User Software
Interface
Software Power
Supply
Power
Supply
Package
Package
Modular system, parallel design

Figure 2.15 Coupled System vs. Modular System
design will have a simpler relationship between functional require-

ments and design parameters. When the functional requirements
have to change, it is much easier to gure out how you have to change
design parameters of uncoupled designs to match the new functional
requirements than that of coupled designs.
3. Simpler operation. An independent design will also have a sim-
pler relationship between adjustment factors and system func-
tion, because they are a part of functional requirements and design
parameters. Example 2.1 illustrates this fact quite well.
2.3.1.3 The Information Axiom The second axiom of axiomatic design

deals with the complexity in design. The formal statement of Axiom 2
is as follows:
Minimize the information content: Among alternative designs that satisfy
Axiom 1, the best design has the minimum information content.
In Suhs Principles of Design (1990), information content is defined
as the total number of bits of information that is needed to store the
complete information for the design. In recent years (Suh 2001), the
definition of information content by Suhs school is mostly focused on
the probability of success for delivering all the FRs. Suh used the fol-
lowing entropy model to define the information content:
The information content, that is, entropy Ii, for a given FRi, is defined
as follows:
1 (2-5)
Ii = log 2 = log 2 Pi
Pi
Where Pi stands for the probability for FRi to be delivered successfully.

For a group of m independent FRis, the success probability for the
whole system, P, is
m
P= Pi (2-6)
i =1
Then the total information content, that is, total entropy I, is
1 m m
I = log 2 = log 2 P = log 2 Pi = log 2 Pi (2-7)
P i =1 i =1
Intuitively, Axiom 2 basically states: if you have several alternative

designs, and every design can deliver the functional requirements and
satisfy the independence axiom, then the simplest design is the best
design.
In recent axiomatic design practices, however, most real-world case
studies are related using Axiom 1. The information contents defined
by equations (2.5) through (2.7) are difficult to quantify in practical
circumstances.
I believe that Axiom 2 is a very fundamental and powerful design
principle, and more comprehensive and practical complexity metrics
should be developed to foster the application of this axiom. Later in
this chapter, we will use Axiom 2 to guide the lean product development
process.
2.3.2 Design as an Information

Production Factory
In his excellent book Managing the Design Factory (1997), Reinertsen
stated that the only purpose of the design process is to produce useful
information economically. In this book, information is what design
engineers are producing in their work, such as drawings, test results,
specifications, and so on. Reinertsen also used entropy to describe the
information as follows:
1 (2-8)
I = log 2
Pe
44 Chapter Two
where Pe is the probability for a event to occur. Clearly, I will be large

if Pe is small. Reinertsen stated that if an event has a very small prob-
ability of occurring, but it does occur, then this event will contain a lot
of information. Specifically, Reinertsen applied this entropy formula to
engineering test applications as in equation (2.9):
ITest = PFailure IFailure + PSuccess ISuccess (2-9)
where ITest stands for the information contents of a test, PFailure and
PSuccess refer to the probability that a test is a failure or a success, respec-
tively, and IFailure and ISuccess are the corresponding information contents
for a failed test, or a successful test, respectively.
By analyzing equation (2.9), Reinertsen stated that if a test has a high
probability of failure, and the real test result is a failure, then the infor-
mation content in this test is low, because you dont learn much from this
test. On the other hand, if a test has a high probability of success and the
actual test does yield a success, the information content is also low for the
same reason. On the contrary, if a test has a high probability of failure, but
the actual test yields a success, based on equation (2.8), it can be proved
that the information content of this test is high. This makes sense because
a success in a high-risk test contains some rare information. Similarly,
if a test has a high probability of success, but the actual test yields a
failure, then this test must have high information content because it will
shed some light on rare failure causes. However, for most tests, if the
probability of failure is high, most likely, the test will yield a failure, and
if the probability of success is high, then most likely the test will yield
a success. So high success-rate tests and high failure-rate tests are not
likely to provide a lot of information. Mathematically, it can be proved
that if the failure probability of a test is 50 percent, (so the probability of
success is also 50 percent), then the expected information content in this
kind of test will be the highest. Figure 2.16 illustrates this fact.
Based on this analysis, Reinertsen further proposed that in engineer-
ing design practice, you should control the risk level for each project at
around 50 percent. That is, a project, or a test, will have an equal prob-
ability of success or failure, assuming you can learn useful lessons from
both successes and failures. A low-risk level strategy will make learning
things too slow, because you dont learn much from a successful project
if the project is very easy to do. A high-risk-level strategy will also slow
down the learning process, because each success is so hard to get that
you will be buried under a pile of failures. Obviously, this medium-risk,
medium-difficulty level learning strategy makes a lot of sense and it
will enable us to gain information at the highest speed.
Reinertsen also stated that information has time value; the earlier
you get the information, the better. If you get the information too late in
the product development process, the necessary design change will be
Information
Figure 2.16 Failure Probability and

0 0.5 1.0 Information Content
Failure Probability
very difficult to implement. In Reinertsens point of view, early informa-

tion can be obtained by conducting small-scale tests early in the product
development process. He also stated that not every piece of information
has the same valuesome information is critical, some information is
not as critical. In the product development process, you would like to
get critical information earlier than noncritical information. This fact
forms the basis for design job sequencing.
2.3.3 Information and Knowledge Mining

What is the nature of the product development process? What are the
key factors for a superior product development process? How can you
improve your product development process? We are trying to dig out
answers for these questions in this subsection.
The answer to the question what is the nature of the product develop-
ment process? can be found by looking into what is the final result of the
product development process? What are design engineers doing every
day? In general, what is the extended product development team (includ-
ing marketing, management, research and development scientists, and
technicians) doing every day? They are creating documents, compiling
testing reports, doing design analysis, drafting graphs, calculating survey
data statistics, creating specifications, building prototypes, designing and
making tools for producing the product, and developing assembly opera-
tions. In general, they are generating all kinds of information and knowl-
edge. When are their jobs done? Their jobs are done after the production
people have obtained enough useful information to produce the product
effectively, reliably, economically, and with good quality, and the products
46 Chapter Two
shipped to customers are free of after-sale problems. Clearly, the nature of

the product development process is an information generation factory.
2.3.3.1 Defining Information and Knowledge Data, information, and

knowledge are closely related. Data consists of numbers, characters,
images, and signals. Data is usually unorganized. Information is the
result of processing, manipulating, and organizing data in a way that
creates meaning to the person receiving it.
However, data and information alone cannot fully explain the prod-
uct development process. Without human involvement, the information
is static and not dynamic. At least for now, you cannot build a 100 percent
computerized product development system that just takes care of itself
with zero human involvement. Knowledge plays an important role. There
are many different definitions of knowledge. According to Ballard (1987),
knowledge is information plus theory; other researchers define knowledge
as information combined with experience, context, interpretation, and
reflection. It is a high-value form of information that is ready to apply to
decisions and actions (Davenport et al. 1998). In my point of view, knowl-
edge is a personal or organization-based information-processing mecha-
nism; knowledge is a combination of an information bank, interpreter,
organizer, enabler, improver, and creator; knowledge should represent
truth, and it should be actionable. All of this is illustrated in Figure 2.17.
A product development organization can be modeled as shown in
Figure 2.18. Inside the product development organization, there are
many knowledge cells, which can be individual professionals, and many
knowledge centers, which can be teams. All kinds of information flow
around, into, and out of knowledge centers and cells. Each knowledge
center and cell will process information and create modified or new
information after processing.
In the product development process, there are three dominant types
of information processing: information mining, information transforma-
tion, and information/knowledge creation.
2.3.3.2 Information Mining Information mining is the extraction of valu-

able information from raw information sources. For example, in research
and publication work, the literature survey is information mining work.
The raw information sources include professional journals, the Internet,
professional conferences, and leading professionals. The process of a lit-
erature survey is to extract relevant and high-quality information, and
Modified
Information Knowledge and/or New
Information
Figure 2.17 Knowledge and Information

Information Flows
Knowledge Centers
Knowledge Cells
Product Development Process as a Connected Network of
Figure 2.18
Knowledge Centers and Cells
then process and compile it. The literature survey report is the extracted,
condensed information. Similar to metal that has been mined as com-
pared to ore, the literature survey report will serve as a stepping-stone
for further research.
In the product development process, there are two major types of
information mining work: mining of the voice of the customer, and
mining of technological information. Figure 2.19 shows how these two
types of information mining fit into the product development process. In
addition, as the figure shows, the voice of the customer can come from
both external and internal customers. In general, the steps in informa-
tion mining are as follows:
1. Prospecting to locate information sources

2. Exploration to nd and then dene the extent and value of informa-
tion sources
3. Exploitation to extract information from information sources
4. Processing and rening the extracted information for the informa-
tion users
Information Mining of the Voice of the Customer According to the

theory of axiomatic design, product development is a sequence of map-
ping processes. The very first step is mapping from the customer domain
to the function domain; this step is really a rigorous product definition
step. The whole value of the product is largely determined by whether
the product will be welcomed by potential external customers, or buyers.
48 Chapter Two
Internal Customers
Y=f(x)
FRs DPs
External CAs PVs
.
Customers .
.
. . .
.
. . .
. Y=f(x) .

Domain Domain Domain Domain
Technical Information Sources
Mining of Technology Information
Mining of Voice of Customer

Figure 2.19 Information Mining in the Product Development Process
If customers dont want your product, it will be a failure. So the ability

to accurately mine potential customers minds is critically important in
the product development process. There is a saying, garbage in, garbage
outif you totally ignore what customers want, or you did not do a
superb job in discovering what customers really want, your very first
input to the product development process might be totally flawed, and
whatever you do in the subsequent steps, you will produce a garbage
product. It is not an exaggeration that accurately capturing the voice of
the customer is like striking gold.
There are two kinds of customers: external customers and internal
customers. The external customers are the customers who buy or use
your products. The internal customers are people inside your company
who are the users of your work. For example, production people are the
customers of the product development team, because they are the down-
stream users of product design and manufacturing process design.
The specific steps in mining the voice of the external customer are
described as follows:
1. Prospecting to locate information sources. The most important infor-

mation source is the prospective real customers. You have to get most
of this information from the people, not from the literature, electronic
media, and so on.
2. Exploration to nd and then dene the extent and value of informa-
tion sources. At rst, you need to determine which customer group
and what kind of customers you need to locate, and then how to
select sample customers to work with. The value of the information
source is similar to the value of ore in metallurgythe value of the
information depends on how much valuable information you can get
from this source, and how easy it is to extract the real valuable infor-
mation. For example, lead users and competitors customers might
be really valuable information sources. The lead users can provide a
lot more in-depth opinions than regular users, and provide informa-
tion on the future market, because what lead users are doing with
the product might be the mainstream application of the product in
the near future. The competitors customers can provide information
about why some people dont like your products. This information is
very difcult to get from your existing customers.
3. Exploitation to extract information from information sources. This
step deals with what kind of information you really want to get, and
how to extract high-quality information from customers.
For the first question, the purpose of extracting the voice of the cus-
tomer is to support the whole product development process. For example,
you need to get enough VOC information to derive full sets of product
functional requirements, based on Figure 2.19. Common problems for
this step include:
You did not get the right type of information.
You did not get enough information or clear enough information to
support your job.
For example, suppose that I eat in a restaurant and I am given a cus-

tomer survey form to fill out. However, all the questions are something
like: Do you like our food? or Do you like our drink? and ask me to fill
in numerical ratings. How much can this information help to improve
the service? If I fill in the food is bad, then the question is, which menu
item is bad? And how can it be improved? This response doesnt provide
the kitchen with the right type of information, or enough information, to
improve the food. Chapter 3 of this book will provide detailed informa-
tion about what kind of information do I really want to get?
For the second question, how do you extract high-quality informa-
tion from customers? you first need to know what defines high-quality
information. In my opinion, high-quality VOC information should have
some of the following characteristics:
It accurately reects what customers really want deep in their hearts,

not just what they say.
It uncovers some unmet customer needs, or unarticulated customers
needs in the current marketplace.
50 Chapter Two
It provides a basis for predicting future or near-future market trends

and needs.
It provides accurate and sufcient information to feed the product
There are many effective methods for extracting the voice of the
customer, such as customer surveys, focus groups, and ethnographic
approaches. These methods will be discussed thoroughly throughout
this book.
The information users of the voice of the customer are the product
development team, mostly design engineers. So they are the customers
of the people who collect the VOC information. At first, you need to know
what kind of information the product development team really needs, so
that the people who collect VOC information will collect relevant and suf-
ficient VOC information. Secondly, the VOC information should be trans-
lated into language that information users will understand. Chapter 6 of
this book will discuss how to process VOC data. Quality function deploy-
ment (QFD) is also an effective tool to relate the voice of the customer to
design requirements. QFD will be discussed in Chapter 7.
Mining of Technology Information and Knowledge Whether you de-

velop a technology push-type product or a market pull-type product,
you need to have technological know-how and information to make the
voice of the customer into reality. In this information mining work, the
quality and speed of the information mining are very important.
The quality of this information mining means that you are getting
the best technology in terms of performance, the cost of the technology,
and the robustness of the technology. The performance of the technology
means that this technology can create the right functions, and it per-
forms better than other competing technologies. The cost of technology
means that its value-to-cost ratio is high; you should not adopt technol-
ogy that has top-notch performance but is prohibitively expensive. The
robustness of the technology is also very important; robustness means
that this technology can deliver its functions consistently under vari-
ous usage conditions and does not require excessive operator or user
requirements. Many very new technologies have excellent performance,
but their technical bugs have not been fully cleaned up yet; introducing
too much immature technology into the product could create potential
quality and reliability problems.
The speed of this information mining is also critically important; the
ideal technology information mining process has the ability to provide the
right technology information to the right people at high speed. In many
companies, this information mining work is a slow and ineffective process,
with a lot of unnecessary information and a lot of reinvention. Later in
this chapter, we are going to discuss how to improve this process.
The specific steps in mining the technology information are described

as follows:
1. Prospecting to locate technology information and knowledge sources.

As discussed earlier, information is organized data that delivers
meaning. Knowledge is the distilled information that is understood,
believed to represent the truth, and actionable.
The sources of this kind of information could be internal and exter-
nal. Internal sources of information include:
Company databases, company libraries, and document
storage This type of information is relatively easy to retrieve.
Everywhere in employees computers, binders, and docu-
ment folders Usually, this kind of information is difcult to
retrieve; people dont even know where it is. This information is
constantly missing, falling into cracks, misplaced, and distorted.
In the product development process, a lot of time and manpower
is lost in retrieving and re-creating this kind of information. I will
discuss how to improve the management of this information source
in a lean product development process.
The external sources of information include:
Suppliers Many companies contract out a lot of product devel-
opment tasks, so the suppliers have a lot of this technology
information.
Universities and research labs
Individual researchers and small-sized providers This
includes a lot of new startup research-oriented companies, consul-
tant companies, and even individuals.
Public domain, such as publications, patents, and the
Internet
The sources of knowledge are more complicated, because you cannot

isolate the knowledge from the people who have the knowledge.
Based on Nonaka and Takeuchi (1995), there are two kinds of knowl-
edge: explicit knowledge and tacit knowledge. Explicit knowledge is
knowledge that can be explicitly expressed in a stand-alone form,
such as oral language, written language, computer codes, mathemat-
ical formulas, and/or visual media. Written reports, presentations,
research papers, books, instructional manuals, procedures, sound
and visual records, and software are examples of explicit knowl-
edge. On the other hand, tacit knowledge is knowledge that cannot
be expressed explicitly in a stand-alone form. Tacit knowledge is
buried deeply inside people and it is difficult to express and transmit
52 Chapter Two
it explicitly. However, it usually can be transmitted only through

a lot of people-to-people interaction. The simplest example of the
nature and value of tacit knowledge is that one does not learn how
to ride a bike or swim by reading a textbook, but only through per-
sonal experimentation, by observing others, and/or being guided by
an instructor. Similarly, nobody can learn NBA basketball skills by
reading or by watching audio/video media. In product development
and in manufacturing in general, a significant portion of knowledge
is of this typechief engineers, high-level technical experts, experi-
enced senior managers, and high-level technicians all possess a great
deal of tacit knowledge.
The sources of explicit knowledge are very similar to the sources
of information that we just discussed. The sources of tacit knowledge
are more complicated, including:
The companys own valuable employees
High-caliber new hires from outside the company
Training and coaching practice through internal and external
sources
2. Exploration to nd and then dene the extent and value of information
sources. A world-class company needs an effective information and
knowledge management system. This management system includes
the personnel department, training, the senior management of prod-
uct development, IT support, and so on. An ideal system would be the
system that hires the right people, conducts the right training at the
right time, and acquires the right external technical information and
explicit knowledge. In many companies, however, there is enormous
waste in this information and knowledge acquisition processfor
example, the massive layoff of experienced personnel by American
automobile manufacturers in recent years had an enormous cost in
product development capabilities. The attitude that engineers and
technical specialists are disposable commodities is very harmful to
information and knowledge management practice, because the right
kind of tacit knowledge takes a lot of time, experience, and manage-
ment to create.
3. Exploitation to extract information from information sources. This
step deals with the question of what kind of information do I really
want to get? and how to extract high-quality information from the
information sources. A technology roadmap (Phaal et al. 2001) can
be used to support this step.
4. Processing and rening the extracted information for the information
users. The users of this information are the product development
team, mostly design engineers. The people who acquire and process
this information are also often the product development team.
The quality of this step highly depends on the quality of product
development engineers.
2.3.4 Information Transformation

As you can see, information mining is a key part of the product devel-
opment process; however, the mined information is only the input to
the design process. In the design process, it is only the first step; you
need to transform this input into all necessary design specifications.
I call this process the information transformation process. Information
transformation is similar to the mappings in the axiomatic design. In
the fuzzy front end of the product development process, you are mostly
doing information mining, mining of VOC information, and mining of
technology information. You do some information transformation in the
fuzzy front end; this information transformation is basically process-
ing, refining, and clarifying the raw voice-of-the-customer information.
Based on axiomatic design, there are three important mappings in the
product development process. The first mapping is the transformation
of customer attributes into product functional requirements. The second
mapping is the physical mapping, which is the transformation from
functional requirements to design parameters. The third mapping is the
process mapping, which is the transformation from design parameters
to process variables.
Another kind of information transformation is working from the high-
level system to subsystems, to components, and finally to all detailed
specifications. In system engineering, this transformation process is
also called flow down.
Unless there is some information and knowledge creation work, that
is, creating new ideas, new technology and so on, most of the informa-
tion transformation work deals with existing knowledge. For example,
in automobile product development, body design and assembly is a big
chunk of design work. Body styles have to change to make cars catch
current fashion trends. But there is very little new knowledge needed
in this design workstyle designers design the body style, then the
stamping dies have to be made, then stamped metal panels are welded
to form subassemblies, and then subassemblies will be welded to form
the whole automobile body-in-white (body without paint). Though there
is no new technology or knowledge involved, simply the new combina-
tion of the parts and the new shape of the parts, there will still be a lot
of unknowns to be worked out. In this automobile body design example,
you still need to work out all the issues about fit and finish, that is, can
these subassemblies fit together seamlessly?
54 Chapter Two
The types of work in this information transformation usually

include:
Given a need, pull design solutions. For example, if you need to bind
parts together, then we can pull many kinds of fastener solutions. This
category is very broad; selection of material, selection of modules, and
so on all belong to this category.
Design of interfaces
Shape and form design
System ow down and integration
Design analysis and simulation
Testing
Prototype building
For complicated products with thousands of parts or more, the scope of

this information transformation work can be quite extensive. You need
a whole organization with many people, with different tasks, different
knowledge backgrounds, and different experience to work together, as
illustrated in Figure 2.18. The information and knowledge need to flow
from one team, or one person, to another team or teams, another person,
and so on. There are several potential dangers in this knowledge and
information flow network:
With the ow of information from one place to another, the informa-

tion can be misunderstood and distorted. Because people have differ-
ent backgrounds, a lot of meaning can be lost in the translation.
The information may not ow as it should. Some teams may have done
some test, or some analysis, and made some design changes, but the
people on other teams didnt get this information. This will create a
lot of technical bugs and inconsistencies in the design.
The ow of the information could be very slow, or very ineffective.
Based on the preceding discussion about this information transfor-

mation process, we can summarize the following key factors for a good
process:
A good bank of information and knowledge, from which people can pull
all the necessary information and explicit knowledge to create design
solutions as needed. This information bank should be a live one that
keeps improving and enriching itself. This information and knowledge
bank can exist in many different formats. It can be a well-managed
computer information system with a nice structure, containing high-
quality information that is constantly updated with any new informa-
tion by product team members. It can also be a war rooma big room,
in which all important design engineers and managers are present, with
a good best-practice database. This best-practice database records the
best design practices for all kinds of parts, subsystems, and so on, as well
as best testing practices. These practices are mature, veried practices
that have very low technical risks. A design engineer can pull solutions
either by asking other engineers in the war room, or pulling them from
the best-practice database. This combination of war room and best-prac-
tice database is one of the key features of Toyotas product development
process.
Well-managed training programs for product development team mem-
bers. The training should enable the engineers to know all necessary
methods and tools, and learn best design practices, best design prin-
ciples, teamwork, and so on.
Good design support tools and methods. One way to preserve knowl-
edge is to build the knowledge into computer packages and similar
systems. For example, the statistical software MINITAB contains so
much knowledge in statistical methods that you can pull and use
any of those methods very readily. There are many counter-examples.
There are many fancy research consortiums with many participants
from top-notch universities and multiple companies, which may create
a lot of new knowledge during their operations. However, after the
projects are done, this hard-earned knowledge usually gets lost very
fast. Many methods can help this information transformation work
quite well, such as design of experiments and robust design. These
methods can accelerate the design improvement process and work out
the technical bugs very effectively.
Application of good design principles, such as axiomatic design, system
engineering, and modular design. You could design every part per-
fectly, but if the system integration is bad, you still create very com-
plex vulnerable designs. In a later part of this chapter, I will discuss
some details of these good design principles.
A well-designed and well-managed product development process. Just
like any product, the product development process is also a man-made
process that needs to be well designed. This process needs to be well
managed as well. This product development process should have the
following features:
A customer-valuebased design process You need to mine the
voice-of-the-customer information adequately and accurately, and
translate that VOC into correct engineering requirements.
A front-endloaded process The earlier you get the critical infor-
mation, the better, so you should work more in the concept develop-
ment stage. I will discuss detailed approaches for this later in this
chapter.
56 Chapter Two
A simple and easy-to-manage process This product develop-

ment process should enable high-volume, accurate, and fast infor-
mation ow across the product development organizations. This
product development process should enable the jobs and projects
to be performed smoothly, quickly, and with high quality.
2.3.5 Information and Knowledge Creation

In the product development process, after youve done all the required
information mining and transformation, you may find that there are
still some design tasks for which no ready solutions can be pulled from
anywhere. These tasks involve creating new information and new
knowledge. Here are some of the scenarios:
Resolution of some technical bottlenecks that nobody has

accomplished before For example, the fuel efciency of the inter-
nal combustion engine is low, and with increasing petroleum prices,
this technical difculty needs to be resolved.
Development of the new generation of product You want to
drastically improve your products performance, cost, and so on in
order to move ahead of competition. This improvement is not merely
a ne-tuning of existing product.
Develop a product with a new marketing concept For example,
the development of sport utility vehicles in the 1980s and 1990s.
Technology pushes product development Many research
results are coming out from universities, research institutions, and
so on, and many patents are created every year. Manufacturers are
bringing these new technologies into their products.
The following list summarizes some common difficulties in this infor-

mation and knowledge creation work:
Difculties in nding new technical solutions

Excessive cost and lead time in developing new technical solutions
Customers rejection of the new product
Excessive time in ironing out technical bugs and wrinkles
Quality and reliability problems due to immature technologies
There are many methods and approaches to overcome these difficul-

ties; Ill discuss some of these methods and approaches in the following
sections.
2.3.5.1 Theory of Inventive Problem Solving (TRIZ) I discussed TRIZ

earlier in this chapter. Basically, the TRIZ method is based on the
studies of patents and inventions. The basic idea of TRIZ is to reuse the
ideas from similar patents and inventions to solve your own problems.
Because the similar patents and inventions are verified ideas and proven
to work, the solutions derived from TRIZ may have a higher probability
of success than ideas drawn out of the blue. TRIZ is used to resolve the
difficulties in finding new technical solutions. TRIZ will be discussed in
detail in Chapter 10.
2.3.5.2 Knowledge Creation Spirals Knowledge creation not only in-

volves idea generation, but also involves gaining deeper understanding,
perfecting the concepts, ironing out technical bugs, and extending the
successful new design concepts to other products. In their book The
Knowledge-Creating Company, Nonaka and Takeuchi (1995) stated that
a successful knowledge management and knowledge creation system is
crucial for success for many Japanese companies. As I mentioned previ-
ously, there are two kinds of knowledge: explicit knowledge and tacit
knowledge. A good knowledge management program needs to convert
internalized tacit knowledge into explicit codified knowledge in order
to share it, but also for individuals and groups to internalize codified
knowledge and make it personally meaningful once it is retrieved from
the knowledge management system.
In a traditional information system, tacit knowledge does not get
enough recognition because it is difficult to codify and share it. However,
tacit knowledge is very valuable and it shouldnt be wasted. Polanyi
(1967) is one of the pioneers of tacit knowledge. One of Polanyis famous
aphorisms is: We know more than we can tell.
Even for engineers and researchers, tacit knowledge is everywhere. If a
group of people takes an identical set of training courses, their actual work
performances will differ. You cannot reproduce a top scientist by reading
exactly the same material, taking exactly the same training, and working
on exactly the same projects that the scientist did. The top scientist will
always have some more hidden and difficult-to-share knowledge.
The creation of human knowledge usually goes through a spiral as
Explicit Knowledge
At Next Level
Tacit Knowledge
At Next Level
Explicit
Knowledge
Figure 2.20 Knowledge Creation
Tacit
Spiral
Knowledge
58 Chapter Two
For example, a person starts a job as a quality technician. He works

in a production line, making measurements, keeping data, and observ-
ing what people are doing in the plant, so he starts to find out what it
feels like to be a quality technician. This is the start of a tacit knowledge
buildup. After he works for a year, he has gained a lot of insight, such as
how the situation in the production line keeps changing; sometimes the
quality is good, sometimes it is bad, especially when incoming material
has changed, or operators change in shift. The data may fluctuate a lot,
even when everything works smoothly. After one year of work, he starts
his Masters program in a local university, and takes a lot of courses,
such as statistical process control, quality management, and so on. After
his training, he fully appreciates the explicit knowledge he got from the
courses; now, not only he can do measurement and plot the data, but also
he can perform analysis, do some troubleshooting, and lead some quality
improvement projects. Also, after he fully appreciates and understands
some advanced statistical concepts, he starts to perceive the data differ-
ently; he can sense something such as a mean shift, excessive variation,
or lot-to-lot differences, which never crossed his mind before. This step is
the jump from explicit knowledge to tacit knowledge of the next level.
In Nonaka and Takeuchis book (1995), this spiral-of-knowledge-
creation mechanism can be used to extend personal knowledge into team
knowledge, and gradually into organizational knowledge. There are four
kinds of knowledge creation spirals as shown in Table 2.2 (Nonaka and
Takeuchi 1995).
In Nonaka and Takeuchis point of view, tacit and explicit knowledge
are two sides of a coin, and they are complementary to each other. In
addition, both tacit and explicit knowledge can be at a personal level,
as well as an organizational level. Since product development is team-
work, it is essential to spread knowledge from the personal level to the
organizational level. Nonaka and Takeuchi believe that there are four
modes of knowledge conversion, that is, socialization (from individual
tacit knowledge to group tacit knowledge), externalization (from tacit
knowledge to explicit knowledge), combination (from separate explicit
knowledge to systemic explicit knowledge), and internalization (from
explicit knowledge to tacit knowledge).
Socialization is a process of spreading personal tacit knowledge to
organizations or other individuals. This process is often characterized
TABLE 2.2 Knowledge Creation Spirals

To tacit knowledge To explicit knowledge
From tacit knowledge Socialization Externalization
From explicit knowledge Internalization Combination
by well-planned and organized people-to-people interaction. For

example, a top chef does show and tell to his coworkers to share
his tricks in cooking. Externalization is a process of transforming
tacit knowledge into explicit knowledge by observing, analyzing, and
coding the tacit knowledge. This transformation can be in the form of
person to him/herself, person to person, or person to a team. Observing
and digitally recording NBA players play, extracting their playing
patterns, and coding them into a computer game is such an example.
Combination is a process of combining many individual explicit
knowledge pieces into a system of explicit knowledge. One example
of combination is to compile a quality control handbook by combining
many new individual method modules into an existing handbook.
Internalization is the process of transforming explicit knowledge into
tacit, operational knowledge. Using standard operating procedures
to train new employees in real work environments is an example of
such internalization.
2.3.5.3 Customer-Centric Innovation Creativity is about the generation

of new ideas. However, new ideas may not always be good ones in the mar-
ketplace. Innovation is about making new ideas into commercial success.
Therefore, creativity and knowledge creation have to work hand-in-hand
with the voice of the customer. This approach is called customer-centric
innovation. Ethnographic methods and testing markets are powerful
tools to support customer-centric innovation. Ethnographic methods are
thoroughly discussed in Chapter 5 of this book.
2.3.5.4 Robust Technology Development Robust technology develop-

ment means building robustness into newly developed generic technol-
ogy, or new technology at its infant stage. The examples of such new
generic technology include new memory chips, new electronic bonding
technology, new materials, and so on. New technologies are usually de-
veloped at research laboratories with ideal conditions, small batches,
and on a small scale. After a generic new technology is developed, prod-
uct developers will try to integrate it into new products. But usually
there are a lot of hiccups in this integration process; the new technology
that works well in the lab may not work well after integration, and its
performance may not be stable and up to peoples expectations. It usu-
ally takes a great deal of trial and error to make it work right.
Robust technology development is a strategy that tries to reduce the
hiccups and make new technology integrate faster with products and
production. Robust technology development proposes to conduct robust
parameter design on the new technology when the new technology is
still in the research lab. Robust technology development is discussed by
Yang and El-Haik (2003).
60 Chapter Two
2.3.5.5 Kaizen for Product Development In the product development

process, if your design involves new technology, new parts, new inter-
faces, and so on, there will be some technical bugs and wrinkles. These
bugs and wrinkles will bring some quality and reliability problems
down the road. Therefore, fixing all the bugs and ironing out all techni-
cal wrinkles quickly is a critical factor to achieve quick product develop-
ment lead time and low product development cost. In addition to some
systematic methods such as robust design and design of experiments, a
steady step-by-step improvement effort, such as Kaizen, can also play a
big role in bringing out a quality product in a short time. Product devel-
opment Kaizen has been used effectively in such Japanese companies
as Toyota and Honda, and has achieved great success. One simple ap-
proach is visual management. In this approach, a designated area inside
the company is devoted to visually displaying the status of progress for a
product development project. This area is often called a project control
room (Oobeya in Japanese). In this project control room, a big product
development project is subdivided into a hierarchy of subtasks of differ-
ent levels, and wall spaces are color-coded and are allocated to subtasks
so that each project subarea can post its own dashboards and project
reports. In each dashboard, key tracking metrics for product develop-
ment tasks, such as task progress versus schedules, project tracking
data, and testing results are posted. This project control room approach
is a center of project information sharing and learning. People can learn
from each other, and duplicated tasks or overlapping work can be easily
discovered and eliminated. Both the management and team members
can see the big picture and discover the bottlenecks in the product de-
velopment project. This approach helped Japanese auto manufacturers
to cut the lions share of product development lead time and cost.
2.3.6 The Ideal Product Development

Process
In the theory of inventive problem solving (TRIZ), there is an approach
called the ideal final solution. The ideal final solution is an extreme
description of what an ideal product is. For example, what is an ideal
pen? The ideal final solution for an ideal pen would be: A mark-making
device that can make any kind of mark you wish, make any change
you wish, without trace of corrections. This device uses no resources,
produces no harm and no waste, and costs no money. This kind of
ideal final solution may sound unrealistic, but it provides an ultimate
goal. By comparing the ideal final solution with your current solution,
you can find the gaps in performance and visualize where you need
to improve.
2.3.6.1 The Ideal Final Solution For the product development process,
what would be the ideal final solution? We can easily list the following
features:
The customers reception of the product is overwhelming! It is absolutely
a hit! Everything is designed exactly the way the customer wants!
The product performance is absolutely amazing; it beats all the com-
petition right there!
The cost is extremely low, much lower than all competitors cost!
The product is so unique that nobody else delivers a similar product.
It is the rst and only one in the market; it will take competitors
tremendous effort to catch up.
The quality and reliability is the best in the class; it is much better
than what customers are expecting.
2.3.6.2 Reality Gaps Obviously, this kind of product does exist, but
there are very few. Based on the previous description of the nature of the
product development process, and based on the history of many success-
ful and unsuccessful product development stories, you can figure out
what can go wrong in each of the five categories in the preceding list:
1. Customer reception is low. This means that you didnt mine the voice
of the customer in a perfect manner. You may have gotten the wrong
voice of the customer, or you may have gotten an incomplete voice of
the customer. You may have gotten only what customers know how
to say, and did not discover the hidden and unmet customer needs.
You may have paid attention to minor concerns, and lost the big pic-
ture. You may have predicted the right voice of the customer at this
moment, but failed to predict the voice of the customer in coming
years. Most important, if you captured the wrong voice of the cus-
tomer, no matter how well you do in product performance, quality, and
reliability, you designed the wrong product and very few people will
want it, unless you sell it at such a low price that it is almost free.
2. Product performance is inadequate. Assuming you mined the right
voice of the customer, the low performance is caused by a awed
information transformation process and information and knowledge
creation process. You did not translate the voice of the the customer
into the best possible design solutions. Specically, you can easily list
the following possible aws:
a. You dont have a good enough bank of knowledge management
and information; the design solutions that you pull are far from
the best.
62 Chapter Two
b. The design team is inadequately or incorrectly trained, so they

dont know the necessary tools and methods, and do not apply best
practices.
c. The design support tools and systems are poorly managed. People
use all kinds of software, tools, and buzzwords, leading to a lot of
incompatibilities and misunderstandings.
d. Poor information ow. Communications among people are poor,
and meetings are unproductive.
e. Poor design architecture. Flawed design principles, excessive com-
plexities, too many custom-made parts, and so on.
f. Poorly managed product development process. Everybody is busy,
and there is a lot of re-ghting, too many rushed, half-cooked
projects.
g. Brain drain. People hop jobs too often; you need to retrain people,
and reinvent wheels all the time.
h. Lengthy and ineffective quality improvement process. Too much
money and time is spent in debugging the new products.
3. Excessive cost. As discussed earlier, there are three major cost com-
ponents for the product:
a. Product development cost. Excessive cost in product development
is usually caused by an inefcient product development process.
b. Production cost. Excessive product production cost is usually caused
by poor product design and poorly run production systems.
c. Cost of running supporting business processes. Excessive business
cost is usually caused by waste and an inefcient business process.
Lean operation practices could be used to resolve this.
4. Excessive competition in the marketplace. You could design your
product too similar to your competitors product, so you have to
endure brutal competition. You are struggling with either competing
on cost or competing on performance. There is a product development
strategy called the blue ocean strategy, which calls for carefully
examining your market position and your customer value proposi-
tion. This strategy has a roadmap of revising your product value
position to break out of this brutal competition. I will discuss this
blue ocean strategy in Chapter 3 of this book.
5. Inadequate quality and reliability. The common causes for inade-
quate quality and realiability include the following:
a. You did not debug the product adequately.
b. The product design is awed or it is too complex.
c. There are too many immature technologies.

d. The product development process lacks discipline, and has too
many holes.
e. There is poor quality control on suppliers.
f. The product design lacks integrity.
2.3.6.3 Deriving the Ideal Product Development Process I have dis-

cussed extensively the nature of the product development process,
which consists of information mining, information transformation,
and information and knowledge creation. The mining of information
includes mining voice-of-the-customer information and mining techni-
cal information. The best mining process is to mine the gold! That is,
you have to mine from the right customer, from the right source, obtain
the right voice of the customer, and get adequate VOC information to
populate the product requirements and figure out the core values of
customers. The purpose of information transformation is to pull VOC
information and technical information and effectively transform it into
perfect product design solutions. So the best information transforma-
tion process is a process that has the highest speed and quality in pull-
ing both the VOC and the technical information and transforming it
into designs. The purpose of information and knowledge creation is
bringing out innovative and breakthrough design solutions. The best
such system should be able to bring out innovation and breakthrough
solutions of the highest quality and at the highest speed. Within a prod-
uct development system, there are many people and teams, so there are
a lot of information flows. The purpose of information flow is to move
information to the right people who need it. The best information flow
is the one that delivers information accurately, smoothly, at the right
time, and in the right amount when people need it.
The product development process is an information creation process.
The consumer of this information is the product. If you design a high-
end, complicated product such as commercial airplanes, it consumes a
lot of information. You need an enormous amount of design, drafting,
testing, simulation, and tooling design to generate enough information
so that you know how to build the aircraft successfully. Axiom 2 of the
axiomatic design principles says that the best design is the design that
delivers all product functions and has the lowest possible information
content. It makes a lot of sense here. For example, in an automobile
design project, if you end up using a lot of specially-designed parts,
extremely high part counts, and a lot of new and immature technologies,
then you need to spend a lot of engineering hours to design, analyze,
test, and troubleshoot these parts and subsystems. Overall, you need to
generate a lot of information in these design tasks. So this design will
64 Chapter Two
consume a lot of information that was created by hard work. However,

if you adopt the right design principle, you can use off-the-shelf parts to
replace specially-designed parts, use value engineering and design for
manufacturing practice to reduce a large number of parts, and reduce
the number of immature technologies introduced into this product. If
the product still can deliver the same functions as the previous design
by doing all these, you reduce the information content in the design
greatly, so you dont need to generate all that information any more,
and you save a lot of engineering hours.
Therefore, the ideal product development process is such that it
creates information and knowledge at the highest efficiency, speed,
and quality, but the consumption of information for each good qual-
ity product is minimal. Obviously, if you generate information at the
highest possible quality, efficiency, and speed, but the consumption
of the information on each piece of good-quality product design is at
a minimum, you will have the most effective product development
process.
In recent years, the Toyota product development system (Liker 2005;
Kennedy 2003) has been getting a lot of attention. Toyota is famous for
its lean manufacturing practices. Some authors have called Toyotas
product development process the lean product development process. In
the next section, I will describe a value-based lean product development
process, which is based on the combination of best practices in the prod-
uct development process.
2.4 Customer-ValueBased Lean Product

Development Process
At the end of the last section, I derived a very important result; that is,
the ideal product development process is such that it creates informa-
tion and knowledge at the highest speed, efficiency, and quality, but the
consumption of information for each good quality product is at a mini-
mum. In this section, I will discuss how to achieve this ideal product
The lean operation technique is a very important tool in achieving the
ideal product development process. Ill start by discussing the general
lean operation in detail.
2.4.1 Lean Operation Principles

Lean manufacturing is a very effective manufacturing strategy first
developed by Toyota. During a benchmarking study for the automo-
bile industry in the late 1980s (Womack, Jones, and Roos 1990), Toyota
clearly stood above their competitors around the world with the ability
they developed to efficiently design, manufacture, market, and service
the automobiles they produced. This ability made a significant contri-

bution to both their companys profitability and growth, as consumers
found their products to simultaneously exhibit both quality and value.
The focus on recognizing and eliminating wasteful actions and utilizing
a greater proportion of their companys resources to add value for the
ultimate customer was found to be the key of their operating philosophy
by the researchers. Lean production is first mentioned in this study,
used to describe the efficient, less wasteful production system developed
by Toyota, called the Toyota Production System. Lean production, in
comparison to mass production, was shown to require one-half the time
to develop new products, one-half the engineering hours to design, one-
half the factory hours to produce, and one-half the investment in tools,
facilities, and factory space (Monden 1993; Ohno 1990; Shingo 1989).
Although the lean manufacturing approach was originally developed
in the traditional manufacturing industry, lean manufacturing mostly
deals with production systems from an operational viewpoint, not from
a hard-core technology viewpoint. It has been found that most lean
manufacturing principles can be readily adopted in other types of pro-
cesses, such as product development processes, office processes, and
service factory processes.
The key objective of the lean operation is to eliminate all process
wastes and maximize process efficiency. The key elements of lean opera-
tion include the following items:
Waste elimination in process
Value-stream mapping
One-piece ow
Pull-based production system
I will discuss each of these key elements in detail.
2.4.2 Waste Elimination in Process

In observing mass production, Tachii Ohno (Ohno 1990; Liker 2004), an
engineering genius of Toyota, and the pioneer of the Toyota Production
System, identified the following seven wastes in production systems:
1. Overproduction: Producing too much, too early

2. Waiting: Workers waiting for machines or parts
3. Unnecessary transport: Unnecessary transporting of moving parts
4. Overprocessing: Unnecessary processing steps
5. Excessive inventory: Seminished parts between operations and
excessive inventory of nished products
66 Chapter Two
6. Unnecessary movement: Unnecessary worker movements

7. Defects: Parts need rework or are scrap
These seven wastes are called muda. Muda is a Japanese term for
missed opportunities, or slack. These items are considered waste because
in the eyes of customers, these activities do not add value to the products
they desire.
In lean operation principles, the seven wastes can be identified mostly
by a value-stream mapping method. The waste caused by overproduc-
tion can be reduced or eliminated by a pull-based production system.
The waste caused by excessive inventory, waiting, unnecessary trans-
port, and unnecessary movement can be greatly reduced by one-piece
flow and work cells (cellular manufacturing).
2.4.3 Value-Stream Mapping

Value-stream mapping is a good method to chart a process and iden-
tify and quantify the waste in that process. Value-stream mapping was
developed to map and analyze production processes, especially batch
flow shops and flow shop processes. A value stream is all the activities
(both value-added and non-value-added) required to bring a product
through the main flows.
Figure 2.21 is a simplified value stream map for a production process.
Clearly, based on the definition of seven wastes, the staging, transportation,
setup, and inspection are non-value-added steps; casting, machining, and
assembly are value-added steps. In Figure 2.21, the horizontal length
of each step is in proportion to the time required to perform the step.
Transportation
Staging Setup Inspection Staging
Casting Staging Machining Assembly
Time
Value Added Non-Value

Time Added Time
Figure 2.21 A Simplied Value-Stream Map for a Production Process

The total time duration from the beginning of the process to the end
of the process is often called the process lead time. Clearly, in this example,
the value-added time is a small portion of the total lead time.
You can see that this simple value-stream map identified and quanti-
fied waste in the process and provided the clue for process improvement.
Clearly, the process can be improved if you can shorten the non-value-
added time.
This kind of simple value-stream mapping can also be used to analyze
a service process. Figure 2.22 shows a simplified value-stream map for
a sales order process.
In Figure 2.22, the dark boxes are value-added steps; the light boxes
are non-value-added steps. The first box, search, does not really add
value for customers, but it is an essential step, so for now, a color of
yellow is used.
In many production systems, there are huge amounts of muda
(seven types of waste) in the process. The ratio of value-added time
over total lead time can be used as a measure of process efficiency,
specifically:
Process Efciency = Value Added Time (2-10)

Total Lead Time
Start
2 min search 1 min 2 min wait in 3 min add

book value batch header
6 min add 60 min wait 1 min 60 min wait in

items, print quotes for cross cross check folder
check
End
3 min 60 min
confirm wait for
PC
Figure 2.22 A Simplied Value-Stream Map for a Sales Order Processing

Process
68 Chapter Two
TABLE 2.3 Process Efficiency for Various Processes
Typical Process World Class Process

Process Type Efciency Efciency
Machining 1% 20%
Fabrication 10% 25%
Assembly 15% 35%
Continuous manufacturing 30% 80%
Transactional business processes 10% 50%
Cognitive business processes 5% 25%
The major goal of a lean operation is to increase the process efficiency.

A process that has high efficiency will have much less waste, shorter
lead time, and lower cost. As a rule of thumb, a process is considered
lean if the process efficiency is more than 25 percent. Based on research
by Michael George (2003), the typical process efficiency and world class
efficiency for many types of processes are summarized in Table 2.3.
Clearly, the process efficiency of typical processes is very low. A big
proportion of process lead time is not used to do value-added work,
but to do non-value-added work, that is, muda. Lean operation tries to
redesign the process flow and layout so that the portion of process time
in doing non-value-added work is greatly reduced.
The most frequently used techniques in lean operation include the
following:
One-piece ow
Work cells (cellular manufacturing)
Pull-based production
2.4.4 One-Piece Flow

There are several types of manufacturing processes, such as the job shop
process, the batch flow shop process, and the line flow process. The job
shop process is also called machine village, which means that similar
machines are grouped together. The job flow patterns of such production
systems can be quite erratic and messy, as illustrated in Figure 2.23.
A job shop process is featured by low utilization, long delays, high work
in process inventory and long lead time. The advantage of the job shop
process is that it can take on a large variety of tasks.
Many service processes are also in the form of job shop processes. For
example, in most organizations, the departments are grouped together
by function, such as the personnel department, accounting department,
benefits department. If a new employee wants to finish all his or her
Figure 2.23 Typical Flow Pattern of a Job Shop
paperwork, he or she will go through all these departments. In many

organizations, paperwork has to be approved by many departments, so
each piece of paperwork will first go to one department, then it goes
through interdepartmental mail to the next one. If there are any mis-
takes, the paperwork could be sent back for correction. It is quite usual
for the documents to get lost or buried in the paper trail so that it takes
a long time for them to be completed.
A batch flow process is better in its flow pattern, as illustrated by
Figure 2.24.
However, there are still a lot of work-in-process inventories in a batch
flow shop. Most of the lead time is spent on inventory waiting in the
stock. It is better than the flow pattern of a job shop, but it is still inef-
ficient.
One-piece flow, or single-piece flow, is the solution proposed by the
lean operation principle. One-piece flow is illustrated by Figure 2.25.
A one-piece flow means that the work piece is worked on one piece at
a time, not one batch a time. This will eliminate the work-in-process
inventory completely.
Raw Product
Material
Batches of Semi
Finished products
Figure 2.24 Flow Patterns of a Batch Flow Shop
70 Chapter Two
Raw Product
Material
Figure 2.25 One-Piece Flow
On the other hand, in a single-piece line flow process, any error or

defect in any process step will cause the whole line to stop. In tra-
ditional Western operation management, the work-in-process inven-
tory, or buffer inventory, is used to temporarily feed the downstream
process steps so the line will not stop. However, the Toyota production
system believes that the buffer inventory has more disadvantages than
benefits; buffer inventory ties up money, and hides hidden problems. In
the Toyota production system, zero buffer inventory is used to expose
all the hidden problems in the production process. It forces you to debug
all hidden problems so that eventually, you will have a zero-defect pro-
duction process.
2.4.5 Pull-Based Production

A pull-based production system means a demand-driven production
system. The pull-based production system is modeled after supermarket
shelf replenishment operation. On a supermarket shelf, there are lots of
goods, such as milk, eggs, and orange juice, that are ready for custom-
ers to pick up. The customers pull the goods from the shelf, and then
depending on how many items are taken away, the inventory person in
the supermarket will refill the same amount of items by pulling them
from the warehouse. Then the warehouse person will order roughly the
same amount of items that are pulled from the warehouse.
Restaurant operation is a perfect example of pull-based production.
The customer places the order, and then the kitchen produces exactly
what the customer ordered. In general, the key feature for pull-based
production is that the information flow direction is opposite to the mate-
rial flow. The information flow means the production control order. In
the restaurant case, the production control is the order from kitchen to
cook; this order direction is clearly from the customer to the kitchen.
On the other hand, the direction of material flow, which is the flow of
food in the restaurant case, moves from the kitchen to the customer.
Clearly, the information flow direction and material flow direction in
the restaurant kitchen are opposite.
The opposite of pull-based production is push-based production. The
key feature for push-based production is that the direction of informa-
tion flow is the same as that of the material flow. In push-based produc-
tion, each work stop will send the work downstream in the operation.
That is, it pushes the work downstream, without considering whether
the downstream can make use of it. Typically, activities are planned
centrally but do not reflect actual conditions in terms of idle time, inven-
tory, and queues.
Agricultural production is a typical push-based production. Because
the production cycle is very long, there is no way that farmers can pro-
duce only the amount of food needed, based on real-time demand. The
production plan is purely based on market forecast, and sometimes just
based on last years production. The production command will flow in the
same direction as the workflow. It is well known that agricultural pro-
duction often suffers from oversupply and market fluctuations. Clearly,
pull-based production, whenever possible, will create much less overpro-
duction, so the waste caused by overproduction can be reduced.
2.4.6 Lean Principles for Product

Development
Lean operation practices have achieved a great deal of success in both
the manufacturing industry and many service industries, such as bank-
ing, insurance, and health care. Can these lean operation principles
achieve the same drastic results in the product development process?
The answer to this question is positive. However, there are many dis-
tinct differences between the product development process and the
manufacturing process, so the lean principles have to be modified to
work well in the product development process.
The birthplace of lean manufacturing, Toyota, definitely has an edge
in the product development process compared with North American
automobile companies. Table 2.4 summarizes the performance dif-
ferences between Toyota and North American automobile companies
around 1990 (Womack 1990).
Toyotas product development system gained a lot of attention (Liker
2005; Kennedy 2003). However, there are many other best practices as
well, such as Apple Computers innovation approach and Samsungs
Design for Six Sigma practices, and all of them have achieved great suc-
cesses. In this chapter, we will outline a lean product development strategy
TABLE 2.4 Product Development Performance Comparisons
North American Automobile

Measures Toyota Companies
Average engineering hours per new 1.7 3.1
vehicle development (million hours)
Average development time (months) 46.2 60.4
Employees per team 485 903
Ratio of delayed project 1 in 6 1 in 2
Achieve normal quality After Launch 1.4 11
72 Chapter Two
that combines many best practices and several sound design principles
(Suh 1990; Nonaka 1995; Reinertsen 1997; Huthwaite 2004).
There are many significant differences between the manufacturing
process and the product development process. For manufacturing pro-
cesses, what you are going to produce is very clear in the beginning. The
product that you produce has already been designed. For the product
development process, you are very fuzzy about what the product will
be like; and the outcome of the product development process is not a
certainty. For a manufacturing process, the rework is treated as waste;
for the product development process, iterative improvement on product
design is quite common. Even the goal of lean operation is different
between a manufacturing process and the product development process.
For a manufacturing operation, the goal of lean operation is to minimize
waste and increase speed; for the product development process, devel-
oping a top-notch product design that can lead to high sales and high
profitability is the goal, so you are trying to maximize the value of the
design, as well as reduce waste, and increase development speed.
Based on these differences, the following definition applies to this
value-based lean product development process:
The value-based lean product development process aims to deliver
more customer value by using fewer resources through
Thoroughly capturing the voice of the customer and accurately under-
standing the customer value
Effectively transforming the voice of the customer into high-quality
design with high speed and low cost
Relentlessly decreasing the wastes in the product development
process
Based on this definition, you can see the difference between lean manu-
facturing and lean product development quite easily. While lean manufac-
turing focuses on reducing waste and increasing speed, the lean product
development process focuses on both increasing the customer value and
reducing waste and increasing speed.
Unlike the seven wastes in the manufacturing process, there are
no universally agreed-upon waste categories for product development.
However, the following waste categories are commonly found in the
product development process:
Excessive design requirements Excessive tolerances, excessive

material specications, excessive operator requirements, and so on.
Excessive complexity in design The simplest design is the best
design, assuming that it can deliver all the product functions.
Poor product architecture Poor product architecture often leads

to redesign, mismatch, and performance problems.
Reinvention If someone else has already done this work, reinven-
tion certainly is a waste of manpower and resources.
Mismatch of subsystems Many design rework problems happen
in unexpected subsystem interactions.
Information loss and re-creation This happens a lot in most
companies.
Miscommunication Miscommunication among product develop-
ment team members often leads to doing the wrong work, and then
having to redo it.
Searching for information, waiting for critical information
This is certainly not a value-added activity.
Overburden on the people or resource Excessive workload and
unrealistic deadlines often lead to half-cooked projects and bug-ridden
designs, and eventually they lead to rework.
Unproductive meetings Meetings consume person-hours.
Here are some of the most important lean concepts in the product
development process:
Value The capturing of true customer value.

Value stream The ow and transformation of the customer value
and VOC information into completed product design, through the
stages of product development phases.
Flow Flow of information and knowledge.
Pull Pull the right kind of VOC and customer value information
from customers; pull the right kind of technical information and
knowledge from the right sources, at the right time, in the right
amount, by the right user.
Perfection No wastes, no information distortion and loss, perfect
information ow, no reinvention.
After these thorough discussions on the lean principles for product

development, I can define the value-based lean product development
strategy. This strategy has five main components:
1. Mining the voice of the customer to capture value

2. Maximizing technical competence
3. Front-loading the product development process
74 Chapter Two
4. Optimizing information transformation and ow

5. Creating a lean product
I will discuss these five main components in subsequent subsections.
2.4.7 Mining the Voice of the Customer

to Capture Value
The ultimate success of the product depends on the customers reception
of the product. If you had magic powers so you could read your custom-
ers minds completely and accurately, and were able to deliver what they
want with low cost, then you would be guaranteed to get rich in no time!
The same is true in discovering the real, accurate voice of the customer.
You need to find the right sources for the voice of the customer, locate
the sources, use the right method to extract good information from the
source, and process the result into useful information for the product
development team.
There are three objectives for the mining of VOC information:
1. Obtain sufcient information to develop product requirements. From
the earlier discussions on axiomatic design, you know that product
design is a sequence of mapping processes. The voice of the customer
is the source for the product functional requirements. You need to get
enough accurate information from customers to derive these product
requirements. Chapter 3 of this book provides a discussion of what
kind of voice of the customer we need to collect to serve this purpose.
Chapters 4 and 5 give comprehensive descriptions of all methods and
tools that are necessary to capture the voice of the customer. Chapter 6
discusses the technical aspects of how to rene and analyze the
voice-of-the-customer data. Chapter 7 discusses the method of qual-
ity function deployment; this method can be used to transform VOC
information into product requirements.
2. Obtain and analyze customer value information. Customer value is the
value that the customer perceives for a given product. Customer value
information is derived from the voice of the customer, because only the
customers themselves can tell what kind of product benets they really
like, and what kind of liabilities they dont like. The kind of customer
value that your product should have is a strategic decision in product
development; I call it the customer value proposition. Customer value
capturing and analysis is thoroughly discussed in Chapter 3. Chapter 8
and Chapter 9 discuss two effective methodologies to improve customer
value and deploy that value into product development.
3. Explore alternative customer value propositions. Many market-
places are ooded with competitive products, where the competition
is brutal. Many successful innovative product developments offer
a completely new kind of product that has a different customer value

proposition than its competitors. For example, in the late 1980s, when
Chrysler introduced the SUV (sport utility vehicle) to the automobile
marketplace, it was a completely new type of product with a new set
of customer benets, and thus, a new customer value proposition.
Developing a new customer value proposition also needs voice-of-
the-customer inputs. This strategy is called the blue ocean strategy,
and is discussed in Chapter 3.
A big portion of this book is about mining the voice of the customer and
delivering customer value, so that these factors can serve as a technical
manual for this value-based lean product development component.
2.4.8 Maximizing Technical Competence

Capturing the voice of the customer accurately is only a necessary condi-
tion for delivering valuable products. By capturing accurate VOC infor-
mation, you will know what customers really want, but it does not tell
you how to make this kind of product with low price, high performance,
and high quality, and or how to bring this product to market quickly.
Not only do you need the voice of the customer, but you also need strong
technical competence. In Likers book, The Toyota Product Development
System (2006), he explained that one of the 13 Toyota lean product
development principles is develop towering technical competence in
all engineers. Clearly, superior technical competence is an important
pillar for a first-class product development team.
In the value-based lean product development process, you want to
develop this capability in a cost-and time-effective manner. This strat-
egy has the following key components:
An effective information and knowledge management system

This is a good bank of information and knowledge from which people
can pull all the necessary information and explicit knowledge to create
design solutions as needed. This information bank should be a live one
that keeps improving and enriching itself.
Effective organizational knowledge creation Adopt the
knowledge-creation spirals mechanism for organizational knowledge
creation, which was originally delineated in The Knowledge-Creating
Company (Nonaka and Takeuchi 1995). We discussed this approach
in detail earlier in this chapter.
Applying TRIZ to reduce reinvention
Applying DFSS tools to support all stages of product
development Design for Six Sigma (DFSS) (Yang 2003) is an
effective technically intensive approach to supporting the product
76 Chapter Two
development process. DFSS provides a roadmap and many effective

tools to ensure product performance excellence and superior qual-
ity in every stage of the product development process. DFSS can go
hand-in-hand with a lean product development process. DFSS is
about high product performance, robustness and quality; the lean
product development process can provide help in reducing waste
and improving efficiency. DFSS and lean product development can
work together very well in the product development process.
Table 2.5 summarizes many DFSS tools that support the first four
stages of the product development process.
2.4.9 Front-Loading the Product

Development Process
As we discussed earlier, the product development process is an informa-
tion and knowledge generation process. Information has time value; you
want the key information to be available earlier, rather than later. In
Toyotas product development process, Principle 2 of the 13-principle
Toyota lean product development process calls for front-loading the
product development process to explore thoroughly alternative solutions
while there is maximum design space (Morgan and Liker 2006).
TABLE 2.5 DFSS Tools and Product Development Stages
Product/Service Life What DFSS Tools Can

Cycle Stages Accomplish DFSS Tools
Stage 0. Impetus/ Ensure that new technology Robust technology
ideation or ideas are robust for development
downstream development
Stage 1. Customer and Ensure that the new product QFD
business requirements concept can come up with the
study right functional requirements
that satisfy customer needs
Stage 2. Concept Ensure that the new concept Taguchi method/robust
development can lead to sound design and is design
free of design vulnerabilities TRIZ
Ensure that the new concept Axiomatic design
is robust for downstream DOE
development Simulation/optimization
Reliability-based design
Stage 3. Product/service Ensure that the designed Taguchi method/robust
design/prototyping product (design parameters) design
can deliver desired product DOE
functions over its useful life Simulation/optimization
Ensure that the product design Reliability-based design/
is robust for variations from testing and estimation
manufacturing, consumption,
and disposal stages
The technical approach for this front-loading principle is the set-

based design practice.
The set-based design approach is used in the concept design stage,
and it works with modular design practice. Modular design is a design
practice in which a product is broken into smaller subsystems. The sub-
systems are connected together via standard interfaces. In this case, the
subsystems become decoupled, that is, the designing of one subsystem
is not dependent upon other subsystems. Therefore, the design work for
each subsystem can be conducted in parallel.
For each subsystem, you start with the concept design. In regular
design practice, you start with a small number of design concepts, then
you select one seemingly good concept, and move into detailed design.
After the design is ready, you perform an evaluation test. If the test
shows the concept is acceptable, you move this concept to the param-
eter design and prototyping stage. If the test shows that this concept is
not acceptable, you start another raw concept and do another round of
development. You may iterate this process until an acceptable design is
found. This regular design practice is illustrated in Figure 2.26.
On the other hand, set-based design simultaneously starts with sev-
eral concepts. The initial sets of concepts are coming from
Current knowledge
New technology from recent research and development efforts
Idea generation through brainstorming or TRIZ
This initial set of concepts should include at least one concept that is
relatively mature and reliable. After the initial set is selected, you divide
the engineers into teams, such that each team works on one concept.
Each team will grow its concept by detailing, design evaluation, and
tests. It is very important that the set-based design stay in the concept
design stage; you can do some low-cost computer-aided design (CAD)
simulations, alpha prototypes, small-scale lab tests, and so on. However,
you dont launch high-cost prototype construction and validation tests
at this time. This ensures that the set-based design approach will not
be expensive and time-consuming. Figure 2.27 illustrates how set-based
design works.
Iterate (if required)
Figure 2.26 Regular Concept Design

Few Select Detail Test Process
concepts
78 Chapter Two
Figure 2.27 Set-Based Concept Design

Many concepts
Each subsystem
The whole set-based design will be subdivided into several mini-stages

during its progress; the concepts are evaluated and tested, and weaker
concepts are eliminated. The advantages of set-based design include
discovering better concepts early in the design process, and gaining
a lot of knowledge about what works well, and what doesnt. Cross-
pollination of several good concepts may create an even better one. For
example, in a university capstone design project, if you assign different
projects to different teams of students, then every team will probably
create an average, mediocre design. If you assign the same project to
many teams, and each team works on it independently, then you will get
many different designs. The best design will almost certainly be better
than the average design, and by closely examining the top two or top
three designs, you may end up with an even better concept by combining
all the advantages of these designs, and eliminating the disadvantages.
By sharing the results with all participants, everyone will gain a lot of
knowledge in this subject.
2.4.10 Optimizing Information

Transformation and Flow
Product development involves a lot of information transformation and
information flow. Information transformation usually involves pulling
design solutions to satisfy design needs. Information flow involves deliv-
ering information to needed people. The following is a list of types of
waste that often occur in the information transformation and informa-
tion flow processes:
Reinvention
Mismatch of subsystems
Information loss and re-creation
Miscommunication
Searching for information, waiting for critical information

Overburdening the people or resources
Unproductive meetings
The optimized information transformation and information flow

should be able to reduce those wastes to a minimum. There are several
very effective approaches to improving information transformation and
information flow process. However, in order to fully appreciate these
approaches, we need to review some background knowledge in human
factors and queueing theory.
2.4.10.1 Human Factors People do the product development work. In

order to find ways for people to work effectively, we need to study some
basic principles of human factors:
Attention span Attention span is the amount of time a person

can concentrate on a single activity. The ability to focus ones mental
or other efforts on an object is generally considered to be of prime
importance to the achievement of goals. When people are taking
classes or training, or attending meetings, they can only focus for
a certain period of time; after that period, their attention wanders.
Some researchers state that a persons attention span is 10 + Your
Age minutes, and that anything taught after that is not taken in, but
taking a 5- or 10-minute break after this time may help to replenish
the persons attention span.
Mental focus and multitasking When people are doing some work,
it takes some time to achieve mental focus on the job, and it takes some
time to get even a subtask done. When people are constantly inter-
rupted, the productivity will be very low. Also, when people handle
several tasks, if they switch tasks very often, the progress on each task
will usually be slower than when they are doing one thing at a time.
2.4.10.2 Queueing Theory Queueing theory is the mathematical study

of queues and waiting lines, called the queueing model. In the queueing
model, jobs or customers are entering the system for service, the system
has one or several servers, or even some complicated departments and
sequences of service operations. The customers can be served in many
ways, such as first-come first-serve, single queue, or multiple queues
(multiple waiting times). There are several queueing performance met-
rics, such as average waiting time in the queue, queue length, and uti-
lization of servers.
Queueing theory has many established results about how to improve
queueing systems, such as how to improve the arriving process, and
queueing the service process to improve job service times and reduce
80 Chapter Two
waiting times. These results can be readily used in helping the product
development process as well. The product development process involves
many projects. Each team or engineer will work on several projects
during the whole product development process, so some projects have
to wait in the queue until the current project is finished so the team or
engineer can be free to work on them. In this case, how you sequence the
jobs and how you assign the workload and timing will make a lot of dif-
ference in overall product development progress. Queueing theory can
provide great help in reducing waiting time and improving throughput
(number of projects finished/unit time).
There are several important results in queueing theory that are rel-
evant to the product development process.
Batch queueing is inefficient Batch queueing means the jobs are

coming into the queue in big groups, or batches. For example, if a plane-
load of international travellers arrives in an airport custom inspection
and immigration inspection stop, a long queue will form immediately,
and it takes a while for everyone to get cleared. If most people get out of
work at about the same time, a trafc jam will form immediately and it
takes a long time for everyone to get home. If the same number of people
get out of work at different times, the trafc time for each person will be
much shorter. The implication for product development practice is that
if you load product development teams or engineers with big chunks of
work, instead of staggering the work, your throughput will be low.
Nonlinear relationship between capacity utilization and queue
length Queueing theory states that the relationship between server
utilization and average waiting time is a nonlinear relationship, as
illustated in Figure 2.27. Capacity utilization is dened as the per-
centage of time that the server is busy. What this relationship indi-
cates is that when the server is partially loaded, say 50 percent, the
waiting time will be very low. However, if you increase the capacity
utilization by just 25 percent, queue length will grow to several times
longer. When the server is 100 percent loaded, the waiting time will
be extremely high. You can see this fact in your real-life experience.
On a three-lane freeway that is not fully loaded, if a trafc accident
suddenly occurs and just one lane is blocked, a long queue will form
immediately, even though theoretically there is enough room to let
every car go through. The implication for the product development
process is that overburdening the product team or engineers will
make the product development lead time much longer.
Constant job arrival rate vs. variable arrival rate Assume
that you have two scenarios in the waiting queue system. The rst
one is such that every job arrives at exactly the same time interval,
say exactly every 10 minutes. The second one is such that every job
Average Time in Queue The Product Development Process 81
Figure 2.28 Queue Length vs.

0 10 20 30 40 50 60 70 80 90 100 Capacity Utilization
Capacity Utilization
arrives at variable time intervals, say one job arrives 2 minutes after
the previous job, the next job 18 minutes later, and so on, but the aver-
age inter-arrival interval (for example, also 10 minutes) is the same
as the rst case. The average waiting time and queue length for the
rst queue system will be shorter than the second queue system. The
implication for the product development process is that if you load
jobs to engineers evenly, the throughput will be higher.
Uneven job size vs. similar job size Again, we are comparing two
queue systems. In both queues, the arriving time patterns are the same.
The rst queue system is such that every incoming job takes about the
same amount of time for the server to nish, for example, 10 minutes.
The second queue system is such that every incoming job is a different
size. Some jobs are big, and they take a longer time to nish. Some jobs
are small, and they take a shorter time to nish, even if the average
job processing time is the same as the rst queue system (for example,
also 10 minutes). Then the waiting time and queue length of the rst
queue system will be shorter than the second one. The implication for
product development is that loading engineers with similar job sizes
for each task will increase the job throughput.
After reviewing this background about human factor and queueing

theory, I can now present several effective methods to greatly improve
information transformation and flow.
2.4.10.3 Visible Knowledge Visible knowledge is a very important

component of Toyotas product development system. Visible knowledge
is knowledge that has been captured and illustrated so that it is easier to
share with other people within the organization. Here we will show two
important visible knowledge tools, the A3 report and the planning wall.
A3 Report An A3 report is a report that uses A3 size paper (11" 17").

Each report will use exactly one piece of A3 size paper, which is equivalent
82 Chapter Two
to a two-page report for regular 8" 11" size paper. Based on the objec-
tives, there are primarily three kinds of A3 report:
Knowledge sharing
Problem solving
Project status report
The contents in each A3 report for a knowledge-sharing objective

usually include the following contents:
Problem statement
Current situation
Prior research or work
Root cause analysis
Experimental methods
Data analysis
Recommendations
For other objectives, the contents will be different. An A3 report is

primarily used in
Communication before meetings

Communication on the planning wall
In Toyota, before a meeting takes place, even a one-to-one meeting,

the participants usually e-mail an A3 report to each other. Before the
meeting starts, the participants can get enough information on the sub-
ject from each other so that the serious discussion can take place more
quickly. In this way, meetings can be very productive. In many meetings,
a lot of time is wasted in trying to understand what other people are
really up to; it may take a long time just to get problems defined. After
an hour or so, people are tired and stop paying attention, so the meeting
is not productive.
The advantage of the A3 report is that it offers just about the right
amount of information for people to digest. A three-line e-mail is too
short and not enough information is provided. A 20-page report is too
long and intimidating. An A3 report makes communication effective and
the meeting productive.
Planning Wall Planning walls are the walls in the project control
room for many Japanese companies. Various A3 reports are displayed
on its walls, giving team members an immediate bigger picture view of
the project objectives and how they relate to overall corporate objectives,
Activities Checkpoint
A Meeting
Proceed to
B
next phase
C
E
D Redirect
Project
Figure 2.29 Stage Gate
Developmet Phase
visually oriented progress reports for all parts of the project (color-coded,
to show which metrics are on target and which need immediate atten-
tion), and much more. Centralizing and distilling all of this project data in
one location creates, in effect, a set of project management dashboards
that team members can learn from, discuss, and collaborate around.
Clearly, the visible knowledge tools such as A3 reports and planning
walls help to make information transformation and flow visible to all
team members, so people will know what has been done, and what has
not. They can also learn from other peoples experience. This helps to
reduce the waste caused by information searching, miscommunication,
knowledge loss and re-creaction, and so on.
2.4.10.4 Stage Overlapping The stage-gate approach is widely applied

in many companies. The stage-gate approach brings a lot of discipline
to the product development process and it plays a role in reducing
quality problems during this process. However, if the stage gate is
implemented in a very strict manner, no work in the next stage can
start unless every check in the current stage is passed, as illustrated
in Figure 2.29.
However, this approach is simply a batch queue, which is illustrated
in Figure 2.30. In Figure 2.30, the projects are like customers in the
queue. The strict stage-gate practice means that nobody can cross the
Phase 2 Phase Gate Phase 1
Figure 2.30 Stage Gate as a Batch Queue

84 Chapter Two
phase gate unless all the customers are served by the server in phase 1.
The finished customers have to stay in phase 1 until the last customer
is served, then all the customers flood into the next phase.
Obviously, this is a slow and ineffective approach. Clark and Fujimori
(1991) discussed the stage-overlapping approach to overcome this defi-
ciency. The idea of stage overlapping is illustrated in Figure 2.31.
In Figure 2.31, the product design and process design are two phases
in the product development process. Stage overlapping means that the
second phase does not need to wait until the first phase is completely
finished. The second phase can start as soon as a necessary portion of
the first phase is finished. In this way, the total duration needed to finish
both phases will be much shorter.
2.4.10.5 Information Flow across Stages Clark and Fujimori (1991)

also discussed the importance of letting information flow across differ-
ent stages of the product development process, and among engineers.
Clark and Fujimori defined information flow quality, as illustrated in
Figure 2.32. The first row in Figure 2.32 indicates that a strictly phased
and batched information flow has poor information flow quality; a frag-
mented information flow under overlapped stages has higher informa-
tion flow quality. This is common sense, just as we eat bite by bite,
rather than eating a big chunk at a time.
The second row of Figure 2.32 shows that the throwing over the wall
type of batch document communication has poor information flow qual-
ity; a face-to-face, person-to-person communication has high information
flow quality. The rest of Figure 2.32 is self-explanatory.
2.4.10.6 Reduction of Product Development System Complexity In

Toyotas product development process, there are 13 lean product devel-
opment principles. In this chapter, we have already discussed a few of
these principles in earlier sections. If you can understand the importance
of information flow and some basics of human factors and queueing the-
ory, these principles are very easy to understand. Here we will discuss
several more principles of Toyotas lean product development principles
(Liker 2006).
Product Design
Process Design
Time
Figure 2.31 Stage Overlapping

Sequential Stage Overlapping

Timing of Upstream-Downstream Activities
(phased) (simultaneous)
Documents
Face-to-Face
Computer Network Richness of Information Media
(high bandwidth)
(low bandwidth)
Batch
Fragmented
Transmission Frequency of Information Transaction
(piece by piece)
(one shot)
Bilateral
Unilateral Direction of Communication
(feedback)
Late release of Early Release

Complete Timing of Upstream-Downstream Information Flow of Preliminary
Information Information
Low High
Figure 2.32 The Quality of Information Flow
Principle 3: Create Leveled Product Development Process Flow This

principle calls for synchronizing activities across different functional
departments in a product development organization. It also calls for
an evenly distributed workload to various departments and engineers;
rather than having extremely busy days mixed with extremely idle
days, you will want steady loads for all the people. This approach cre-
ates a steady workload and job flow so the tasks will flow through the
organization smoothly and waiting line will be unlikely to occur. Even
if you get a huge data set to be distributed to other department, you
will cut the big data set into chunks and give the people one chunk at
a time.
Overall, this principle stresses the importance of even job flow, avoid-
ance of batch queues, and overloads the capacity of the queueing system.
This principle makes perfect sense in queueing theory.
Principle 4: Utilizing Rigorous Standardization to Reduce Variation

and Create Flexibility and Predictable Outcomes This principle calls
for applying the following four kinds of standardization across the
product development organization:
1. Design standardization: Engineering checklist, standard architecture,

sharing of common components
86 Chapter Two
2. Process standardization: Standardizing common tasks, sequence of

tasks, and task duration
3. Skill set standardization
4. Standardized skill inventories
This principle uses the fact that standardization will reduce complexi-
ties and confusion in communications among engineers. Standardization
will make each job more transparent and uniform, so you can have more
predictable outcomes. Standardization will also reduce the waste caused
by reinvention, mismatch, information loss, and re-creation.
Principle 11: Adapt Technology to Fit People and Process This prin-
ciple calls for achieving a balance among technology, people, and the
current product development process. Newer and fancier technologies
may bring more capabilities, but they may also create mismatch, con-
fusion, and communication problems on current product development
systems and product development teams. You should try to integrate
new technology seamlessly into existing technologies and existing prod-
uct development systems before using it. The goal of introducing new
technology is to improve and support the lean product development
process, not to damage it.
Overall, this Toyota principle calls for streamlining and simplifying
the overall system of people, technology, and process. Certainly it makes
sense according to the information axiom of the axiomatic design; if the
system can do the job, then the simpler system is a better system.
Principle 12: Align Your Organization Through Simple, Visual

Communications Principle 12 is the principle of visible knowledge.
Principle 13: Use Powerful Tools for Standardization and

Organizational Learning Principle 13 addresses the issue of organi-
zational learning and knowledge creation.
2.4.11 Creating a Lean Product

Finally, recall the definition of the ideal product development process:
The ideal product development process is such that it creates informa-
tion and knowledge at the highest efficiency, speed, and quality, but the
consumption of information for each good quality product is minimal.
The previous strategies all deal with creating information and knowl-
edge at the highest efficiency, speed, and quality. We still need to develop
a strategy to reduce information consumption for each good-quality
product design. The information consumption is proportional to the
complexity of the product design. So we need to trim all unnecessary
complexities out of product design. A product design that is free of

unnecessary complexities is called a lean product.
The complexity in engineering design is related to
Number of functions and parts

Complexity in product architecture (how different modules and design
parameters are related to each other)
Uncertainty (such as uncertainty caused by variation in quality and
technical immaturity)
Complex relationships between design parameters and product per-
formance
Based on the work of Huthwaite (2004), the following approaches can

be used to reduce the unnecessary complexities in the product design
and create lean products:
Reducing unnecessary product functions and parts

Loosening up unreasonable tolerances
Using standard or off-the-shelf parts
Controlling technical immaturity
Avoiding complicated user/operator requirements
Avoiding complicated interface requirements
Chapter
Customer Value and the

3
Voice of the Customer
The bottom line for every company is not its short-term profitability but
the value of its products in the eyes of potential customers. Short-term
profitability reflects a companys recent history and past strengths, but
without continuing enthusiasm from customers, profitability may not
last. It is customers opinions that will determine the price level, the size
of the market, and the future trends for a product family.
Customers opinions of the value of a product comprise the customer
value. A product that has high customer value often also has increas-
ing market share, increasing customer enthusiasm toward the product,
word-of-mouth praise, increasing name recognition, a reasonable price,
and a healthy profit margin for the company that produces it. Clearly,
the ability to design and deliver service products that have high cus-
tomer value is the key to success for service organizations.
There are plenty of books that discuss issues related to customer
value. The famous book Market Ownership by William Sherden (1994)
has an excellent chapter on customer value. Bradley Gales book
Managing Customer Value (1994) also presents workable methods for
surveying customer value and using that information in product and
service design.
The customer value for a given product or service may change over
time, and a new product or business model that better fits the changing
customer value could be a breakthrough product. Creating products or
business models that fit customer value shifts is called a blue ocean
strategy (Kim and Mauborgne 2005).
89
90 Chapter Three
Overall, in this book, we will develop a comprehensive strategy that

integrates several wonderful methods to create superior customer value
for service products:
Well-designed surveys are essential for obtaining customers opinions,
and there is some excellent literature on this topic. Customer survey
design and analysis will be discussed in Chapter 4.
The ethnographic method is a new and popular approach for collecting
the voice of the customer. This method is rooted in anthropology and
it is based on detailed observation on the customers turf. This method
can capture a lot of unarticulated and unspoken customer needs. The
ethnographic method is discussed in Chapter 5.
After collecting the voice of customer, the raw VOC data needs to be
analyzed and processed into well-defined, quantitative quality met-
rics. VOC data analysis is discussed in Chapter 6.
Quality function deployment, or QFD (Cohen 1985, 1989), is an excel-
lent method developed in Japan, and it can be used to deploy custom-
ers wants into product designs. QFD will be discussed in Chapter 7.
Customer value is also highly related to brand recognition. Usually,
customers are willing to pay more to buy a product with a well-established
brand name than a similar product with no name recognition, so brand
building should be an important strategic consideration in product
design and customer value enhancement. Brand development will be
discussed in Chapter 8.
Value engineering is a technique that can systematically guide
design engineers to develop high-value products with low cost. Value-
engineering books (Park 1999) provide detailed definitions and methods
for value analysis and cost reduction. The value-engineering technique
will be thoroughly discussed in Chapter 9.
Innovation and uniqueness are also huge factors for customer value.
If the new product concept is right, customers may be willing to pay
a premium price for a unique or first of its kind product. We will
discuss the theory of inventive problem solving in Chapter 10.
This chapter will outline the customer value creation strategy and
provide an overview for all these methods. Section 3.1 will formally
define customer value and its components. Section 3.2 will discuss a
practical analysis framework for collecting and analyzing customer
value data, and Section 3.3 will discuss how to use the results in prod-
uct development. Section 3.4 will discuss how customer value changes
over time and how we can develop breakthrough products by capturing
the changes in customer value. Section 3.5 discusses the relationship
between customer value and the voice of the customerspecifically,
Customer Value and the Voice of the Customer 91
what kind of VOC data we should collect in order to capture all aspects
of customer value.
3.1 Customer Value and Its Elements

Value is a frequently used term, yet the concept is a confusing one.
The nature of value has been extensively studied by many researchers,
including value-engineering or value-analysis professionals. According
to Park (2001), one of the leaders in the field of value engineering, cost
is a fact; it is a measure of the amount of money, time, labor, and any
other expenses necessary to obtain a requirement. Value, on the other
hand, is a matter of opinion of the buyer or customer as to what the
product is worth, based on what it does to him/her. In addition, a per-
sons measure of value is constantly changing to meet a specific situa-
tion. Dictionaries define value in terms of an items equivalent in goods,
services, or money; an items market price; or its worth, usefulness, or
importance to someone.
Value is related to worth, with worth being either a synonym for value
or the quality that gives an item value. Carlos Fallon (1980) states that
worth is a simple concept; it becomes value when it is related to cost.
He further states that cost is a necessary component of value. Chris
OBrien (1982) further defines value as the ratio of worth and cost. In
this definition, worth is an appraisal of the properties of a productit
is essentially an appraisal of the function of the product. In other words,
value is the ratio of function to cost, with the function being what the
product does for the customer.
Many value-engineering researchers and practitioners have devel-
oped precise definitions for value, such as the following:
Wants + needs SellFunctions + UseFunctions

Bryant Value V = =
Resources Dollar + People
Worth
Harris V=
Effort
Functions
Kaufman V=
Cost
Function Utility Performance

Wasserman V= = =
Cost Cost Cost
Objectives
Fallon V=
Cost
In each of these definitions, the denominator is a unit that can be
measured by dollars, effort, resources, manpower, and so on. In one way
92 Chapter Three
or another, they can all be converted to dollars. We can say that all
these definitions converge to having the denominator be a measure of
cost. The numerators converge to being a measure of function, or perfor-
mance. Therefore, in value engineering, value is measured primarily as
a function-to-cost ratio. A product with better functionality and lower
cost has a higher value. If a product or method can accomplish a given
function with a lower cost than all competitors, this product or method
is the best value.
Clearly, a higher function-to-cost ratio is important for increasing
value in the eyes of the customer. However, in my opinion, the function-
to-cost ratio alone is not sufficient to adequately measure customer
value. Consider these two points:
In value engineering the best value of a function is sometimes also

defined as the lowest cost to accomplish that function, measured in
dollars. But this definition is difficult to reconcile with value is func-
tion divided by cost. (Does this imply that function is measured in
units of dollars squared?)
There are many cases where people are willing to pay different prices
for two products that have exactly the same functions. For example,
the Toyota Corolla is exactly the same car as the Geo Prizm, but people
are willing to pay $300 more to buy the Toyota Corolla. Similarly, the
same item in a neighborhood convenience store will usually sell at
a significantly higher price than in a large discount chain store. The
function-to-cost ratio cannot explain the items value in the eyes of
customers adequately.
Sherden (1994) and Gale (1994) provided a broader definition for

customer value. In their definition, customer value is perceived benefit
(benefits) minus perceived cost (liabilities):
Value = Benefits Liabilities
The benefits include the following categories:
Functional benefits
Product functions, functional performance levels
Economic benefits, revenues (for investment services)
Reliability and durability
Psychological benefits
Prestige and emotional factors, such as brand name reputation.
Perceived dependability (for example, people prefer a known brand
rather than an unknown brand).
Social and ethical reasons (for example, environmentally friendly

brands)
Psychological awe (many first-in-market products not only provide
unique functions, but also give customers a tremendous thrill; for
example, the first copy machine really impressed customers)
Psychological effects of competition. For example, if there are many
competitors producing a same or similar product, that usually will
create a perception that this product is a commodity and it doesnt
carry much value.
Service and convenience benefits
Availability (how easy is it to access the product or service?)
Service (how easy is it to get service in case of product problems or
failure?)
The liabilities include the following:
Economic liabilities
Price
Acquisition cost (such as transportation and shipping costs, time
and effort spent to obtain the service)
Usage cost (additional cost to use the product or service in addition
to the purchasing price, such as installation)
Maintenance costs
Ownership costs
Disposal costs
Psychological liabilities
Uncertainty about the dependability of the product or service
Self-esteem liability of using an unknown brand product
Psychological liability of poor performance of the product or service
Service and convenience liability
Liability due to lack of service
Liability due to poor service
Liability due to poor availability (such as delivery time, distance
to shop)
Clearly, this customer value definition by Sherden and Gale contains

much more information than simply the products function and cost,
and it also is measured in dollars, or monetary worth. I will use this
customer value definition throughout this book.
94 Chapter Three
3.1.1 Value and Other Commonly Used Metrics

There are other product metrics that can easily be combined with the
concept of value, including price, performance, cost, and quality. Lets look
at the similarities and differences between value and these metrics.
3.1.1.1 Value and Price Some economists define value as price, but as
we saw earlier, price is only one factor that affects the value. Specifically,
price is one important element in customer value (economic liabilities).
In general, customers accept a high selling price for a product if they
believe the product offers superior benefits.
A higher price may provide a higher profit margin for the company
that sells the product, but the sales volume can be sustained only if
customers think that this product will provide more customer benefits
than costs; that is, if it offers high customer value (more benefits for
a lower price). The gap between customer-perceived benefits versus
liabilities determines the products overall attractiveness to customers,
and thus determines the size of the market, the sales volume, and the
market share.
3.1.1.2 Value and Performance Performance is also called function.

The function of a product is what the product is supposed to do for
customers, and it is only one component of the products benefits;
customer value is the benefits minus all costs. Therefore, more and
better functions may not always give a better value. Also, a function
can create value only if that function is something customers need or
want. In 1979 the American Can Company thought that a product with
more and better functions would always sell, so it designed a stronger
paper towel called BOLT. It looked and performed like cloth, and it was
sold at a higher price than regular paper towels. However, this product
was a total failure, because customers did not perceive a benefit in a
paper towel that could be washed. Actually, in many cases, customers
enjoy a product change that adds more functions without increasing
price, and often they are delighted to see a product change that reduces
both functions and price. Adding and improving functions while also
increasing the price is a risky strategy.
3.1.1.3 Value and Cost of Production The costs of producing a product

usually do not relate to customer value. A company could provide a
product with a lot of features and high cost that customers do not ap-
preciate. However, reducing cost will either increase the profit margin
or create more room for price reduction.
3.1.1.4 Value, Quality, and Perceived Quality Like value, quality is a

tricky concept to define. Even quality gurus do not agree on a definition.
The American Society for Quality (ASQ) defines quality as a subjective
term for which each person has his or her own definition. In technical
usage, quality can have two meanings: 1. the characteristics of a prod-
uct or service that bear on its ability to satisfy stated or implied needs.
2. a product or service free of deficiencies (www.asq.org).
In ASQs definition, the characteristics of a product or service that
bear on its ability to satisfy stated or implied needs sounds a lot like
customer-preferred performance and function; and a product or ser-
vice free of deficiencies is definitely related to dependability and reli-
ability. In comparing quality with customer value, it is clear that value
is a much broader concept. Quality is a part of value, not all of value.
Quality is subjective, as mentioned in the ASQs definition, and it has
a psychological component. For example, a brand-name drug may have
exactly the same functionality and manufacturing quality as a generic
drug. However, a substantial portion of consumers will still consider
the drug with the brand name to have higher quality. This overall
customer opinion of the quality of a particular product or service is
also called the perceived quality, and it is a better indicator of customer
value than any objective measures of quality. There are primarily two
components in perceived quality: the technical component that relates
to performance, functionality, dependability, and level of defects; and
the psychological component, such as brand image.
3.1.2 The Versatility and Dynamics of Value

Customer value can mean different things for different peoplewe call
this the versatility of value. Value may also change over time with peo-
ples preferences and lifestyleswe call this the dynamics of value.
The versatility of value reflects the fact that the marketplace consists
of different people, and it is difficult to find even two people who have
exactly the same opinion. In marketing science, people can be divided
into market segments for a particular kind of product or service; the
customers within each segment display similar behavior and opinions.
Some products or services can only find customers in a particular market
segment; this is a niche market. For example, a good state-of-the-art
computer engineering book can only be sold at university campuses and
to computer engineersthis book is useless to an animal trainer. Some
products and services, however, can find customers in mass market; for
example, vegetables, fruits, pencils, and personal banking services have
customers all over the social spectrum.
Given the versatility of value, it is important to know the types of
markets that your product or service is in, and to understand what value
your product or service can offer to customers. Even for a single product
or service, its value can be broken down into several categories:
Use value Properties that make something work; this can also be
called functional value.
96 Chapter Three
Esteem value Properties that make something desirable to own;

this can also be called emotional or psychological value.
Exchange value Properties that make it possible to exchange one
thing for another.
For example, the use value of an airline ticket is the ability of a

customer to take an airplane from point A to point B; even a coach-
class ticket provides that use value. A business-class ticket provides
a little more functional value, such as better seating and better food,
but it also provides substantial esteem value the feeling that I am
special, I have special status. Coach-class tickets are in a mass market;
the business-class ticket is in a niche market.
If you want to enhance the value of a product or service so that it will
be more successful in the marketplace, you need to determine what will
make it more valuable. Because value is a matter of customer opinion,
you need to understand what will motivate and excite people.
Abraham Maslow developed a simple scale to define peoples psycho-
logical needs. He called this scale the hierarchy of needs, as shown
in Figure 3.1. Maslow said that people are motivated to do different
things at different levels of psychological development, or at different
levels of society. He divided these motivational factors into five basic
needs. As each need is satisfied, other higher needs arise. Although the
lower level needs may never disappear, they become weaker or less
important. A person may have several needs at the same time but one
need is dominant.
Self-
Esteem Actualization
Belonging
Safety
Relative Importance
Survival
Psychological Development
Figure 3.1 Value and Psychological Development
Maslows theory provides a lot of insight into customers buying motiva-

tion. For example, customers in developing countries usually prefer prod-
ucts that address basic needs, which are robust in harsh user conditions,
and which are low in cost and without fancy features. This reflects the fact
that survival and safety needs are predominant needs for these consumers.
However, for many children in affluent countries, printing sports stars fig-
ures on cereal boxes might be a very effective way to increase sales because
it addresses the esteem and belonging needs for such children.
Many factors can also change customers preferences and the tradeoffs
theyre willing to make regarding products. This is the dynamic nature
of customer value (changing over time). It can be affected by economic
conditions, new technologies, and the changes in customers psychological
development.
For example, changes in oil prices will affect peoples attitudes toward
the types of cars they like. Similarly, bad economic times, fierce competi-
tion, and a tough job market can make discount chain stores a favorite
shopping place. A strong competitors emergence in a market segment
will greatly change the expectations for a particular product or service.
3.2 Customer Value Analysis

In the last section, we established that customer value is the difference
between benefits and liabilities:
Value = Benefits Liabilities

where the benefits consist of the following elements:
Functional benefits
Psychological benefits
Service and convenience benefits
and the liabilities consist of these elements:
Economical liabilities (customer costs)

Psychological liabilities
Service and convenience liabilities
For any product or service, the company that provides more value
to its customers than its competitors will eventually gain in sales and
profitability. However, for each particular product or service, the profile
of benefits and liabilities will be very different, and customers will give
the benefits and liabilities different relative importance. When a prod-
uct or service has several competitors, it is very important to do better
in the areas most important to customers.
98 Chapter Three
Bradley Gale (1994) developed a systematic approach to maximiz-

ing customer value in providing products and services. This approach
consists of the following steps:
1. Conduct customer surveys to get information about the relative impor-

tance to customers of aspects of a particular product or service, and to
get customers ratings of competing products or services. Gale referred
to this step as compiling the market-perceived quality profile.
2. Collect information about the price or combined customer costs of
the companys product or service, and the same information for com-
peting products or services. Gale referred to this as compiling the
market-perceived price profile.
3. Complete a comprehensive customer value evaluation of the com-
panys product or service compared with its competitors. Gale called
this compiling the customer value map.
4. Complete an area-to-area competitive analysis, to identify the critical
areas in order to gain a competitive advantage.
5. Deploy the critical improvement into the product or service design.
In a customer value management approach, all non-cost-related

attributes, such as functional benefits, psychological benefits, and
service and convenience benefits, are considered to be components of
market-perceived quality. If a product or service can offer higher market-
perceived quality than all competitors, yet have its cost under control,
this product will have a competitive edge.
3.2.1 Market-Perceived Quality Profile

A market-perceived quality profile is a detailed quality scorecard
that provides quantitative quality ratings for a companys own product
versus its competitors products on all important non-cost-related attri-
butes. According to Gale (1994), creating the market-perceived quality
profile involves the following steps:
1. In forums such as focus groups, ask customers in the targeted

market, including both your customers and competitors custom-
ers, to list the factors other than price that are important in their
purchase decisions.
2. Establish how the various non-price factors are weighted in the cus-
tomers decision, usually by simply asking customers to tell you how
they weigh the various factors, distributing 100 points among them.
3. Ask well-informed customers, including both yours and your com-
petitors, how you and your competitors perform on the various
quality attributes. Then for each attribute, divide the score of the
product or service you are studying by the scores of competitors
products. That gives you the performance ratio on that attribute.
Multiply each ratio by the weight of that attribute, and add the
results to get an overall market-perceived quality score.
Table 3.1 shows an example of a market-perceived quality profile
for frozen chicken (Gale 1994). The first column lists all of the impor-
tant non-cost-related quality attributes in the chicken business: yellow
bird, meat-to-bone, and so on. The second column lists the relative
importance rating for the attributes listed in column 1; the relative
importance ratings will add up to 100%. In this example, yellow bird
accounts for 10% of relative importance, and fresh accounts for 15%
of relative importance. These importance ratings are obtained from a
specially-designed customer survey. The third column records the aver-
age customer rating of the product from our business on each quality
attribute; the highest possible score is 10.0, and the lowest possible score
is 1.0. These scores are also computed based on a specially-designed
customer survey. The fourth column lists the average customer ratings
of the competing products for each customer attribute.
The remaining columns are calculated from the preceding ones. The
fifth column lists the ratio of column 3 to column 4; that is, the average
customer rating of our product versus the average customer rating of the
competitors products. Clearly, if this ratio is less than 1, it means that in
this quality attribute category our product performs worse than the com-
petitors average; if this ratio is greater than 1, it means that in this cat-
egory our product performs better than the average competitors. The last
column lists the ratio from column 5 multiplied by the relative importance
score in column 2. Clearly, if all the ratios are equal to 1, it means that our
product is an average product in comparison with our competitors. In that
case, the total score in column 6 would be equal to 100, and our market-
perceived quality score would be 100. A product with a market-perceived
quality score larger than 100 is considered to be a competitive product,
and the higher the score, the more competitive the product is.
The information used to compile the market-perceived quality profile
can be obtained by conducting a special kind of customer survey. A form
for this kind of customer survey is illustrated in Table 3.2.
The survey population should include the customers of our company,
as well as all consumers who purchase this kind of product or service,
including the customers of competitors. For example, if our company is
McDonalds, then the survey population should be the customers for all
fast food chains and should include the customer population of Burger
King, Wendys, and so on. Clearly, if we collected enough finished survey
forms from customers, we would have all the information needed in a
market-perceived quality profile study.
100
TABLE 3.1 Market-Perceived Quality Profile in the Chicken Business

Industry Comparison
Quality Scores
Customers Perdue
Weight of (Our business) Ratio (Ours/Competitors)
Attributes (1 = lowest, Average (Ratio>1.0 means better
Quality Attributes (Total = 100) 10 = highest) Competitors than competitor) Customer Weight x Ratio
Yellow Bird 10 8.1 7.2 1.13 = 8.1/7.2 11.3 = 1.13 10
Meat-to-bone 20 9.0 7.3 1.23 24.6 = 1.23 20
No Pinfeathers 20 9.2 6.5 1.42 28.4
Fresh 15 8.0 8.0 1.00 15.0
Availability 10 8.0 8.0 1.00 10.0
Brand Image 25 9.4 6.4 1.47 36.8
Total 100 126.1 = market-
perceived quality score
TABLE 3.2 Survey Form for Customer Values
Performance Scores
1 10, 1 = lowest, 10 = highest
Quality (Non-Price) Importance Weights Company A Company B Company C Company D
Attributes (Add up to 100) (Our company) (Competitor 1) (Competitor 2) (Competitor 3)
1
2
3
4
5
6
7
8
9
10
Sum of Importance
Weights = 100
Price (perceived transaction price)
Price Competitiveness
101
102 Chapter Three
Besides customer survey data, functional data can also be used

in a market-perceived quality profile study. Table 3.3 gives such an
example.
3.2.2 Market-Perceived Price Profile

For some industries, such as retail sales, the price for a particular item
is paid only once, so it is very clearly understood. In this case, the price
comparison simply involves comparing one dollar amount to another
dollar amount. For many other businesses, the overall customer cost
structure is rather complicated. For example, the cost-related factors
in purchasing a car might involve a trade-in allowance, rebate, and
finance rate, in addition to the purchase price of the car. In such a case,
it is necessary to create a market-perceived price profile, because it will
integrate all the cost factors and compile a combined price score.
The construction of the market-perceived price profile is very similar to
creating a market-perceived quality profile. Customers are asked to list
the factors that affect their perceptions of a products cost. Table 3.4 shows
an example of a market-perceived price profile in the luxury car market.
Because people consider lower prices to be better, however, having a
higher customer satisfaction score in price level represent a better price
is counterintuitive. In Table 3.4, Acuras market-perceived price score
is 118.7this is more than 100, which means that the Acuras overall
price level is more attractive (lower) than other competitors. Using the
inverse score (84.2) is more intuitive.
TABLE 3.3 Quality Profile Studies of Gallbladder Operations
Endosurgery versus Traditional Surgery

Customers
Quality Scores
Weight of Industry Comparison
Quality Attributes Customer
Attributes (Total = 100) Endo Traditional Ratio Weight x Ratio
At Home 40 12 weeks 68 weeks 3.0 120
Recovery Period
Hospital Stay 30 12 days 37 days 2.0 60
Complication 10 05% 110% 1.5 15
Rate
Postoperative 5 0.51 inch 35 inch 1.4 7
Scar
Operation time 15 0.51 hour 12 hours 2.0 30
Total 100 232 = market-
perceived
quality score
TABLE 3.4 Market-Perceived Price Profile: Luxury Cars
Quality Scores
Industry Comparison
Ratio (Ours/
Customers Acura Competitor)
Weight of (Our business) (Ratio>1.0 means
Price Satisfaction Attributes (1 = lowest, Average better than
Attributes (Total = 100) 10 = highest) Competitors competitor) Customer Weight x Ratio
Purchase price 60 9 7 1.29 = 9/7 77.4 = 1.29 60
Trade-in allowance 20 6 6 1.0 20 = 1.0 20
Resale price 10 9 8 1.13 11.3
Finance rates 10 7 7 1.00 10.0
Total 100 118.7 = market-perceived
price score
Relative price ratio 84.2 = (1/118.7) 100
103
104 Chapter Three
If we just want to compare the sticker prices, a simple price ratio

can be used. For example, if Acuras price is $35, 200, and the average
competitors price is $40,000, the price ratio is $35,200/$40,000 = 0.88. If
a percentage score is used, the relative price ratio is 0.88 100 = 88%. If
the relative price ratio of a product is less than 100, then its price level
is lower than its competitors.
3.2.3 Customer Value Map

A customer value map is a very useful tool for identifying the competi-
tive position of a particular product in comparison with other competi-
tors products. A product is competitive if it has high customer benefit
and low customer cost. The customer benefit can be well represented
by the market-perceived quality score, and the customer cost position
can be represented by the relative price ratio.
A customer value map is a two-dimensional plot of the market-
perceived quality score on the horizontal axis versus the relative price
ratio on the vertical axis. Figure 3.2 shows what a customer value map
looks like.
In this customer value map, each dot represents a particular product.
The diagonal line in the customer value map represents the range where
the market-perceived quality score is equal to the relative price ratio.
Low value, high cost

Higher
price Worse Customer
Value
Fair value
zone
Relative price ratio
100
High value
low cost
Better Customer
Value
Lower
price Lower score 100 Higher score
Market-perceived quality score
Figure 3.2 Customer Value Map
For example, if a product has a market-perceived quality score of 80,

and its relative price ratio is also 80, the dot representing this product
will be on the diagonal line; this product would have low value and low
price. Similarly, if a product has a market-perceived quality score of 120,
and relative price ratio of 120, it would also be on the diagonal line; it
would be a high-price, high-value product.
Overall, the region around the diagonal line can be called the fair-
value zone, and products in this zone can be considered average prod-
ucts. The products in the lower-right corner of the customer value map
feature lower relative price ratios and higher market-perceived quality
scores. These are high-value, low-price productsthey have superior
competitive positions and are poised to gain market share. The prod-
ucts at the upper-left corner of the customer value map feature high
relative price ratio and low market-perceived quality scores. These are
low-value, high-price products that have inferior competitive positions
in the marketplace and are vulnerable to losing market share.
Example 3.1 Toaster Customer Value Analysis Table 3.5 lists 15 brands
of toasters and their market-perceived quality scores and relative price
ratios.
TABLE 3.5 Market-Perceived Quality Profile and Price Profile for Toasters
Market-Perceived Relative
Name of Toaster Quality Score Toaster Price Price Ratio
1. Cuisinart CPT-60 128 $70 215
2. Sunbeam 119 28 85
3. KitchenAid 117 77 237
4. Black & Decker 112 25 77
5. Cuisinart CPT-30 109 40 123
6. Breadman 107 35 108
7. Proctor-Silex 22425 104 15 46
8. Krups 101 32 98
9. Oster 94 45 138
10. Toastmaster B1021 91 16 49
11. Proctor-Silex 22415 87 35 108
12. Toastmaster B 1035 84 21 65
13. Betty Crocker 84 25 77
14. Proctor-Silex 22205 83 11 34
15. Rival 80 13 40
Average 100 $33 100
106 Chapter Three
250
3
1
200
150
9
Relative price
5
6
100 11
8
2
13
4
12
50
10 7
15
14
Figure 3.3 Customer Value
0 Map of Toasters
0 50 100 150
Market-perceived quality score
By using the data from Table 3.5, we can draw the customer value
map shown in Figure 3.3.
In the customer value map, the products above the diagonal line are
the products that have low customer value:
Products 1 and 3 (Cuisinart CPT-60 and KitchenAid) have good

market-perceived quality scores but very high prices.
Product 9 (Oster) has below average market-perceived quality but a
high price.
Products 5, 6, and 11 (Cuisinart CPT-30, Breadman, and Proctor-Silex
22415) are also relatively high in price and low in performance, but
they are close to the fair value zone.
Products 2, 4, and 7 (Sunbeam, Black & Decker, and Proctor-Silex)
are low-price, high-performance products.
Products 10, 14, 15, and 12 (Toastmaster B1021, Proctor-Silex 22205,
Rival, and Toastmaster B 1035) are low in price and have reasonable
performance.
Products 8 and 13 (Krup and Betty Crocker) have better than average
customer value but are in the fair value zone.
Overall, the products that are located in the lower right of the chart are
better customer value products, and the further they deviate from the
fair value line, the better customer value the product has. In Figure 3.3,
products 2, 4, and 7 are in the lower-right portion and are the furthest
from the fair value line, so they are the products with the best customer
value. Product 2 has a higher price, so it is a best-value product at the
higher price level. Product 7 has a lower price, so it is a best-value prod-
uct at a low price level. Similarly, the products that are located at the
upper-left corner of the chart are products with worse customer value,
and the further they are from the fair value line, the worse customer
value the product has.
3.2.4 Competitive Customer Value Analysis

A competitive customer-value analysis is a graph that compares prod-
ucts in important aspects of customer value. This analysis can show you
the areas where you can improve your product most effectively.
Table 3.6 shows a market-perceived quality profile for two print-
ers, printer A (our printer) and printer B (a competitors printer), and
Table 3.7 shows the market-perceived price profile.
TABLE 3.6 Market-Perceived Quality Profile of Two Printers

Industry Comparison Quality Scores
Printer Ratio (Ours/
Customers A (Our Competitor)
Weight of Printer) (Ratio>1.0
Attributes (1 = lowest, Printer B means
Quality (Total = 10 = (Competitors better than Customer
Attributes 100) highest) Printer) competitor) Weight x Ratio
Machine 25 8 7 1.14 = 8/7 28.5 = 1.14 25
Up Time
Print Speed 15 9 8 1.13 17.0 = 1.13 15
Image 15 7 8 0.88 13.2
Quality
Ease of Use 5 4 7 0.57 2.85
Service 15 5 7 0.71 10.65
Response
Time
Repair Time 15 5 6 0.83 12.45
Quality of 10 7 7 1.0 10.0
Service
Total 100 94.65 =
market-
perceived
quality score
108
TABLE 3.7 Market-Perceived Price Profile: Printers

Industry Comparison
Customers Weight Printer B Quality Scores
Price Satisfaction of Attributes Printer A (Competitors
Attributes (Total = 100) (Our Printer) Printer) Relative Price Ratio Customer Weight x Ratio
Purchase Price 40 $508 $585 0.87 = 508/585 34.8 = 0.87 40
Service and Repair 30 $60/year $65/year 0.92 = 60/65 27.6 = 0.92 30
Toner 20 $235/year $235/year 1.0 20
Paper 10 $124/year $101/year 1.23 12.3
Total 100 94.7 = relative price
ratio
0.6 0.7 0.8 0.9 1.0 1.1 1.2 1.3 1.4
Machine Up Time
Print Speed
Image Quality
Ease of Use
Service Response Time
Repair Time
Quality of Service
Figure 3.4 Area-to-Area Customer Value Chart for Printers
Figure 3.4 shows an area-to area customer-value chart that compares

printer A and printer B. Each bar represents a market-perceived quality
characteristic. The horizontal dimension of the bar shows how much the
product is better or worse than the competitors, and the thickness of
each bar is proportional to the relative importance of each characteristic.
So the total white area represents our advantage, and the total shaded
area represents our disadvantage. Our goal is to maximize the white
area and minimize the shaded area in the most effective way.
Figure 3.5 shows a head-to-head market-perceived price ratio chart
for printers, and each bar represents a customer cost component. Its
horizontal dimension represents how much better or worse our products
price is compared with the competitors price, and the bars thickness
represents the relative importance of that cost component in the eyes
of customers. So the total white area minus the total shaded area rep-
resents our products cost advantage. The larger the cost advantage, the
more competitive our product is in price.
3.3 Customer Value Deployment

After we have completed our competitive customer-value analysis and
relative price-ratio analysis, we need to find an effective way to overcome
our disadvantages and strengthen our existing advantages to improve our
customer value and win over the competition. To do that, we need to iden-
tify the critical areas of the company that are related to our key market-
perceived quality factors and market-perceived customer price areas.
110 Chapter Three
0.6 0.7 0.8 0.9 1.0 1.1 1.2 1.3 1.4
Purchase Price
Service and Repair
Toner
Paper
Figure 3.5 Area-to-Area Market-Perceived Price Ratio Chart for Printers
A quality function deployment (QFD) template can be very useful in

deploying key customer values into our process improvements.
Table 3.8 shows the customer value deployment matrix for the printer
example. In this matrix, correlation scores of 9, 3, 1, and 0 are used. A
score of 9 means very much related, a score of 3 means related, a score
of 1 means slightly related, and a score of 0 means not related.
TABLE 3.8 Customer Value Deployment Matrix for Printer Example
Sales
Quality and
Attributes Design Manufacturing QC Service Distribution Marketing
Machine 3 9 9 0 0 0
Up Time
Print 9 3 3 0 0 0
Speed
Image 9 9 9 0 0 9
Quality
Ease of 9 0 0 9 0 3
Use
Service 0 0 0 9 9 0
Response
Time
Repair 0 0 0 9 9 0
Time
Quality of 0 0 0 9 0 0
Service
For example, in the machine up time category, quality control and

manufacturing are very critical in ensuring printer dependability; prod-
uct design is also related to the dependability of the printer. In the ease
of use category, design is very important in creating a printer that is easy
to use. However, sometimes there is a gap between customer-perceived
quality and the real quality level. For example, printer A may actually
be easy to use, but because of poorly written customer instructions, poor
service support, and poor marketing, significant numbers of custom-
ers may think that printer A is hard to use. To overcome this problem,
redesigning the printer may not be the right solutiona comprehensive
strategy that includes improving customer service, rewriting customer
instructions, and developing a better marketing message might be the
best approach.
3.4 Evolution of Customer ValuesBlue

Ocean Strategy
In todays marketplace, we are flooded by all kinds of products and
services. With the development of information technology and a global
economy, the pace of product development and cost cutting is acceler-
ating. You could have a top-notch product with high customer value,
and thus high customer appreciation, and your product could have a
commanding lead in the marketplace this year. But very quickly, you
will find competitors growing like mushrooms, and soon there will be
many similar products in the marketplace with reasonably good quality,
similar product characteristics, and cutthroat prices. You will be forced
to enter a bloody competition, head-to-head, to compete on every item of
the current customer values. You will be forced to make improvements
on every item of the current customer values, but still cut prices. With
this brutal competition, your profit margin will start to disappear.
This scenario is very common in todays marketplace in the airline
industry, automotive industry, and many others. This situation is called
the red sea, as an analogy to fishermen along a crowded coastline
competing for dwindling fish. In general, the red sea refers to a satu-
rated market with fierce competition, crowded with companies provid-
ing similar services or goods.
The opposite of the red sea is the blue ocean, which is an analogy
to fishermen in a wide-open blue ocean, with plenty of fish and no com-
petition. In general, the blue ocean refers to an untapped and uncon-
tested market, offering little or no competition, since the market is not
crowded. To create a blue ocean, you need to do something different from
everyone else, produce something that no one has yet seen. The blue
ocean strategy is a business strategy for capturing uncontested market
space, making competition irrelevant (Kim and Mauborgne 2005).
112 Chapter Three
Specifically, the blue ocean strategy enables a company to unilater-

ally alter the commonly adopted key customer value items to form a
new set of customer values. Thus a new type of product or service is
formed, without competition. Example 3.2 illustrates how the blue ocean
strategy works.
Example 3.2 U.S. Wine Industry The United States has the third
largest consumption of wine worldwide, and this $20 billion wine in-
dustry is very competitive. California wines dominate the domestic
market, capturing two-thirds of all U.S. wine sales. California wines
compete head-to-head with imported wines from France, Italy, Spain,
and South America. In recent years, there are also many emerging
wine producers from such places as Oregon, Washington, and New York
State. The number of wine brands keeps increasing, but overall market
size remains stagnant, so the wine market becomes a typical red sea
market. (Kim and Mauborgne 2005)
In the U.S. wine market, almost all wine makers compete based on
the following seven factors:
Price per bottle of wine

An elite, refined image in packaging, including labels announcing the
wine medals won and the use of esoteric enological terminology to
stress the art and science of wine making
Above-the-line marketing to raise consumer awareness in a crowded
market and to encourage distributors and retailers to give prominence
to a particular wine house
Aging quality of wine
The prestige of a wines vineyard and its legacy (hence the appella-
tions of estates and chateaux, and references to the historic age of the
establishment)
The complexity and sophistication of a wines taste, including such
things as tannins and oak
A range of wines to cover all varieties of grapes and consumer prefer-
ences from Chardonnay to Merlot, and so on
Based on these seven factors, Kim and Mauborgne analyzed the wine
market and drew the graph shown in Figure 3.6. The seven factors are
shown on the horizontal axis, and the vertical heights indicate the level
or values of each factor. For example, the premium wines have higher
value in price than the budget wines, which means that the premium
wines are higher in price. For other items, such as aging quality, the
higher value for premium wines means that premium wines have a
High
Premium Wines
Budget Wines
Low
Price Above-the-line Vineyard prestige Wine
Use of enological marketing Aging and legacy Wine range
terminology and quality complexity
distinctions in wine
communication
Figure 3.6 Value Curves of the U.S. Wine Market
higher aging quality than budget wines. Kim and Mauborgne (2005)
call this graph the value curves, or strategy canvas.
We can easily see that this value curve has a lot of similarities with
customer values. In Kim and Mauborgnes value curves, price is always
the first item, and in a customer value analysis, price, and in the broader
sense, the market-perceived price profile, is also always an important
customer value metric. In Kim and Mauborgnes value curves, factors
other than price are always key competitive factorsthey are very simi-
lar to the market-perceived quality profile in the customer value. We can
call all these factors the value factorsprice is always one of the value
factors, and the other value factors are quality-related factors.
One of the differences between Kim and Mauborgnes value curves and
a customer value analysis is their graphing method. Figure 3.6 shows that
although there are more than 1600 wineries in the U.S. wine industry, most
wines fall into one of two categories: premium wines are high in price, but
are also rated high in the remaining six key value factors, so premium
wines are high-quality, high-price wines; budget wines are lower in price,
but are also rated lower in the remaining six key value factors, so budget
wines are low-price, low-quality wines. The striking fact is that all wines
are competing head-to-head on the same seven factors. The premium wines
are trying to win by increasing the ratings of quality, which is indicated by
the six value factors (other than price). Increasing these quality aspects
114 Chapter Three
requires investment, thus increasing the cost, and to recover the cost, the
premium wines have to sell at a higher price. The budget wines are selling
at a lower price, and to save production costs, the ratings of the six value
factors (other than price) will have to be lower. When the overall market
size is limited, competing either way is not very profitable, and the potential
for growth is very poor.
The key for a blue ocean strategy is to systematically change these
value factors:
Which value factors that the industry takes for granted should be
eliminated?
Which value factors should be reduced well below the industrys
standard?
Which value factors should be raised well above the industrys
standard?
Which value factors should be created that the industry has never
offered?
Casella Wines, an Australia winery, noticed that almost all wine

makers focus on overdelivering on prestige, and on the traditionally
accepted quality factors at their price points. However, Casella Wines
noticed that the alternative drinks to wine, such as beer, spirits, and
ready-to-drink cocktails, were a market three times as big as tradi-
tional wines. Casella also found that a lot of American consumers think
traditional wines are too complex, confusing, and difficult to appreci-
ate. These consumers think that beer and ready-to-drink cocktails are
sweeter and easier to drink. Based on careful analysis, Casella Wines
developed a brand new wine called Yellow Tail. Yellow Tail is a wine that
has a new combination of wine characteristics that produced an uncom-
plicated wine structure. The wine is soft in taste and approachable like
ready-to-drink cocktails and beers. The sweet fruitiness of the wine also
keeps peoples palate fresher, allowing them to enjoy another glass of
wine without thinking about it. The result was an easy-to-drink wine
that does not require years to make. In line with this fruity sweetness,
Yellow Tail dramatically reduced or eliminated all the value factors that
the wine industry has long competed ontannins, oak, complexity, and
aging. Figure 3.7 illustrates the value curves of this new wine.
We can easily see that Yellow Tails value factors are very different from
those of conventional wines. Three value factors for conventional wines
wine packaging, marketing, and aging quality of wineare reduced to a
minimum; another three value factors for conventional winesvineyard
prestige, wine complexity, and wine rangeare also reduced; and three
completely new value factors are addedeasy drinking, ease of selection,
and fun and adventureand they are positioned at a very high level.
High
Premium Wines
[Yellow Tail]
Budget Wines
Low
Price Above-the-line Vineyard prestige Wine Ease of
marketing and legacy range selection
Use of enological Aging Wine Easy Fun and
terminology and quality complexity drinking adventure
distinctions in wine
communication
Figure 3.7 Value Curves of Yellow Tail
Yellow Tail is a great success. Two years after its launch, it became
the fastest-growing brand in the history of both the U.S. and Australia
wine industries, and it is the number one wine imported into the United
States, surpassing the wines of France and Italy. By August 2003, it was
the number one red wine in 750ml bottles sold in the United States.
Moreover, this happened in the context of a global wine glut.
In summary, Yellow Tails drastic change in value factors created a
social drink for everyone: beer drinkers, cocktail drinkers, and non-wine-
drinkers. This new wine works on different value factors and opened
up a much broader market with very little competition. Thus it created
a blue ocean.
We can derive the following observations:
1. In a red sea marketplace, every competitor delivers a similar product

or service with an almost identical set of value factors; the competi-
tion is head-to-head; each competitor either delivers a high-price,
high-quality product, or a low-cost, low-quality product; the market
will soon be saturated; and profitability will be very marginal.
2. Customers value systems may change over time, even for the same
product or product class. The current customer value factors that
the industry competes with may have been established long in the
past, and customers may welcome a change in these customer value
factors.
116 Chapter Three
3. If a producer captures the right strategy in changing customer value

factors, and develops a new set of customer value factors that resonate
with customers desires, the producer will create a blue ocean for its
products with great market growth potential and little competition.
Clearly, creating a blue ocean is like hitting a jackpot or a gold mine.

In the business literature, we can see many success stories resulting
from bold changes in customer value factors, but there are also many
blunders and failures due to bad moves. How can you formulate the
right changes in customer value factors to create blue oceans, and are
they even possible? Kim and Mauborgne proposed several ideas for
systematically formulating blue ocean strategies by making the right
changes in customer value factors.
3.4.1 Formulating a Blue Ocean Strategy

Kim and Mauborgne noticed that many companies tend to do the fol-
lowing in their business practices:
Define their own industry as others define it, and fight head-to-head
to be the best.
Stick to their industrys accepted strategic groups, such as luxury
automobiles or pickup trucks, and strive to stand out in their strategic
group.
Focus on the same buyer group, be it purchasers (for office equipment
industry), or retail customers (in the clothing industry), or influencers
(such as doctors for the pharmaceutical industry), even though there
are usually many tiers of buyers.
Define the scope of the products and services offered by their industry
in much the same way as their competitors do.
Adopt customer value factors similar to their competitors, and set
similar priorities on functional and psychological benefits.
Focus on the same point in time, and often on the current competitive
threat.
These commonly adopted business practices are actually the main

forces that lock businesses into red seas. In order to break out of red
seas and into a blue ocean, you need to go beyond these business prac-
tices. However, breaking out of the currently accepted customer value
factors in a given industry is also risky. You cannot arbitrarily aban-
don a well established set of customer values and replace them with
unknowns.
Based on many years of research and observation, Kim and Mauborgne

proposed formulating the blue ocean strategy by looking at the follow-
ing six aspects:
Alternative industries
Strategic groups
Buyer groups
Complementary product and service offerings
Functional-emotional orientation
Time
3.4.1.1 Looking Across Alternative Industries For a given industry, the

alternative industries are the industries that provide different forms of
products or services that serve the same purpose. For example, in one
sense, restaurants and theaters serve very different kinds of products,
but both restaurants and theaters give customers nice places to be and
activities for their evening time. Another example is tax filingyou
can hire a CPA, do it yourself, or buy tax-filing software. These are very
different choices, but they serve the same purpose.
Different industries within the same alternative industries class may
have quite different customer value factors; for example, the customer
value factors for a restaurant are very different from those of theaters,
and the customer value factors of a CPA are quite different from those
of tax-filing software. However, these alternative industries serve the
same customer groups for the same or similar purposes. In this case,
some crossover of the features of these alternative industries can create
new products or services that will be smashing successes. The case of
Cirque du Soleil is illustrated in Example 3.3.
Example 3.3 Cirque du Soleil Cirque du Soleil (French for Circus of

the Sun) is a successful entertainment enterprise based in Montreal,
Quebec, Canada, and founded in Quebec in 1984. Though Cirque du
Soleil is called a circus, it doesnt involve many animals as other tradi-
tional circuses do; it mainly depends on human performance. Cirque du
Soleil not only inherited the tradition of circus, but it also integrated
many elements of street performances and busking, opera, ballet, and
rock music into its performance. Usually, the traditional circus empha-
sized stunts and skills in the performance without paying too much
attention to story lines. Cirque de Soleils performance is centered on
story lines. Cirque du Soleil is also characterized by its beautiful colors,
nice live music, and a mix of talents from all over the world.
118 Chapter Three
From the customer value factors point of view, Cirque combined the
customer value factors from several alternative industries: regular circus
shows, opera, Broadway-style musical shows, Las Vegas style shows, and
even Chinese acrobats. This cross-pollination created a unique enter-
tainment style and a great success. Cirque du Soleil attracted custom-
ers for both the traditional circuschildrenas well as customers who
would usually not go to see a circuscorporate clients and adults.
Figure 3.8 shows the value curve for Cirque du Soleil. You can see that
it eliminated four customer value factors that are commonly adopted
by a regular circusstar performers, animal shows, aisle concessions,
and multiple show arenas. Cirque du Soleil kept three customer value
factors for a regular circusfun and humor, thrills and danger, and a
unique venueand it introduced four new customer value factors
theme, refined watching environment, multiple productions, and artistic
music and dance. These last four customer value factors are associated
with musicals, Las Vegas shows, and others. Cirque du Soleils new
and unique customer value curve makes it a unique circus, and it has
become so successful it is the largest cultural export of Canada.
3.4.1.2 Looking Across Strategic Groups Within Industries In any in-

dustry, there are usually several strategic groups. A strategic group is
usually a group of companies or entities within an industry that pursue
a similar strategy. For example, within the automotive industry, there
High
Ringling Bros. and Barnum & Balley Value Curve
Cirque du Soleil
Value Curve
Smaller
Regional
Circuses
Low
Price Animal Multiple Thrills and Theme Multiple
shows show danger productions
Star Aisle arenas Fun Unique Refined Artistic
performers concessions and venue watching music
humor environment and dance
Figure 3.8 Value Curves of Cirque du Soleil

are several strategic groups, such as luxury cars, family cars, and low-
price cars. Within the restaurant industry, there are fast food restau-
rants, family restaurants, and high-end restaurants.
Usually, strategic groups can be partitioned by price and performance;
that is, high price and high performance, medium price and medium per-
formance, and low price and low performance. The distinction between
the luxury car, family car, and low-end car is one example of this kind of
partition. Strategic groups can also be partitioned by functional varia-
tions within an industry. For example, in the automotive industry, there
are strategic groups such as family sedans, trucks, and vans. Different
strategic groups with an industry usually also have different customer
value factors. Crossover of customer value factors among different stra-
tegic groups may lead to breakthrough products or services.
In the automotive industry, Toyotas Lexus combined the quality of
high-end Mercedes, BMW, and Jaguar, with the price level of the Cadillac
and Lincoln, thus creating a new market segment of Lexus consumers.
This is an example of crossover between two strategic groups within the
automotive industry. Similarly, the Chrysler minivan was a crossover
between family cars and regular vansthe minivan drives like a car,
but it is more spacious and has more seats and cargo space like a van.
Chryslers minivan was a great success that led to the SUV (sport utility
vehicle) revolution for American automobile makers.
3.4.1.3 Looking Across the Chain of Buyers For most industries, there
is usually a chain of buyers for the industrys products or services.
For cloth manufacturers, their products may be bought first by retail
chains; the products are then distributed to retail stores, and finally
the products in retail stores are bought by individual shoppers. For
the pharmaceutical industry, the products (medicines) are first bought
by drugstore chains, and then they are distributed to individual drug-
stores. For prescription drugs, medical doctors will write the prescrip-
tion, and the actual users will bring the prescription to the drugstore
and get the medicine, although the medicine may be paid for by insur-
ance companies.
The chain of buyers can be quite complicated, but generally there are
three types of people in this chain of buyers: purchasers pay for the prod-
ucts or services; users actually use the products or services; and influ-
encers play very important roles in the buying decisions. For example,
doctors are the influencers in the pharmaceutical market place. These
three types of people may overlap, but often they do not.
Though the products or services are used solely by users, the man-
ufacturers usually pay the greatest attention to the most influential
player in the buying decision-making process. For example, cloth manu-
facturers may pay a lot more attention to buyers from big chain stores,
120 Chapter Three
such as JCPenney, Marshall Fields, and Wal-Mart, because the bulk of

their products will be sold to them. Pharmaceutical companies usually
pay most of their attention to medical doctors, because it is doctors who
write the prescriptions that lead to the sales of medicines. This conver-
gence of attention on the most influential players in the purchasing
decision usually leads to a fixed marketing and product strategy for the
whole industry. Soon, every competitor in the industry plays the same
game, and the whole marketplace is a red sea.
A path to business innovation is to look across the chain of buyers.
Instead of focusing on the most influential player like most competi-
tors, you can focus on other players in the chain of buyers. For example,
Internet-based stores sell goods directly to users, instead of to retail
chains, or stores. Another example is Novo Nordisk, a Danish insulin
producer, discussed in Example 3.4.
Example 3.4 Novo Nordisk Insulin is used to regulate the level of

blood sugar for diabetes patients. Historically, the insulin industry
focused its attention on the doctors, who are the key influencers in
medicine-buying decisions. Before the 1980s, the insulin industry
was competing on who can produce the purest insulin. By 1980, the
technology of producing insulin of the highest purity was developed
and now many producers can do this. As a very common scenario,
when many competitors have similar technological capabilities, the
brutal price competition phase began in and the profit margin for each
producer was quickly disappearing.
Novo Nordisk, a Danish insulin producer, who also had this insu-
lin purification technology, thought about another strategy other than
this price competition. Novo found that diabetes patents were supplied
with insulin in vials. But self-application of these insulin vials involves
the complex and unpleasant task of handling syringes, needles, and
insulin, and of administering doses according to the individuals needs.
Instead of competing on producing high-purity insulin vials to impress
doctors, Novo developed a product called NovoPen in 1985, the first
user-friendly insulin delivery solution, which was designed to remove
the hassle and embarrassment of administering insulin. The NovoPen
resembled a fountain pen; it contained an insulin cartridge that allowed
the patient to easily carry, in one self-contained unit, roughly a weeks
worth of insulin. The pen had an integrated click mechanism, making
it possible for even blind patients to control the dosing and administer
insulin. Patients could take the pen with them and inject insulin with
ease and convenience without the embarrassing complexity of syringes
and needles.
This product is a hit, not because medical doctors liked it, but because
many patients requested NovoPen for its easy and hassle-free application.
After NovoPen, Novo Nordisk introduced NovoLet, which is a prefilled

disposable insulin injection pen with a dosing system that provided users
with even greater convenience and ease of use. And in 1999 it brought out
the Innovo, an integrated electronic memory and cartridge-based delivery
system. Innovo was designed to manage the delivery of insulin through
built-in memory and to display the dose, the last dose, and the elapsed
timeinformation that is critical for reducing risk and eliminating wor-
ries about missing a dose.
In this case, Novo Nordisk shifted its attention from medical doctors
to the patients themselves. Novo Nordisk developed a product that fits
the customer value of diabetes patients rather than that of medical
doctors. This product made Novo Nordisk the dominant player among
insulin producers.
3.4.1.4 Looking Across Complementary Product and Service Offerings

Most products or services are not consumed alonethey are consumed
with other products or services. In the automobile industry, cars and
trucks are the main products, but without auto insurance, auto repairs,
auto parts suppliers, and so on, consumers would not be able to use cars
and trucks in the long run. In this case, insurance, auto repairs, and
parts are complementary products and services for the automobile in-
dustry. For most industries, there are many hidden untapped values in
complementary products and services. To open up the market boundar-
ies and create blue oceans, we should not only focus on the original in-
dustry itself, but also on these complementary products and services.
For many utility industries, such as electrical power generation, pro-
ducing and selling power generation products, such as turbines, is the
main focus. However, there are many competitors in this market, and
power-generation equipment usually has very long life cycles, as much
as 30 to 40 years. The number of units that each company can sell fluc-
tuates greatly. If equipment producers only count on their revenue from
equipment sales, they will be in a very precarious position. However,
the demand for complementary products and services, such as parts,
upgrades, and maintenance, is very steady and these items can be cash
cows. As a result, most equipment manufacturers will not only provide
equipment, but will also be very actively involved in providing these
complementary products and services.
Many successful enterprises look at both the core products and ser-
vices, and at complementary products and services to create highly
successful business models. Starbucks not only provides coffee, but it
also provides other services associated with coffee consumption, such
as meeting and chatting places, Internet connections, and nice set-
tings. Compared with traditional bookstores, where only books are
sold, Borders and Barnes & Noble redefined the scope of the services
122 Chapter Three
they offer. They transformed the product they sell from the book itself,
into the pleasure of reading and intellectual exploration, adding lounges,
knowledgeable staff, and coffee bars to create an environment that cel-
ebrates reading and learning. Both Borders and Barnes & Noble are
very successful.
3.4.1.5 Looking Across Functional or Emotional Appeals to Customers As

mentioned earlier in this chapter (Section 3.1), for every product or ser-
vice, the customer value always has functional and psychological benefits.
Usually one of the most important psychological benefits of a product or
service is the emotional benefit or emotional appeal of the product. For ex-
ample, a very successful businessman may like to drive a luxury car. From
a functional point of view, a car provides a means of transportation, and
to serve this functional purpose, an economy car is sufficient. However,
a luxury car is also a status symbol, and the businessman who drives it
shows that he and his business are both successful.
In a given industry, some products compete mostly on price and func-
tion, and others compete largely on emotional appeal. When a product
adopts its unique combination of functional and emotional benefits,
it tends to stick with them. The producer of the product gets familiar
with the strategy, and often adopts the mentality of if its not broken,
dont fix it. On the other hand, customers of the product also get used
to it and develop a fixed customer image for the product. When sev-
eral products are adopted with similar combinations of customer value
components, you get the red sea type of competition. To break out
this market boundary, you can think about redefining the combination
of functional and emotional benefits to create a unique product and a
blue ocean market.
For example, in the airline industry, almost all airlines adopted simi-
lar customer value strategies and competed on the same things, such
as getting high-end business travelers. This meant that many airlines
built business lounges, supplied good meals, and so on. This approach
offers a fair combination of functional and emotional benefits. However,
Southwest Airlines adopted a very different combination of customer
values: it reduced a lot of the emotional benefits, reducing the quality
of meals, not offering luxury lounges, and so on. But Southwest drasti-
cally increased the convenience of boarding airplanes and offered fre-
quent and flexible departures, and it offered this more functional service
with much lower cost. The bulk of customers are willing to take a very
basic, but reliable and safe airline service, with low cost. They dont
care much about in-flight meals, business lounges, and so on. Southwest
achieved a lot of success by adopting a low-cost, purely functional cus-
tomer value.
There are also many success stories that work the other way around,
that is, adding more emotional benefits on top of functional benefits.
For example, regular coffee shops only provide coffee, and if you adopted
a function-based customer value combination, you might just try to
improve the quality of the drink by using high-grade coffee beans, making
fancy coffee drinks, and selling them at a higher price. Starbucks Coffees
strategy can be seen as a fair combination of functional benefits and
emotional benefitsit provides reasonable quality coffee drinks, but on
top of that, Starbucks provides a social setting, a place to take a break
or to have brief business meetings, and so on. It adds a lot of emotional
benefits.
3.4.1.6 Looking Across Time All industries are subject to external

trends that will profoundly affect them over time. These trends include
technological developments, life-style evolution, climate change, global
economic changes, and so on. For example, the rapid development of
computational technologies, data storage technologies, and microelec-
tronics will greatly affect the way people communicate and interact
with each other. When technology or lifestyle undergoes a quantum
leap, it greatly affects customers attitudes and value systems.
Many companies only pay attention to the needs of the marketplace in
the very short term, and fail to pay attention to these trends. Companies
that pay attention to these trends and catch them ahead of time may be
able to develop breakthrough products much earlier than their competi-
tors and capture a huge share of the market. One such example is the
development of the Toyota Prius hybrid automobile. One fundamental
trend in the world is that nonrenewable resources will get scarcer and
scarcer. When Toyota put in a lot of effort to develop the Prius in the late
1990s, petroleum was still quite cheap. If Toyota had just paid attention
to the short-term market, it would not have made sense to develop this
high fuel-efficiency, but still immature technology. However, the drastic
oil price hikes in the early 2000s made fuel efficiency one of the most
important factors in car-purchase decision making. The reputation of
Toyotas fuel efficiency and of the Prius, in particular, greatly increased
Toyotas market share worldwide.
3.5 Customer Value and the Voice

of the Customer
So far in this chapter, we have thoroughly considered the concept of cus-
tomer value. Customer value consists of key factors that determine how
well customers will appreciate a given product or service. Investigating
and analyzing customer value and planning with it in mind will help
you determine the strategic positions for products or services in the
124 Chapter Three
marketplace. In a broader product or service development process, how-

ever, more needs to be done.
Product development is an information mining, transformation, and
creation process, as illustrated in Figure 3.9. For a customer-centric prod-
uct development, the first step is to mine the information about what
customers really want. This information is the customer attributes. Even
for a product based on new technology, knowing what customers like is
still very important. Mining the customer attributes is often referred to
as capturing the voice of the customer. As you can see in Figure 3.9, com-
plete information about what customers really want is the foundation for
a good product development process because it is both the first step and
the source of all information regarding product development.
The first mapping in the product development process is to trans-
form customer attributes into functional requirements for the product
or service. Functional requirements are a minimum set of indepen-
dent requirements that completely characterize the functional needs
of customers with regard to the product or service that they purchase.
Specifically, products or services are designed and produced to deliver
functions, and customers purchase or use products for their functions.
The functional requirements define those functions.
For example, customers purchase and use pens to make marks on
paper, so making marks is one of the main functions of a pen. For
complex products, such as an automobile, the functions that it delivers
are usually more complex and hierarchical. Automobiles can deliver
many functions such as
Moving people from A to B

Providing a nice riding environment
Providing a nice image for passengers
Matrix A Y=f(x)
CAs FRs DPs PVs

. . . .
. . . .
. . . .
Matrix B
Y=f(x)
Customer Functional Design Process
Attributes Requirements Parameters Variables
Figure 3.9 Product Development as a Mapping Process

For each of these main functions, such as moving people from A to B,

there are many details, such as move with varying speed, change direc-
tions, stop, and parkthese are the second-level functions. And for
each second-level function, such as move with varying speed, there are
many third-level functions, and so on.
The second mapping in the product development process is the trans-
formation from functional requirements to design parameters. Design
parameters are the elements of the design solution in the physical
domain that are chosen to satisfy the specified functional requirements.
For example, the power-train system in an automobile is the design
parameter that provides power to move the car around. Clearly, design
parameters can be very complex and hierarchical, taking the form of
systems, subsystems, components, and parts.
The last mapping is the transformation from design parameters to
process variables. Process variables characterize the manufacturing
process that satisfies the design parameters. Essentially, this transfor-
mation involves figuring out how to design the manufacturing process
and equipment to make the physical products. The process variables
are usually complex and hierarchical as well.
From Figure 3.9, you can see that the nature of product development
is information mining, creation, and transformation. If we do not cap-
ture the VOC data accurately in the first stage of product development,
we may develop products that customers do not like. If we do not capture
sufficient amounts of VOC data, we will not be able to link every product
requirement to what customers really like when we develop functional
requirements, and we would have to fill in those customer requirements
arbitrarily, or by guesswork. This could lead us to develop products with
the wrong performance specifications.
The transformation from functional requirements to design param-
eters also depends on the accuracy and completeness of the VOC data we
capture at the beginning of the product development process; if we derive
the wrong functional requirements, we will end up transforming these
wrong requirements into wrong product design parameters, and we will
develop the wrong product. On the other hand, the transformation from
functional requirements to design parameters requires a good under-
standing of technology, science, engineering, and economics, and it also
requires knowledgeable product development team members, as well as
a good product development process. Similarly, the transformation from
design parameters to process variables also depends on the accuracy and
completeness of the VOC data captured, as well as on knowledge and a
good product development system.
We can conclude the following:
Importance of VOC data Clearly, customer value defines the

strategic position of the product or service in the marketplace.
126 Chapter Three
Information on the customer value of our product can only come

from the voice of the customer. In detailed product development, the
voice of the customer is also the source of information on the product
development process; we need to get enough information so that the
three mappings in the product development process can be carried
out flawlessly. Therefore, the VOC is the source of information for both
the strategically important customer value proposition, and all the
building blocks in the product design.
Accuracy of VOC data Because of the importance of the voice of
the customer, we need to capture VOC data accurately. Only with accu-
rate VOC data can we develop an accurate customer value proposition
and product performance specification. There are many methods for
capturing the voice of the customer, but arbitrary use of these meth-
ods will not lead to capturing the voice of the customer accurately.
We will discuss how to accurately capture the voice of the customer
in Section 3.6.
Completeness of VOC data capturing Not only do we need to
accurately capture the voice of the customer, but we also need to cap-
ture a sufficient amount of VOC information to define customer value
and benchmark our customer value position with our key competitors.
Sometimes, we need to collect enough VOC information to figure out
ways to redesign our customer value position to create breakthrough
new products. The voice of the customer is also the source of infor-
mation for product development, and we need to capture sufficient
VOC data to drive the product development process. Specifically, we
need to get sufficient VOC information to drive the transformations
from customer attributes to functional requirements, from functional
requirements to design parameters, and from design parameters to
process variables.
3.6 Capturing the Voice of the Customer

Quality can be defined as meeting customer needs reliably and provid-
ing superior value, and meeting customer needs requires that those
needs be understood. The voice of the customer is the term used to
describe the stated and unstated customer needs or requirements. The
voice of the customer can be captured in a variety of ways, including cus-
tomer interviews, customer surveys, focus groups, ethnographic studies,
and so on.
The purposes of capturing VOC data are usually the following:
1. Product development
2. Market and product planning
3. Customer feedback and product improvement

4. Customer-centric innovation and market opportunity search
For different purposes, the objectives of capturing VOC data will be
different, as will the strategy and methods used. In general, though,
the objective of VOC capturing is to capture all relevant customer attri-
butes so we will have enough information to create detailed functional
requirements, design parameter specifications, and manufacturing pro-
cess variables for all details.
3.6.1 Plan for Capturing the Voice

of the Customer
Once our customer value proposition and product plan are established,
the next step is to plan how to capture the customers needs. This
includes determining who our target customers are, which customers to
contact in order to capture their needs, who will do VOC data collection,
what methods to use to collect the data, and scheduling and estimating
resources required to capture the VOC data.
3.6.1.1 Who Are the Target Customers There is no single voice of the
customer. Even for the same product, there are diverse voices from cus-
tomers about what they think the product should be like. For example,
in the automobile market, some customers would prefer larger sizes
and more power, while some customers would prefer low gas mileage,
low cost, easy maintenance, and so on. We usually can find that there
are several major segments in this market, and within a market seg-
ment, the customers opinion are similar; however, there are signifi-
cant differences among different market segments. In this case, we may
want to capture the voice of the customer from all market segments
and develop a portfolio of products to satisfy different needs. In order
to do this, we should target a representative set of customers from each
market segment to get complete information on customers needs.
For many products, there might be multiple voices from each buying
unit. When buying a car, a husband and wife may have different opinions.
When making a purchasing decision about commercial equipment in an
organization, the voice of the direct user, the voice of the purchasing
department, and the voice of support personnel might be very different.
In this case, all kinds of voices should be considered and the correspond-
ing product design should take into account all these voices.
For a given product, we also need to capture the voice of the following
types of customers:
1. Current customers. These are customers who are buying our product;
we will try to know what they want in order to keep them.
128 Chapter Three
2. Competitors customers. These are customers who need this kind

of product, but they are not buying our product, so we will try to
determine what they want in order to improve our product to capture
more market share.
3. Potential customers. These customers are not buying our product or
competitors products. They are not customers of our industry; we
will try to know why they are not and integrate their VOC data into
our product design, and hopefully they can be our customers in the
future.
4. Lead customers. These are either our current customers or competi-
tors customers. Lead customers are those customers who are the
most advanced users of the product, customers who are pushing
the product to its limits, or customers who are adapting an exist-
ing product to new uses. The voice of lead customers is important
to catch the future market trend and develop a new generation of
product.
How many customers should we talk to? The number depends on the
complexity of the product, diversity of the market, product use, and the
sophistication of customers. The goal is to get to the 9095% level in
capturing customer needs. Research for a range of products indicates
that, on average, this is 20 customers.
3.6.1.2 Who Should Collect VOC Data Traditionally, marketing depart-

ments have had responsibility for defining customer needs and product
requirements, but this approach has many flaws. First of all, design is
a mapping processthe voice of the customer will have to be trans-
formed into functional requirements, and usually it is the engineering
people who actually perform this transformation. If engineers are out
of the loop, and marketing people do all the work in capturing the voice
of the customer and compiling market study reports, and then throw
it all over the wall to the engineers, severe distortions will happen as
marketing language is translated into engineering language. The ac-
curacy of the voice of the customer will be severely compromised.
Second, marketing people are not engineers; they dont do engineering
design. It is up to engineers to define all the necessary functional require-
ments, and to derive design parameters in fine detail. If the marketing
people have not captured sufficient amounts of the right VOC data, the
engineers will not be able to derive accurate functional requirements for
the product. For example, in a power saw design project, the marketing
people might tell design engineers that customers would like a wider
handle for the saw. The design engineers might then be puzzled by why
a wider handle is needed, and how wide is wide enough.
Therefore, product development personnel need to be directly involved

in researching customer needs. This may involve visiting or meeting
with customers, observing customers using or maintaining products,
participating in focus groups, or rotating development personnel through
marketing, sales, or customer support functions. This direct involvement
provides them with a better understanding of customer needs, the cus-
tomer environment, and product use; it develops greater empathy on the
part of product development personnel, minimizes hidden knowledge,
overcomes technical arrogance, and provides a better perspective for
development decisions. These practices have resulted in fundamental
insights, such as engineers of highly technical products recognizing the
importance to customers of ease of use and durability as opposed to the
latest technology.
3.6.1.3 What Kind of VOC Data Should Be Collected? As we saw earlier,

the purpose of collecting VOC data is to get sufficient and accurate in-
formation to finish the product development process. Most importantly,
we need to get accurate and sufficient data to derive complete functional
requirements for the product from VOC data. This clear purpose will
determine what kind of VOC data should be collected.
According to Ulwick (2005), the following four kinds of VOC data are
not suitable for product development purposes: solutions, specifications,
needs, and benefits. However, three kinds of VOC data are good choices:
jobs to be done, desired outcomes, and constraints.
Solutions Many customers offer their own solutions for a particular

product design when they are asked what they want in a product.
For example, in a customer survey study about laptop computers,
many customers said that they wanted the laptop computer to have a
larger screen. However, in the product development process, technical
solutions are actually part of the design parameters, which are usually
derived in a later phase of the product development process. Most
customers are not engineers, technical specialists, or scientists, and
they do not always have the best solutions; their solutions may also
interfere with other product requirements. In the laptop computer
example, one of the manufacturers actually adopted the customers
solution and produced a laptop with larger screens, but they found
that very few customers wanted to buy this kind of laptop, because it
was too big to carry around. Later the computer manufacturer found
that customers actually want a clear image on the screen, which
can be produced with a screen the same size or smaller, but with
higher resolution. This example emphasizes that the first mapping
in the product development process is to transform the voice of the
130 Chapter Three
customer into functional requirements, which should be stated in a

solution-neutral environment. Functional requirements should not
be tied to any particular solutions. Once all the relevant functional
requirements are defined, we can determine the best set of technical
solutions from our design process. Customer-provided solutions
simply confuse the process.
Specifications Customers often give their own opinions on prod-
uct specifications, such as size, weight, color, shape, and so on. They
may want something like a wider handle, sleeker look, or lighter
weight. Again, specifications are part of the design parameters, and
design parameters should be derived after functional requirements.
Customers are not engineers, and accepting their specifications may
limit other design solutions. For example, when customers want wider
handles, they may actually want a better grip, which could be accom-
plished by many technical means. If we adopt a wider handle, we
may design a tool that is too big for many customers.
Needs Needs are often expressed as very vague, high-level descrip-
tions of the overall quality of products or services. They are typically
stated as adjectives and do not imply a specific benefit, and they give
no information about how and how much. For example, customers
often say that they want a product to be more reliable, dependable,
robust, or effective. These needs statements give some information
about what customers might care for, but they are very imprecise
statements; they do not provide enough information to derive any
meaningful functional requirements. For example, if customers say
that they want to have more reliable products, how reliable is reli-
able enough? The specific reliability requirement will still be left to
guesswork by design engineers.
Benefits Customers often use benefit statements to describe what
values they would like a new product or service to deliver. They often
use words like easy to use, faster, or better. Again, these state-
ments are imprecise and they do not provide sufficient information
to derive functional requirements.
Clearly, solutions, specifications, needs, and benefits do not provide

sufficient information to derive functional requirements. Now lets look
at the three types of VOC data that Ulwick recommended be collected.
Jobs to be done The jobs to be done are actually the functions that
a product or service is supposed to deliver. For example, the jobs to be
done for a razor include: remove hair, prevent bleeding, prevent
dry skin, cause little or no pain, and so on. The jobs to be done for
an energy drink may include quench thirst, help people focus, and
recover from exercise. Clearly if we can collect a complete list of jobs to
be done by a product, it will be easy to derive functional requirements.

At least we will know what functions a product should deliver.
Desired outcomes Desired outcomes are specific descriptions of

the end result for each job to be performed. In a power saw product
development project by Bosch (Ulwick 2005, 26145), Bosch collected
many customers statements about their desired outcomes for the
power saw:
Minimize the amount of kick that occurs when starting the saw
Increase the likelihood that the blade will begin cutting precisely
on the cut line
Minimize the amount of time the cut line is blocked from view when
making a cut
Minimize the amount of pressure that must be exerted to keep the
saw flat on the cutting surface
Minimize the likelihood that the blade guard will snag the material
Minimize the likelihood of the cut going off track when approaching
the end of the material/cut
Minimize the frequency with which the cord gets in the way of the
cut path
Minimize the frequency with which the cord (end plug) gets caught
on the material
Minimize the frequency with which the blade binds
Increase the number of cuts that can be made with a single blade
Minimize the amount of damage a worn or dirty table inflicts on the
material when a cut is being made
Minimize the amount of dust/debris that is generated by the saw
These are just a few of the 80 or so outcomes they captured from
customers. For most jobs, even those that may seem somewhat trivial,
there are typically 50 to 150 or more desired outcomes, not just a
handful. Companies must capture information about all the desired
outcomes in order to have sufficient information to derive correct
functional requirements and subsequent design parameters.
Constraints Constraints are physical, regulatory, and environ-
mental road blocks that prevent the product from doing the job.
Identifying the constraints often provides potential chances for devel-
oping breakthrough products. For example, one constraint for diabetes
patients to the use of home test kits is that the patients often have
shaky hands and blurred vision, and the current device needs precise
hand positioning. Obviously, this constraint prevents home testing
132 Chapter Three
from being done successfully. This constraint, however, provided an

unique opportunity for improvementRoche developed a new device
that does not need precise hand positioning, and this design improve-
ment enabled Roches product to increase its market share from 20%
to 45% (Ulwick 2005).
As you can see, VOC data in the form of jobs to be done, desirable out-
comes, and constraints will provide excellent information for deriving
functional requirements. They are a good type of VOC data.
3.6.1.4 Organize VOC Data Once data about customer needs has been
gathered, it must be organized. The mass of interview notes, require-
ments documents, market research, and customer data needs to be dis-
tilled into a handful of statements that express key customer needs.
Affinity diagramming is a useful tool to assist with this effort. Brief
statements that capture key customer needs are transcribed onto cards,
and a data dictionary that describes these statements of need is pre-
pared to avoid any misinterpretation. These cards are organized into
logical groupings or related needs. This makes it easier to identify any
redundancy and serves as a basis for organizing the customer needs.
In addition to stated or spoken customer needs, unstated or unspoken
needs or opportunities should be identified. Needs that are assumed
by customers and therefore are not verbalized can be identified by pre-
paring a function tree. Excitement opportunities (new capabilities or
unspoken needs that will cause customer excitement) are identified
through the voice of the engineer, marketing, or customer support rep-
resentative. These can also be identified by observing customers use or
maintain products and recognizing opportunities for improvement.
Chapter
Customer Survey Design,

4
Administration, and Analysis
One of the most important success factors in designing superior service

products is to know your customers. For a world-class company, the
design decision should be driven by the voice of the customer. Obtaining
detailed customer data is even more important for the survival of a
service institution because maintaining customers loyalty is extremely
vital. Customer surveys are essential tools that can yield information
about customer expectations, customer satisfaction, and strategies for
improvement.
Commercial enterprises use customer survey findings to formulate
marketing strategies and design their products. Television and radio
programs are evaluated and scheduled largely based on the results of
consumer surveys. Libraries, restaurants, financial institutions, and
recreational facilities use customer surveys to gather information about
customers and their desires.
Customer surveys use a sample of customers to learn about the whole
population of a customer base. In a customer survey, information is col-
lected from the sample and analyzed. The sample may contain as few as
30 people (a small sample) to as many as 30,000 or more people (a very
large sample). Survey methods are based on sound statistical principles,
and over 70 years of modern survey practices show that surveys based
on a relatively small sample are usually quite accurate.
4.1 Customer Survey Types

There are three customer survey methods: mail-out, in-person, and
telephone.
133
134 Chapter Four
4.1.1 Mail-Out Surveys

The mail-out survey involves mailing printed questionnaires to a sample
of predesignated potential respondents. Respondents are asked to com-
plete the questionnaire and mail it back to the survey researcher.
The advantages of the mail-out survey include the following:
Low cost Other survey techniques require trained interviewers,

and may incur high travel costs.
Convenience The questionnaire can be completed at the respon-
dents convenience.
Privacy Because there is no personal contact in mail-out surveys,
respondents may feel more comfortable, and their privacy is
preserved.
Lack of time pressure Respondents can take their time to complete
the questionnaire and consult their personal records if necessary.
The disadvantages of mail-out questionnaires include the following:
Lower response rate than other methods The mail-out survey

response rate usually ranges from 2030 percent in the worst case to
8590 percent in the best case. Therefore, adequate numbers of ques-
tionnaires need to be mailed out, and many follow-ups are generally
needed to achieve the desired sample size.
Comparatively slow It usually takes a few weeks for question-
naires to be returned for a mail-out survey.
Self-selection The survey researcher only receives the returned
questionnaires, and the people who choose to fill out and return the
questionnaires may not adequately represent the population. For
example, people who have low reading and writing proficiency may
never return their questionnaires, and thus will not be represented
in the sample.
4.1.2 In-Person Interviews

In-person interviews involve an interviewer directly talking to respon-
dents to get the information.
The advantages of in-person interviews include the following:
High response rate The response rate of in-person interviews is

much higher than that of mail-out surveys.
Ability to contact hard-to-reach populations Certain groups
are difficult to reach by mail or telephone, and in-person interviews
are the only way to reach them.
Customer Survey Design, Administration, and Analysis 135
Greater complexity Because of the direct interaction between

interviewer and respondents, complex questions can be asked, and
the questions can be explained.
Comparatively quick It takes less time to finish the survey com-
pared to mail-out surveys.
In-person interviews have the following disadvantages:
High cost In-person interviews can be very costly due to personnel

requirements, travel expenses, interviewer training, and so on.
Interviewer bias Interviewers may subconsciously introduce per-
sonal bias and affect respondents choices in answering the questions.
Greater stress The interviewing process is usually stressful for
both the interviewer and respondent.
4.1.3 Telephone Surveys

Telephone surveys collect information through telephone interviews
between trained interviewers and respondents.
Telephone surveys have the following advantages:
Rapid response Telephone interviews usually get results more

quickly than in-person interviews and mail-out surveys.
Lower cost The cost of a telephone survey is usually much less
than using in-person interviews, and can be less costly than mail-out
surveys.
Privacy Respondents privacy is better preserved than when using
in-person interviews.
Telephone surveys also have the following disadvantages:
Less control The interviewer has less control over the interview
process than when using in-person interviewsthe respondent can
hang up the phone easily.
Greater stress Receiving a strangers call may be very annoying
for many respondents.
Lack of visual material Telephone interviews cannot use visual
aids, such as maps, charts or pictures to explain the survey questions.
4.1.4 Other Methods of Gathering Information

Besides customer surveys, there are several methods that can be used to
gather customer information, such as secondary customer data research,
direct data measurement, and direct product usage involvement.
136 Chapter Four
Secondary customer data research The customer information

may already be available somewhere, such as in libraries, from govern-
ment agencies, or on the Internet, and secondary research is the pro-
cess of retrieving this information. By using data mining techniques
(Berry and Linoff 2000, Edelstein 1999), researchers can extract
important information from these potentially huge data sources and
produce valuable clues that can guide sales and promotion efforts.
Data mining is primarily used by companies with a strong customer
focus, such as retail, financial, communication, and marketing organi-
zations. It enables these companies to determine relationships among
internal factors such as price, product positioning, or staff skills, and
external factors such as economic indicators, competition, and cus-
tomer demographics. Data mining enables companies to determine
the impact of these factors on sales, customer satisfaction, and cor-
porate profitability, and to develop marketing and sales strategies to
enhance corporate performance and cut losses.
Direct data measurement This technique involves directly count-
ing, testing, or measuring data without interviewing customers.
Testing cholesterol levels, monitoring customers arrival time and
time spent in a service institution, and recording and counting the
type and number of errors in insurance claims are typical examples
of direct measurement.
Direct product usage involvement The key idea in this tech-
nique is to ask the product design leaders to play the role of the
consumer and practice using the product. This technique is practiced
by Toyota. One story (Liker 2004) described a Japanese design leader
who had never been to the U.S., but was assigned to design a car for
the North American market. To overcome his lack of knowledge of
the North American market, he landed in the U.S., rented a car, and
drove through all the continental United States and the Canadian
provinces to experience the actual driving and car usage conditions
in North America. As a result of this first-hand experience, he made
some very good changes in the car design.
4.2 Stages of the Customer Survey

Customer survey research is well-established, and a step-by-step proce-
dure is available to guide the whole survey process. These are the stages
of a typical customer survey:
Stage 1: Establish goals and objectives of the survey.

Stage 2: Set the survey schedule and budget.
Stage 3: Establish an information base.
Stage 4: Determine population and sampling frame.

Stage 5: Determine sample size and selection procedure.
Stage 6: Design the survey instrument.
Stage 7: Pretest the survey instrument.
Stage 8: Select and train survey interviewers.
Stage 9: Implement the survey.
Stage 10: Analyze the data and report.
The following sections give a brief overview of each stage.
4.2.1 Stage 1: Establish Goals

and Objectives of the Survey
Determining the goals and objectives of the survey is always the first
stage in a customer survey study. The company that uses the survey infor-
mation has to do this itself, even if the actual survey job will be contracted
out, because only the company itself can define its own needs.
The following aspects are essential in establishing the goals and
objectives of your survey.
1. Determine survey objectives. This is the most critical part in customer

survey design. The survey objectives will determine what kind of ques-
tions will be in the survey. For example, if the objective in the cus-
tomer survey is to gain the customer requirements for a product, as I
described in Chapter 2, then you should design a survey that provides
appropriate questions to extract accurate and sufficient information to
derive product functional requirements and design parameters. If the
objective in the customer survey is to study the key factors in customer
value proposition, as I described in Chapter 3, then you need to design
the questions to extract the customer rating on these factors. The
survey objectives should be spelled out clearly and very specifically.
2. Identify the users of the survey results. The users of the survey
results are the customers of the survey study. For example, in the
product development process, design engineers are often the users
of the survey results. It is very common that the people who do the
survey study and the people who use the survey results are from
different backgrounds, speak different buzzwords, and having differ-
ent mindsets. For example, if the survey study is to be conducted by
the marketing department, but the users are engineers, many times
their interpretations of the same word, such as operation, are quite
different. Many true meanings in the original customer statements
will be lost in translation in this case. In some big corporations,
the VOC information will be transmitted through a hierarchy of dif-
138 Chapter Four
ferent departments, and when it reaches the final user, the VOC
information has become so distorted that it becomes useless. So, it
is necessary to identify all significant survey results users and let
them participate in all stages of customer survey study.
3. Identify what specific information is needed. After identifying the users
of the survey, you need to determine what specific information the
users really need. For example, in the product development process,
you need to consult with design engineers, maybe even manufacturing
engineers, to find out what kind of information they need to determine
the design specifications and product requirements. In this case, good
cooperation between the marketing and engineering departments is
really essential; the marketing department may know the marketing
trends and how to deal with customers, and the engineers know what
information, how much information, and the level of detail in the infor-
mation they need in order to do their design and production job.
Another example is the restaurant business. The professional market-

ing people know how to deal with customers; however, if the customers
dont like the taste of the food, the marketing people will not know how
to fix it. If the marketing people only tell the chef that your food tastes
bad, it wont help very much. So the chef and the marketing person
should work together to figure out ways to get the right information so
the chef can produce the food that customers really want.
4.2.2 Stage 2: Set the Survey Schedule

and Budget
After the goals and objectives of the customer survey are determined,
you need to develop a budget and timeline for this survey project.
If the survey study will be contracted out, then a major cost item in
the budget will include the contract cost. If the survey study will be done
by internal resources, then manpower from several departments will be
needed, and some trade-offs between manpower needs in this project
and other routine jobs may need to be made.
The timeline of the survey project will depend on when you will need
this information. In product development, the availability of the survey
information should fit the product development schedule.
The timetable should be flexible enough to accommodate unforeseen
delays.
4.2.3 Stage 3: Establish an Information Base

Before developing the survey questionnaire, you need to collect detailed
information to populate the survey questions. This should be a joint
effort by the following people:
People who will design and administer the survey

Survey information users
Representatives of customers
A focus group meeting is a popular approach to bring all these people

together in order to reach consensus about the information needs in the
survey. The size of a typical focus group ranges from 5 to 12. In this focus
group meeting, based on the goals and objectives of the survey estab-
lished in Stage 1, an exhaustive list of raw questions will be developed
and they will serve as a basis to develop a survey instrument.
4.2.4 Stage 4: Determine the Population

and Sampling Frame
The population is the entire set of people, organizations, households,
etc., that are addressed by your survey research. For example, for a
fast-food chain, the relevant population will be fast-food eaters. For a
suburban hospital, the population will be nearby residents.
The portion of the population that can be identified for interviews
is called the sampling frame. For example, the population of fast-food
eaters may include all people except homeless and sick people. But if
a telephone interview is to be conducted, only the people with known
telephone numbers can be reached, so people with telephone numbers
will be the sampling frame for fast-food eaters. The concepts of population
and sampling frame will be further discussed in Section 4.4.
4.2.5 Stage 5: Determine Sample Size

and Selection Procedure
The survey researcher will have to select a sample that adequately
represents the population under study. In general, larger samples
will yield greater accuracy than small samples in terms of analysis
results. The sample size is usually determined by balancing the analy-
sis accuracy against the increased cost and time required for surveying
a larger sample size. Once the sample size is determined, the method
of sampling will be chosen. Commonly used sampling methods include
random sampling, stratified random sampling, and cluster sampling.
How sample sizes and sampling methods are chosen will be discussed
in Section 4.5.
4.2.6 Stage 6: Design the Survey

Instrument
The survey instrument, or questionnaire, is a key tool in the customer
survey process. At this stage, the survey researcher must populate the
140 Chapter Four
right questions in order to obtain the information needs in the survey.

It is recommended that the following people should participate again
in the development of survey questions:
People who will design and administer the survey

Survey information users
Representatives of customers
The development of the questionnaire can be a time-consuming

process, requiring great attention to detail. The best-designed
questionnaire is short and concise with well-worded questions. Long
and wordy questionnaires often result in lower response rates and
higher survey costs. The design of the survey instrument is discussed
in Section 4.3.
4.2.7 Stage 7: Pretest the Survey

Instrument
Once a draft questionnaire has been created, it must be pretested
with a small group of respondents. During this pretest, poorly worded
questions can be identified and refined, improving the quality of the
questionnaire.
4.2.8 Stage 8: Select and Train Survey

Interviewers
For telephone and in-person interviews, trained interviewers are
required. Interviewers can be college students, part-time workers, and
so on. Interviewers should be familiar with the questionnaire and know
how to handle uncooperative respondents.
4.2.9 Stage 9: Implement the Survey

The process of administering the survey instrument is a crucial stage of
a customer survey. It is very important to follow the sampling procedure
to ensure the validity of the survey. It is also very important to stick to
the time schedule. Care must also be taken to ensure the privacy of the
respondents and minimize their inconvenience. The implementation of
surveys is further discussed in Section 4.4.
4.2.10 Stage 10: Analyze the Data

and Report
The data from returned questionnaires will be summarized and ana-
lyzed with statistical methods, and the findings will be reported. Section
4.5 discusses the details of data analysis and report.
4.3 Survey Instrument Design

One of the key tasks in surveying customers is developing the question-
naire. The key concerns in questionnaire development are the types of
questions, the wording of questions, the sequence of questions, and the
overall length of the questionnaire.
There are two types of questions that can be used in customer survey:
close-ended questions and open-ended questions. Close-ended questions
allow respondents to select an answer from a list of choices, and open-
ended questions allow respondents to answer using their own words.
4.3.1 Close-Ended Questions

Close-ended questions offer a fixed list of responses and ask respondents
to select one or more options to best answer the question. Close-ended
questions have many different formats.
4.3.1.1 Multiple-Choice Questions Multiple-choice questions have

a list of answers and usually ask for a factual response. In multiple-
choice questions, only one answer is supposed to be applicable. Here is
an example of a multiple-choice question:
What department do you work in?

a. Sales
b. Marketing
c. Manufacturing
d. Research
When you construct this kind of question, you must be careful to

make sure the list of answers is exhaustive and that the answers are
mutually exclusive.
In the preceding example, if someone were working in the personnel
department, he or she would not find an appropriate answer. In such
a case, the question has an inadequate number of answers. However,
the total number of answers in a multiple-choice question should be no
more than 10 or 12.
142 Chapter Four
4.3.1.2 Checklist or Inventory Questions Checklist or inventory ques-

tions ask respondents which items on a list are applicable. Here is an
example of a checklist question:
Please indicate what sources you use to get information about new
music and movies. Check all that apply.
___Radio ___Television ____Internet ______Newspapers
___Friends ___Magazine ____Other (Please specify)
Clearly, in this example, it is possible that more than one answer

may apply.
4.3.1.3 Rating Questions Rating questions ask respondents to use a

given scale to judge something. Here is an example of a rating question:
Please rate your teachers performance in the following categories on a

scale of 1 to 5, where 1 is very poor, 3 is average, and 5 is excellent.
Rating Category
_____ Knowledge of course contents
_____ Instruction
_____ Office hours
4.3.1.4 The Pros and Cons of Close-Ended Questions Close-ended ques-

tions have several advantages:
The set of answers is uniform, so it is easy to compare the differences

among respondents.
The uniformity in the set of answers for each question will make
computer data entry easier.
The fixed list of answers tends to make the question clearer to the
respondent.
Close-ended questions have the following disadvantages:
Close-ended questions will compel respondents to choose the answer

that most closely represents their actual response, which may deviate
from their true opinions.
When respondents are unsure which is the best answer, they may
choose a random answer, which will lead to errors.
4.3.1.5 Level of Measurements in Close-Ended Questions In analyzing

the results of a survey, the survey data will be organized as variables.
A variable is a specific characteristic of the population, such as age,
sex, preference, rating, and so on. Depending on the design of the ques-
tions and answers, the variables used in the survey will have differ-
ent measurement properties, referred to as levels of measurement or
measurement scales. The commonly used measurement scales include
the nominal scale, the ordinal scale, and the interval scale. In survey
design, there is also a specially designed interval scale called the Likert
scale. We will now look at these scales in detail.
Nominal Scale The nominal level of measurement simply divides

survey answers into categories. For example, a variable such as political
party preference in the United States can be categorized using three
classes: Democrat, Republican, or Independent. When using the nominal
scale, survey data is placed into categories and the frequency of occurrences
is counted. There is no ranking or ordering for the categories.
Ordinal Scale The ordinal level of measurement goes one step be-
yond the nominal scaleit ranks the categories by certain criteria. For
example, peoples education levels can be classified using the following
categories: high school graduate or lower, two-year college degree, B.A./
B.S. degree, M.A./M.S. degree, and Ph.D. degree. Clearly, we can rank
these education levels: Ph.D. is higher than M.A., for example. But it
is difficult to define a numerical difference between these educational
achievements.
The Interval Scale The interval level of measurement gives the

greatest amount of information about the variables. It labels and or-
ders them and also uses numerical measurements to indicate the exact
value of each category. For example, variables such as income, age, and
weight use interval scales.
Likert Scale The Likert scale (named after Rensis Likert) is used
to measure attitudes and opinions. A Likert scale may contain several
options such as strongly agree, agree, neutral, disagree, and strongly
disagree. Here is an example:
Firefox is easier to use than Internet Explorer:

1. Strongly disagree
2. Disagree
3. Neutral
4. Agree
5. Strongly agree
Sometimes, a numerical scale is explicitly displayed in the question-
naire:
144 Chapter Four
What is your general impression of how the city government affects your
business?
Highly Negative Highly Positive
1 2 3 4 5
_____ ______ ______ _____ ______
Likert items are ordinally scaled. It is not assumed that the difference
between strongly agree and agree is the same size as the difference
between agree and neutral. However, in survey data analysis, it is
a general practice to treat the Likert scale as an interval scale. For
example, in college course evaluations, there are many Likert scale
questions about a professors teaching abilities, and the scores for each
question from all students are averaged as the evaluation score. Clearly
this treatment assumes that the Likert scale is an interval scale.
4.3.2 Open-Ended Questions

Open-ended questions allow respondents to provide longer, more complex
answers than closed-ended questions do. Here are some examples:
What is your favorite place to go for your summer vacation?__________

How long have you and your family lived in your current location?
_____________
What is the first foreign language you learned? _____________________
There are several disadvantages of open-ended questions:
They will elicit some irrelevant answers.

The number of distinct answers from the respondents may be very
high and answers could be messy and confusing.
They are difficult to analyze with statistical methods, because statisti-
cal methods require some degree of data standardization.
Overall, it is highly recommended that most survey questions be

in the form of close-ended questions to ensure higher response rates,
shorter questionnaire completion times, and easier data analysis.
4.3.3 The Wording of Survey Questions

The wording of survey questions is very important for a successful cus-
tomer survey. Good survey questions should be:
Clear and easily understandable, stated directly and straightforwardly

Specific and precisely stated so that the respondent knows exactly

what is being asked
Unambiguous and unequivocal so there is only one way to understand
or interpret what the question is asking
Simple and brief, rather than complicated, cluttered, and long-winded
Stated in terms that your respondents are likely to be familiar and
comfortable with, without using complex technical terminology,
jargon, or overly sophisticated wording
Good survey questions should not

Lead the respondent toward a specific answer or make some answers
clearly unattractive or undesirable
Ask about two or more things together in the same question
Make the respondent uncomfortable, or put the respondent in a dif-
ficult or compromising position
Example 4.1 Multipurpose Question Here is an example of a multi-

purpose question:
Do you believe the development of the I-696 freeway entrance will affect
the image and property value for our whole subdivision?
Yes_____ No_______
This question might be difficult to answer, because a Yes or No answer
suggests that the image and property value will move in the same direc-
tion. However, it is possible that the image of the community will go up
because of the better freeway access, but the property value will go down
due to increased traffic, noise, and commercial development.
Example 4.2 Ambiguous Question If a survey question about income

is worded as What is your income? it will generate all kind of answers,
such as annual income, hourly pay rate, monthly income, or total house-
hold income. A better wording for the question would be as follows:
What is your total annual income before taxes?

a. Below $20,000
b. Between $20,001 and $40,000
c. Between $40,001 and $60,000
d. Between $60,001 and $80,000
e. Between $80,001 and $100,000
f. Over $100,000
146 Chapter Four
4.3.4 Order of Questions in Surveys

The order in which the questions are presented can affect the overall
customer survey significantly. A poorly organized questionnaire can
confuse respondents, bias their responses, and jeopardize the quality
of the survey.
It is a good idea to start the survey with some easy introductory ques-
tions. Here are some examples:
Do you own a car? Yes___ No____
Do you have an e-mail address? Yes___ No____
How long have you lived at your current address? _________(years)
You should save more complicated questions that may require some
careful thought for later, after you warm up the respondents with intro-
ductory questions.
There are several typical organizational patterns for survey questions:
Chronological order The questions are presented in sequential

or temporal order, for instance, from most recent to least recent.
Funnel pattern The questions are ordered by topic from broad to
specific.
Inverted funnel The questions are ordered by topic from specific
to general.
Tree pattern The questions branch out in different directions,
depending on the respondents answers to earlier questions.
The choice of organizational pattern for the questions should serve the
goals and objectives of the customer survey. Usually, topically related
questions should be grouped together.
Example 4.3 Grouping Survey Questions The following is an exam-

ple of good grouping of questions:
1. How would you describe the current relationship between labor and
management?
Good_____ Fair______ Poor_______
2. During the past five years, do you think this labor-management rela-
tionship has:
Improved____ Remained the same_____ Worsened_____
3. In what way do you think the labor-management relationship can be
improved?
______________________________________________________________
The preceding three questions are related and are presented in a

logical order. The following three questions are confusing when grouped
together:
1. Do you or your coworker participate in a company-sponsored sugges-

tion program?
Yes_____ No______
2. During the last five years, do you think this labor-management rela-
tionship has:
Improved____ Remained the same_____ Worsened_____
3. Would you be interested in a job training program?
Yes_____ No_______
The respondents will be able to answer the questions, but they will get
disoriented after answering 10 or more of these misplaced questions.
4.3.5 Questionnaire Length

The questionnaire should be as concise as possible while covering the
subjects determined by the goals and objectives of the survey. Practice
has shown that when the survey becomes too long, the response rate
and the quality of the answers goes down significantly.
As a general guideline, telephone interviews should be less than
20 minutes, and should ideally take between 10 and 15 minutes. Mail-
out questionnaires should not take more than 30 minutes to answer, and
ideally take about 15 minutes. In-person interviews should be limited
to less than an hour, and ideally to less than 30 minutes.
4.4 Administering the Survey

Once the survey instrument is designed, pretested, and revised, it is
time to administer it. For different survey methodsmail-out survey,
telephone survey, and in-person surveythe method of administering
the survey will be different. In this section, we will look at how these
different types of surveys are administered.
4.4.1 Administering Mail-Out Surveys

In mail-out surveys, the questionnaire should be designed in the form of
a booklet in order to ensure a professional appearance. Any resemblance
to an advertisement brochure should be strictly avoided. The profes-
sional appearance of the questionnaire is very important for ensuring
a satisfactory response rate. There should be adequate spacing between
questions, and the questionnaire should be designed so that it is conve-
148 Chapter Four
nient for the respondent to mail it back. A good cover letter explaining
the purpose of the survey is very important.
There are two ways to present questionnaires to respondents: personal
delivery and direct mailing. Delivering the questionnaires directly to
respondents is more costly in terms of time and effort, but it is likely to
result in a higher response rate, more rapid responses, a higher percentage
of completed questions in the questionnaire, and perhaps more valid and
accurate responses. Direct mailing will usually result in a lower response
rate, although this can be remedied somewhat by follow-up mailings or
follow-up phone calls. Usually these follow-ups should be done three to
four weeks after the questionnaires are sent by mail. Direct mailing plus
follow-up usually will achieve a 5060 percent response rate. Additional
follow-ups may raise the response rate to over 70 percent.
4.4.2 Administering Telephone Surveys

The telephone survey is less complex to implement than the mail-out
survey. The most important aspect of the telephone survey is the selection
and training of telephone interviewers. A good source of possible inter-
viewers is university students, especially graduate students. Training
usually involves the interviewers learning the questionnaire by them-
selves, learning about pretest results and potential tough issues in the
questionnaire, and learning about general ethical issues. Interviewers
should not introduce any bias in the interview process, and should not
express any opinions in response to the answers from the respondents.
Many companies that conduct telephone interviews use computer-
assisted telephone systems where the interviewers sit at a computer
that dials the telephone and displays the question to be asked so the
interviewer can read it to the respondent. The software that manages
this can take care of the data entry and coding, as the interviewer uses
the computers keyboard or mouse to indicate the respondents answer
to the questions.
4.4.3 Administering In-Person Surveys

In-person interviews are the most expensive type of survey, in terms of
both time and money, and the most intrusive of all the methods. However,
a major strength of the in-person interview is that you can deal with
complex topics. Because you can see how the respondent reacts to the
questions as you ask them, you will have a better idea of how well they
understand the question and what confuses them. You will also have the
opportunity to resolve any glitches in the interview process.
Taking care in selecting and training in-person interviewers is even
more important than for telephone interviewers. The methods for select-
ing and training interviewers are almost the same as for telephone inter-
viewers; the additional abilities that an in-person interviewer should

have include politeness; nice smile; good eye contact; the ability to observe
customers subtle gestures, body language and language tones, and make
sense out of them; and skills to alleviate customers stress.
4.5 Survey Sampling Method and Sample Size

The main goal of a customer survey is to produce an accurate picture of
the population based on information drawn from a scientifically selected
subset of that population. Sampling is necessary because it is impracti-
cal to seek information from every member of the population.
In this sampling process, we first need to determine what our popula-
tion is. Then we need to define a sampling frame, which is a list of ele-
ments in the population that may be selected in the sample. Elements
in the population could be people, business entities, organizations, and
so on. Third, we need to identify a sampling methoda method for
selecting a subset from the sampling frame.
Before we do the sampling, we need to identify an adequate sample
size for the survey to ensure the credibility of the data analysis. In this
section, we will look at these issues.
4.5.1 Population and Sampling Frame

The first consideration in survey sampling is the unit of analysis: the
individual, object, institution, or group that is relevant to the survey. For
example, for a fast-food chain owner, the individual consumer could be
the unit of analysis; for a mortgage lending organization, each house-
hold could be the unit of analysis; for a medical equipment supplier, each
hospital or clinic could be the unit of analysis.
The population is defined as the collection of units of analysis that
the findings of the survey will apply to. For example, the population of
fast-food chain customers is the collection of all potential individual
customers that the chain can reach; the population of customers for a
mortgage lending operation is the collection of all households that the
lending operation could do business with; the population of customers
for a medical equipment supplier will be the collection of all potential
hospitals and clinics that could do business with this supplier.
Usually, in any population, not all of the units of analysis can be iden-
tified and reached. For example, if a population is all the people living
within a metropolitan area, the unit of analysis will be each resident.
From a practical point of view, it is unlikely that all the residents of
this metropolitan area can be identified and reached. People are born
and die, people move in and out. Some people do not have telephones or
stable living places. Usually only a portion of a population is identifi-
150 Chapter Four
able and reachable, and this portion of the population is often called the
working population.
From the working population, it is possible to develop a list of units
of analysis that can be readily reached in a customer survey. This list
is called the sampling frame. For example, if the population is all the
residents in a metropolitan area, and the working population is all res-
idents that can be reached by phone, the sampling frame could be the
residents listed in the local telephone directory. Other possible sources
of sampling frames include the voters list, utility (gas, electric, water)
customer lists, motor vehicle registrants, and magazine or newspaper
subscriber lists.
With most sampling frames, you will have to deal with some of these
problems:
Missing elements Some legitimate members of the population will
not be included in the sampling frame. For example, in some polls of
U.S. elections in 2004, only traditional phone users were polled, and
people who only had cell phones were not included in the poll list.
Therefore, a sizeable proportion of young professionals was left out.
Foreign elements Some peoples names will be listed in the sample
frame, but they are actually no longer in the population. For example,
people could move away but their names would still be in the phone
directory.
Duplicate elements Population members may be listed more than
once in the sample frame.
In all these situations you need to determine how many missing,

foreign, and duplicate elements are in the sampling frame, and how
big a proportion these elements are as a percentage of the whole sam-
pling frame. If this proportion is large and it will affect the accuracy
of the poll, you should consider using a different sampling frame. For
example, in some opinion polls for the U.S. 2004 election, people with
only cell phones were excluded in the opinion poll sampling frame. If
these people were a sizeable proportion of voters and their opinions were
significantly different than those of traditional phone users, the opinion
poll might be unreliable.
4.5.2 Sampling Methods

Sampling methods can be classified into probability sampling and non-
probability sampling. Probability sampling is used when you would
like to draw conclusions for the whole population based on the data
you collect in the sample. If your goal is just to learn something about
the sample, and you do not intend to draw conclusions for the whole
population, you can use nonprobability sampling.
There are two characteristics of probability sampling:
The probability of selection is equal for all elements of the sampling

frame at all stages of the sampling process.
The selection of one element from the sampling frame is independent
of selection of any other element.
For example, consider a sampling frame of 1000 people whose names

are written on equal-sized pieces of papers. The pieces of paper are
thoroughly mixed and selected one by one without looking at them. If
1000 people are selected in this sample, the probability of selecting any
person in the first draw is 1/1000; the probability in the second draw is
1/999; and so on. Finally, the chance of selecting any person in the hun-
dredth draw is 1/901. Though the probability of selecting a particular
person is slightly different in each draw, the probability of selection for
all the available people within each draw is the same. This is in keep-
ing with the first rule of probability samplingthat the probability of
selection is equal for all elements of the sampling frame at all stages
of the sampling process. Also, the probability of selecting a particular
person in any drawing is clearly independent of previous drawings, so
this sampling practice is an example of probability sampling.
There are several methods of probability sampling: random sampling,
systematic sampling, stratified random sampling, and cluster sampling.
The following sections will discuss these probability sampling methods
as well as nonprobability sampling in detail.
4.5.2.1 Random Sampling The best known probability sampling

method is random sampling. In random sampling, each unit in the sam-
pling frame is assigned a distinct number, and then units are chosen at
random by a process that does not favor certain numbers or patterns of
numbers. The chosen units become the sample.
A commonly used method to randomly choose units from a sampling
frame is to use a table of random numbers, as shown in Table 4.1.
Suppose that there are 1000 people in the sampling frame, and we want
to select a random sample of 30 people. Each person will be assigned
TABLE 4.1 A Portion of a Table of Random Numbers

77073 51849 15761 85622 38905 72276
20837 95047 50724 16922 04405 30858
37504 15645 36630 28216 10056 97628
40392 58557 60446 11553 60013 38037
53408 14205 33152 70651 17314 93033
152 Chapter Four
a number ranging from 000 to 999. Using Table 4.1, we can choose to
use any three digits from the five-digit numbers given. For example, we
could choose to use the last three digits, so we would select the people
with number 073, 849, 761, 622, 905, 276, 837, , 033.
For large samples, using a random number table is tedious and time-
consuming, so computer-generated random numbers can be used to
select a random sample.
4.5.2.2 Systematic Sampling Systematic sampling is an adaptation of

the random sampling method that can be used if the sampling frame is al-
ready randomly distributed. It is useful when the sampling frame is quite
large and the sampling units cannot be easily numbered. For example, if
a sampling frame has 3 million people and a sample of 1,500 people is
required, a random sampling approach might be unrealisticnumbering
3 million people is already a big task.
If the original list of these 3 million is randomly distributed, we can
select sample units by selecting them from the list at fixed intervals
(every nth entry; for example, every 20th car on the highway, every 50th
customer in a store). If we are selecting 1,500 people from 3 million, we can
select 1 out of every 2,000 people in the sampling frame (3,000,000/1,500 =
2,000). If we start with a random starting point, and we select every
2,000th person, we will create a random sample of 1,500 people.
4.5.2.3 Stratified Random Sampling Stratified sampling assumes that

the sampling frame consists of several mutually exclusive groups, called
strata. In stratified random sampling, the total number of samples are
divided among strata by a predetermined proportion. Then, random
samples are taken from each stratum.
For example, in a community, assume that 60% of the population is
white, 15% is black, 15% is Hispanic, and 10% is Asian. If a sample of
1000 people is needed, the stratified sampling method will divide these
1000 people into four ethnic groups based on their proportion in the
population. So 600 samples will be allocated to whites, 150 samples
will be allocated to blacks, 150 samples will be allocated to Hispanics,
and 100 samples will be allocated to Asians. Then the 600 people in the
White strata will be randomly selected from the White sampling frame;
150 people in the Black strata will be randomly selected from the Black
sampling frame, and so on.
4.5.2.4 Cluster Sampling Cluster sampling is used when there is a

hierarchy of sampling units. The primary sampling unit is a group (or
cluster), such as counties, cities, schools, subdivisions, and so on. The
secondary sampling units are the individual elements within these
clusters from which the information is to be collected.
For example, suppose we want to study the needs of first and second
graders. It is difficult to directly locate the sampling frame from a raw
population list, such as a telephone directory, but it is easy to identify
the clusters, such as public and private schools, in which there are first-
grade and second-grade classrooms. After we select a subset of class-
rooms, we can randomly select sample units from these classrooms.
4.5.2.5 Nonprobability Sampling In nonprobability sampling, the prob-

ability that a particular unit will be selected is unknown. In this case,
we cannot generalize the findings from within the sample to the overall
population. We cant assume any valid statistical relationship between
the sample and the population, so we cant use such useful probability
distribution models as normal distribution.
However, nonprobability sampling can still be helpful in getting a
feel for what a portion of customers may think, and selecting a sample
is much easier. For example, nonprobability sampling can be used to
quickly select a small sample of respondents (say 30 people) to pretest
a survey instrument. The conclusions from these 30 people cannot be
generalized to the general population, but a lot of shortcomings in the
survey instrument can be identified.
There are a couple of commonly used nonprobability sampling meth-
ods. The most common is the sidewalk survey. For example, for a gen-
eral population of shoppers, an interviewer might interview passersby
near a shopping center. In this approach, the sampling frame is not
explicitly identified and numbered, and the probability of selecting any
particular passerby is unknown. The advantage of this approach is the
ability to get a lot of information quickly.
The other commonly used nonprobability sampling technique is snow-
ball sampling. This method is particularly beneficial when it is difficult
to identify potential respondents. Once a few respondents are identified
and interviewed, they are asked to identify others who might qualify as
respondents. The list of respondents will be increased quickly.
4.5.3 Sample Size Determination

One critical question in a survey project is how many units are needed
in a sample for the results to be generalized to the whole population.
The answer to this question depends on two key factors:
Accuracy The greater the level of accuracy required in the study,

the larger the sample size must be.
Money and time Larger sample sizes will certainly mean higher
costs and more time will be required to complete the survey.
154 Chapter Four
Therefore, the sample size is mostly determined by the trade-off

between the desired level of accuracy on the one hand, and the budget
and time available on the other.
4.5.3.1 Determining Sample Size for Variables Expressed in Proportions In

survey data analysis, many variables are expressed in terms of propor-
tions. For example, we might ask customers a question like this:
Do you like the service of ABC Bank? Yes____ No_____
The proportion of people in the survey sample answering yes, which
is often called the sample proportion, p , is often used as the statistical
estimate of the population proportion, p, where p is the real proportion
of customers who like ABC Banks service. Of course, we would like p to
be as close to p as possible. From the properties of normal distribution,
and if a random sampling method is used, the probability distribution
of p is:
p(1 p) (4-1)
p ~ N p,
n
The 100(1 )% confidence interval for p is:
p(1 p) p (1 p ) (4-2)
p Z p Z
2
n 2
n
p(1 p) p (1 p )
We can use p = Z Z to represent the half width
2
n 2
n
of the confidence interval for p. The magnitude of p represents the
accuracy of p as an estimator of p, because:
P ( p p p p + p ) = (1 )100% (4-3)
p is also called the margin of error.
Example 4.4 Calculating Sample Size with the Margin of Error = 3% In

a customer satisfaction survey, if the preliminary results indicate that
the proportion of unsatisfied customers is very close to the proportion
of satisfied customers, what sample size is needed if we want the ac-
curacy of the survey to be within 3% of the true proportion, with 95%
confidence?
In this case, clearly p 50%, from the statement of the problem. We
want
p(1 p)
p = Z = 3%
2
n
TABLE 4.2 Minimum Sample Sizes for Proportions
Confidence Interval Sample Size

(Margin of Error %) 95% Confidence 99% Confidence
1 9,604 16,590
2 2,401 4,148
3 1,067 1,844
4 601 1,037
5 385 664
6 267 461
7 196 339
8 151 260
9 119 205
10 97 166
Therefore,
2
Z p(1 p)
n= 2 (4-4)
p

is the sample size formula for this case. Specifically,

2
1.96 0.5(1 0.5) = 1067
n=
0.03
for this example, where Z0.025 = 1.96.

So a sample of 1067 or more people is needed to ensure the accuracy
of 3%.
Table 4.2 lists the relationship between sample size, margin of error,
and confidence level.
4.5.3.2 Determining Sample Size for Variables Expressed in Proportions

for Small Populations The sample size rule specified by equation (4-4)
is based on the assumption that the population size is infinite or very
large. In some survey studies, however, the population size is rather
limited. For example, the customer base for a medical equipment sup-
plier will consist of a number of hospitals and clinicsthe population
size will be in the hundreds at most. If the population size, N, is known,
then according to Rea and Parker (1992), the sample size n can be cal-
culated as follows:
Z2 [ p(1 p)]N
n= 2 (4-5)
Z2 [ p(1 p)] + ( N 1) 2p
2
156 Chapter Four
Example 4.5. Calculating Sample Size for Proportional Variables and

a Small Population In a customer satisfaction survey, the preliminary
results indicate that the proportion of unsatisfied customers is very
close to the proportion of satisfied customers, and the population size
is N = 2500. What sample size is needed if we want the accuracy of the
survey to be within 3% of the true proportion, with 95% confidence?
By using equation (4-5),
Z2 [ p(1 p)]N
= 1.96 2 (0.5 0.5)(2500)
n= 2
= 749
Z2 [ p(1 p)] + ( N 1) 2p 1.96 (0.5 0.5) + 2499 (0.03)2
2
This sample size is smaller than the one calculated in Example 4.4.
4.5.3.3 Determining Sample Size for Interval Scale Variables In survey

analysis, some variables are interval scale variables. For example, per-
sonal income, age, and evaluation scores based on the Likert scale are
all interval scale variables. The population means of these interval scale
variables, , are usually our interest. The sample mean of the interval
scale variable, x , is often used as the statistical estimate of population
mean, . We would like x to be as close to as possible.
From the properties of normal distribution, and if a random sampling
method is used, the probability distribution of x is
2
x ~ N , (4-6)
n
The 100(1 )% confidence interval for is

x Z
= x (4-7)
2 n
where is the margin of error for .
By using the relationship

= Z (4-8)
2 n
we can derive the sample size rule:
Z2 2
n= (4-9)
2
Example 4.6 Calculating Sample Size for Interval Scale Variables In

a survey study of household incomes for County Y, the preliminary esti-
mate of average household income is $40,000, and the standard deviation
is estimated to be $6,000. If we would like to determine a survey sample
size so that the margin of error for the average household income is no
more than $1,000, what is the minimum sample size, if a confidence level
of 95% is desired?
By using Equation (4-9):
Z2 2 1.962 60002
n= = = 139
2 10002
Therefore, a minimum sample size of 139 households is required.
4.5.3.4 Determining Sample Size for Interval Scale Variables for Small
Populations The sample size rule specified by equation (4-9) is based on
the assumption that the population size is infinite or very large. Sometimes,
however, the population size is rather limited. If the population size, N, is
known, then according to Rea and Parker (1992), the sample size n can
be calculated as follows:
Z2 2
n= 2
2 + Z2 2 ( N 1) (4-10)
2
Example 4.7 Calculating Sample Size for Interval Scale Variables

and a Small Population In a survey study of household incomes for
County Y, the preliminary estimate of average household income is
$40,000, and the standard deviation is estimated to be $6,000. If we
would like to determine a survey sample size so that the margin of er-
ror for the average household income is no more than $1,000, and it is
known that the total number of households in County Y is 5,000, what
is the minimum sample size, if a confidence level of 95% is desired?
By using Equation (4-10),
Z2 2
n= 2 1.962 60002
2 + Z2 2 ( N 1) = 1.962 60002 = 135
2 10002 +
4999
4.6 Internet Surveys

Since the late 1980s, the Internet has emerged as a popular medium
for spreading, transmitting, and collecting information. The Internet
reaches more and more people every day, and in the not-too-distant
future it will reach an overwhelming proportion of the world population.
E-mail has gradually become a preferred form of communication. The
158 Chapter Four
Internet and e-mail are more convenient, more interactive, more flex-
ible, cheaper, and faster than paper-based information media.
The emergence of the Internet has also affected the way that busi-
nesses conduct customer surveys. In the late 1980s and early 1990s,
e-mail surveys were becoming popular, because e-mail can instanta-
neously transmit many surveys at little or no cost. However, at that
time, e-mail surveys could only be text-based, and the user interface
was poor and prone to response errors. In the mid-1990s, the web-based
survey was gaining the upper hand because its user interface was much
better and it could be highly interactive. After the mid-1990s, e-mail-
ing respondents with a hyperlink to a web survey became one of the
preferred forms for surveys.
In this section, we will discuss all major aspects of Internet-based
surveys, and compare Internet-based surveys with traditional forms
of surveys.
4.6.1 Drawing People to the

Internet-Based Survey
The first step in an Internet survey is to draw a sample of individuals
to participate in the survey. Specifically, a group of individuals must be
identified, selected, and contacted for the survey. This group should fit
the goals and objectives of this Internet survey. For example, if Amazon.
com would like to know what its customers think about the quality of
Amazons online purchasing service, this group should consist of current
Amazon.com customers.
The first stage of an Internet survey should involve the following
steps:
1. Specify the target population.

2. Develop a sampling frame.
3. Choose a sampling method.
4. Determine the size of the sample.
5. Contact the selected sample group.
Lets look at these steps in detail.
4.6.1.1 Specify the Target Population As in all surveys, the target

population is the people relevant to the purpose of the survey. For ex-
ample, suppose Amazon.com would like to know what its customers
think about Amazons online shopping process in order to improve its
processes and make customers happier. In this case, the target popu-
lation should be all Amazon.com online customers. If Amazon wants
to improve its online shopping process to attract more customers, the

target population should be all Internet customers, which would in-
clude not only all of Amazons current customers, but also the online
shoppers of other companies, such as eBays customers, and potential
future online shoppers.
4.6.1.2 Develop a Sampling Frame The sampling frame is the subset of

people in the target population who are readily reachable by the sur-
vey researcher. For example, for the target population of all Amazon
.com online customers, the sampling frame could be all registered
Amazon.com online customers. In this case, survey researchers could
send online surveys by e-mail to all registered customers because they
are readily reachable. However, if the target population is all online
customers, the relevant sampling frame will be much harder to come
by, because there is no single database for all online customers. The
best that survey researchers can do might be to acquire customer lists
from major online shops.
Usually, a sampling frame can be developed in one or both of the fol-
lowing ways:
E-mail list Surveying people by e-mail is very similar to telephone

surveys or mail-based surveys. You send e-mails to the people in the
e-mail list and hope to get responses. However, getting the right e-mail
list to start with may not be an easy task. Many businesses have their
own customer e-mail lists, but these may not provide an adequate sam-
pling frame, because in many survey studies the researchers want to
reach competitors customers and other potential customers. In such
cases, the survey researchers need to acquire e-mail lists from the rel-
evant e-mail list holders, for example, from other companies, or other
sources. Another difficulty in using e-mail lists is that e-mail addresses
in the lists may not be accurate and are subject to change.
Web users Survey researchers can include links to the survey in
relevant websites. While e-mail is the virtual equivalent of postal
mail, communicating with customers through the web is comparable
to advertising through billboards. Considering that there are billions
of web pages, the number of people attracted by a particular web page
is often quite low. However, putting advertisements in popular search
engines such as google.com, or on highly popular websites, could attract
a sizeable number of web users.
4.6.1.3 Choose a Sampling Method After the sampling frame is deter-

mined, the members of the sampling frame will be selected and con-
tacted. As with telephone and mail-based surveys, there are two basic
160 Chapter Four
approaches to the sampling method: probabilistic and nonprobabilistic.

These two sampling methods were discussed in the Sampling Methods
section earlier in this chapter.
The sampling method chosen depends on the purpose of the survey.
If the survey is designed to make inferences to or predictions about the
target population, a probabilistic sampling method is required. However,
if the purpose is only to describe the population, a nonprobabilistic sam-
pling method can be used.
4.6.1.4 Determine the Size of the Sample After establishing the sam-
pling frame, researchers must determine the number of people within
it to solicit for the survey study. The size of the sample will determine
how accurate the survey study will be. Obtaining sufficient sample
size is not usually a problem in Internet surveysthe marginal cost of
contacting an additional individual is very low, and the transmissions
can be done instantaneously, so adding respondents does not generally
increase the cost and time required for the study.
However, there are two common problems with Internet surveys. One
is the extent of invalid contact information; a sizeable proportion of e-
mail addresses might be wrong, or outdated. The other problem is the
response ratepeople usually get a lot of e-mails every day, and e-mail
solicitations for surveys may get ignored or misclassified as junk mail
and may never reach the intended customers.
4.6.1.5 Contact the Selected Sample Group The next step is to contact
the members of the sampling frame and solicit their participation. In
current Internet survey practice, two techniques are used most com-
monly: e-mail-based solicitation and website-based solicitation.
E-mail-Based Solicitation Survey participants can be contacted by

e-mail. The sampling frame may be an e-mail list from a particular
organization, or e-mail lists from several organizations. The samples
obtained from these kinds of e-mail lists are usually probabilistic sam-
ples. Alternatively, the survey participants may be solicited from public
mailing lists, or from paid mailing lists, such as http://lists.indymedia.
org or http://www.infousa.com. In this case, because members of the
target population are usually not equally likely to be included in the
lists, samples from these lists may not be probabilistic.
After securing a list of e-mail addresses, the survey recruitment e-
mails can be sent to the e-mail addresses. However, survey research-
ers should be mindful of the threat posed by spam. Spam is the name
given to unsolicited bulk commercial e-mail, and most e-mail systems
have spam filters that redirect those bulk mailings to a junk mail folder.
Researchers must construct messages that will not only elude spam
filters but can also overcome the readers suspicionmany users ignore
or delete such messages either because they look like spam or because
they fear computer viruses. A very appealing e-mail heading is essential
to overcome these fears.
An e-mail heading is the virtual equivalent of the address information
on a postal envelope. It generally contains six text fields:
From The e-mail address of the sender

To The e-mail address of the recipient
Subject The topic of the e-mail
CC The e-mail addresses of users receiving a disclosed carbon copy
of the e-mail
BCC The e-mail addresses of users receiving a blind carbon copy
of the e-mail
Attachments Any files to be added to the e-mai.
Survey researchers must be careful about what information is entered

on these lines, because e-mail users frequently decide whether or not to
open a message based on these fields.
The From field should contain the e-mail address of someone famil-
iar to the recipient if at all possible, because spam software often filters
out unknown senders. The To field should contain a single recipient.
Though most e-mail software permits multiple names to be placed in
the To field, spam software frequently screens for bulk e-mails, par-
ticularly ones containing unfamiliar addresses.
The Subject field should contain a brief, precise phrase or sentence
inviting the user to participate in a research study. Survey researchers
should avoid words commonly used to market products, such as free,
money, or offer, because spam software often filters on such text. As a
result, incentives are best not mentioned in the subject field. The focus of
the subject line should be on legitimizing the e-mail, either by referenc-
ing the survey researchers home institution or the subject of study.
The cc and bcc fields should be left blank, because spam software
will often filter out bulk e-mails, especially with unfamiliar e-mail
addresses in the cc or bcc fields. Also, e-mail receivers are usually
very skeptical about e-mails with large cc listslong cc lists are usu-
ally a sign of junk mail. The attachment field should also be left
blank. Some people will not open e-mails containing attachments for
fear of being infected with a computer virus.
The body of the message in the e-mail invitation should explain the
objectives, procedures, and expectations of the survey, as well as how
the individuals name and e-mail address were obtained. Spam soft-
162 Chapter Four
ware also screens the body of the message for keywords, hyperlinks, or
images, so survey researchers should avoid using words and phrases
commonly found in product ads.
Website-Based Solicitation The web is also a platform for recruiting

large, diverse, nonprobabilistic samples of the general population because
hundreds of millions of people use the web every day. In website-based
solicitations, advertisements should be posted on popular web pages.
Interested viewers can simply click through the ad and be immediately
directed to the web survey site, where they can be formally recruited.
Survey researchers can also list their website with a search engine, such
as google.com. Search engines direct viewers to these websites based on
keywords the viewer enters in the search engine. The survey websites
can also be listed in the advertisement section of search engine pages.
The content of the advertisement is very important in attracting view-
ers to the survey website. Menon and Khan (2002) found that customers
are more attracted to colorful, image-laden websites than to monotone,
simple websites.
4.6.2 Administering a Survey on the Internet

After recruiting sufficient survey participants, the second stage of the
Internet survey is to administer the survey. Survey researchers need
to construct the survey instrument and distribute it to all participants.
To do this, the researchers need to
1. Select a communication mode

2. Control access
3. Format the survey instrument
4. Help participants finish the survey
In the following sections we will look at these steps in detail.
4.6.2.1 Selecting a Communication Mode Survey administration starts

with selecting a communication mode that transmits the survey instru-
ment to survey participants. Currently, there are three popular meth-
ods: e-mailing the survey instrument directly to survey participants;
posting the survey instrument on the web; and e-mailing participants
with a message and link to the survey website.
E-mailing the Survey Instrument Directly to Survey Participants The

advantage of directly e-mailing the survey is its convenience and effi-
ciency. You can send survey instruments to all participants in no time.
There are two ways to transmit the survey instrument through e-mail.
One is to embed the survey instrument in the e-mail message. However,
some e-mail programs cannot display anything but text e-mails, or for
security reasons are set to only display text, and completing surveys
in text format is not a very convenient process. The other option is to
send the survey instrument as an e-mail attachment. This approach has
several disadvantages as well: Some people are afraid of opening e-mail
attachments for fear of computer viruses, and participants may not have
software that can open the file type of the survey instrument.
Posting the Survey Instrument on the Web Posting the survey in-
strument on a website has quite a few advantages. Websites have a
lot of design flexibilitythey can integrate multimedia and interactive
elements in visually appealing formats. However, web posting is not the
right tool for contacting many participants directly and quickly.
E-mailing Participants with a Message and the Survey Website

Hyperlink This option is a hybrid of the previous two approaches. In
this approach, e-mails are sent to participants, and a hyperlink to the
web page hosting the survey instrument is provided in the body of the e-
mail message. This approach can contact a large number of participants
quickly, and the hyperlink will conveniently guide the participants to
the survey website where a user-friendly and highly interactive survey
instrument can be hosted. However, this approach might require too
much computer knowledge for some people, such as senior citizens who
may have less experience with computers and the Internet.
4.6.2.2 Controlling Access Once the survey researchers have selected

a transmission mode, they have to develop and implement procedures
to control access to the survey instrument. The purpose of controlling
access is twofold: You need to make sure that a participant can only fill
out one survey instrument, and you need to make sure that the right
participants get the right survey instruments. (The second reason is
particularly important if the researchers are conducting more than one
survey, or if the website hosts multiple survey instruments.)
For e-mail solicitation, controlling access is relatively easy; the survey
researcher will send one e-mail for each participant. If the e-mail con-
tains a hyperlink, the researcher can simply include the correct hyper-
link for the appropriate survey instrument. Usually, researchers can
also assign each participant a password, and the password can be set
in such a way that it only works for the right type (or version) of survey
instrument, and so that it only works once for that survey instrument.
For web-based solicitation, the website can also be designed in such a
way that each user will be assigned one password.
164 Chapter Four
4.6.2.3 Formatting the Survey Instrument The next stage of the admin-
istration process involves formatting the survey instrument for pre-
sentation. The formatting of the survey instrument has two important
aspects: questionnaire design and style design.
Questionnaire Design The following list describes some best prac-

tices in Internet survey questionnaire design.
List only a few questions per screen The survey designer should
present only one, or very few, questions per screen so that participants
do not have to scroll down to answer the questions. Excessive scrolling
may stress participants, causing them to abandon the survey.
Eliminate unnecessary questions The survey designer should
avoid unnecessary questions, such as the current date (the computer
can provide this), participants name (which may already be available
from another source), and so on.
Limit the length of the survey to 15 minutes Research (MacElroy
2000) indicates that the dropout rate (the proportion of participants
who abandon the survey) is proportional to the survey length. The
dropout rate is only 9% if the survey can be finished in 10 minutes;
however, the dropout rate jumps to 24% when the survey takes 15 min-
utes to finish, and 35% when the survey takes 20 minutes to finish.
Avoid content that slows down computers Some computer
graphics and fancy animation might take a long time to download
and thus may slow down the computer. This will add to frustration
for the participants.
Use very clear language; avoid ambiguous questions and
answers As in any survey, the questions should be worded very
clearly in such a way that different readers will interpret each ques-
tion identically.
Provide some indication of survey progress With a mail survey,
participants can easily flip through the pages to see how much of the
survey has been completed so far. Without an indicator showing how
much of the survey has been completed, participants will feel that the
survey is endless, leading to stress. Having a progress indicator will
help participants to budget their time.
Allow participants to interrupt and then re-enter the
survey Survey participants should be offered an option of leaving
the survey, saving the survey results, and going back to the same
survey at the same point, especially for long surveys.
Style Design The style design of Internet survey instruments in-

volves the following aspects:

Display The survey researchers must decide how to visually display
the survey instrument. There are three important factors: monitor
resolution, viewable screen area, and the size of the text.
Monitor resolution This is the number of pixels visible on the
screen, with a pixel being a single point projected onto a monitor. For
example, an 800 600 pixel screen can display 800 pixels on each
of 600 lines. The popular monitor configurations are 640 480, 800
600, 1024 768, and 1280 1024 pixels. Creating a large survey
instrument (as measured in pixels) will make the page wider than
some computer screens, forcing participants to scroll horizontally to
read it.
Screen area Different types of desktop and laptop computers
have different screen sizes. Currently, most monitor sizes range
between 14 and 21 inches measured along the main diagonal. A 14-
inch monitor has a 12-inch horizontal screen, and a 17-inch moni-
tor has a 14-inch horizontal screen. Because survey instruments
could be viewed on all kinds of computers, it is a common practice
to put survey instruments into a 10-inch horizontal space in order
to minimize scrolling.
Text size The most popular font sizes in Internet surveys are 12
or 14 points.
Color One of the attractive features of the Internet is the ability
to apply color to the survey instrument. Color can be used to liven
up text, provide contrast, and make images more eye-catching. In
Internet surveys, it is desirable to have a high contrast between the
background color and the text. Black on white, blue on yellow, and
red on green are commonly used as color pairs for the text and back-
ground. Also, research has indicated that dark text on a light back-
ground is easier on eyes than light text on a dark background.
Graphics, multimedia stimuli Internet surveys can add a lot of
features that are not possible in paper-based surveys. Paper-based sur-
veys usually are printed in black and white in order to reduce costs, and
explanations and special help are minimized to save space. Internet
surveys dont have those limitations. They can include the following:
Hyperlinks to other documents can be used to provide help screens
with detailed definitions and explanations.
Graphics and animations can be added to illustrate the survey
contents.
Videos, voice, and sounds can be added to clarify the contents and
stimulate the participants so the survey is not boring.
166 Chapter Four
However, these multimedia stimuli should be used sparingly because

they may slow down the survey process, requiring files to be down-
loaded. Participants also may not have the software required to view
these advanced features.
4.6.2.4 Helping Participants to Finish the Surveys Survey researchers

cant just send the survey instruments to participants and hope every-
thing will be fine. Researchers should offer help and incentives to par-
ticipants in order to get a higher response rate and ensure the quality
of the survey.
Providing Instruction Survey researchers should give clear instruc-

tions to participants on how to access, complete, and submit the survey
instruments. Survey researchers should not assume that any task is
too simple or obvious. Instructions should be designed to inform all
targeted participants, regardless of their education, technical skills, or
online experience, on how to perform each task in the survey.
The instructions should explain how to get access to the survey
instrument, including the hyperlink to the website, how to get a password,
and so on. The instructions should also include a good introductory
statement that describes the objective of the survey and the major
contents in the survey. The instructions should also tell participants how
to enter their responses, and how to submit the survey when finished.
Ensuring Response Rate The response rate is the proportion of sur-

vey participants who complete and submit the survey to those who
were contacted by the survey researcher. Survey researchers can use
these techniques to improve on the response rate:
Personalized contact Personalized contact emphasizes how

important individual participants are to researchers and stimulates
participants to complete the survey. It is helpful to inform the partici-
pants how researchers acquired their name and contact information,
and to emphasize the importance of the survey and how much the
participants are appreciated.
Financial incentives Another option for inducing participation is
to offer a financial incentive either when participants agree to par-
ticipate or once they have completed the instrument.
Sending reminders Sending e-mail reminders to participants once
the administration of the survey instrument is under way can help
remind participants about the survey.
4.6.3 Comparing Paper-Based Surveys with

Internet Surveys
We have now seen how to conduct Internet surveys. For survey research-
ers, though, the real question is which type of survey is better, the paper-
based survey or the Internet survey? And which type of survey should
be used? To answer these questions, we need to compare the following
aspects of paper-based surveys with Internet surveys:
Response rate
Cost
Lead time
Data quality
4.6.3.1 Response Rate Researchers (Schonlau et al. 2002) have shown

that the response rate of e-mail surveys ranges from 6% to 68%, and
the response rate for web-based surveys ranges from 7% to 44%. These
response rates are somewhat lower than those for mail surveys (21%
to 78%). The response rate of Internet surveys is definitely lower than
that of telephone surveys.
4.6.3.2 Cost The cost of Internet surveys consists of the following com-
ponents: the survey instrument design and survey website development
(this is a one-time charge), the cost of contacting participants, and the
cost of compiling the survey results. On the other hand, mail-based sur-
veys have the following cost component: the survey instrument design
(this is a one-time cost), the cost for printing and mailing surveys, and the
cost of data entry for survey responses and compiling survey results.
Based on the study of Schonlau (2002), for Internet surveys the website
design incurs significant cost, but the cost of contacting participants is
much cheaper for Internet surveys than for mail surveys. Internet surveys
also do not require data entry when compiling survey results, because the
participants have already entered their results in electronic form.
Overall, for a small-scale survey (a survey of a few hundred up to a
thousand), the mail-based survey is cheaper, because of the high cost of
website development. For larger-scale surveys (say 1,000 participants
or more), the Internet survey will be cheaper, because website design is
a one-time charge, and the cost of contacting additional people and the
cost of compiling survey results are lower for Internet surveys.
4.6.3.3 Lead Time The time between the kickoff of the survey to
the completion of survey data analysis is the survey lead time. This
includes the time required to contact participants and for them to re-
spond, and the time for following up and compiling the data. Compared
168 Chapter Four
with mail-based surveys, Internet surveys can cut down the time re-
quired to contact participants and compile survey results. But Internet
surveys cant cut down the time required for participants to respond.
Overall, Internet surveys are somewhat faster, but they may not be as
fast as many people hope for.
4.6.3.4 Data Quality Survey data quality usually refers to the qual-
ity of submitted surveys. The data quality is usually measured by the
number of participants who missed at least one question, or by the per-
centage of missed items on questionnaires. For open-ended questions,
longer answers are usually considered to be more informative and of
higher quality. Research (Schonlau 2002) has shown that compared
with mail-based surveys, more participants will miss survey questions
in Internet surveys. However, for open-ended questions, Internet sur-
vey participants usually give longer answers.
In summary, there are pros and cons for both paper-based surveys and
e-mail/Internet-based surveys. Overall, because more and more people use
e-mail/Internet as their dominant way of communication, e-mail/Internet
surveys are gaining ground. However, face-to-face interviews and tele-
phone interviews can provide more direct personal interaction between
survey researchers and customers, and the quality of the survey data is
also better than that from both paper-based surveys and e-mail/Internet
surveys, so they are still indispensable forms of customer surveys.
Chapter
Proactive Customer Information

5
GatheringEthnographic
Methods
In developing a new product, one of the key questions is What do custom-

ers really want? In other words, What is the real voice of the customer?
In the last chapter, we discussed methods for collecting VOC data, such
as surveys and focus groups. These are important and dominant meth-
ods of collecting VOC data, and they have been in use for many decades.
The common feature of these methods is that they are based on what
customers say, either in groups or individually, in person or over the
phone, on paper or on the Internet. However, if product developers rely
on methods that are based on what customers say, rather than on what
they do and how they use the product, they could end up with significant
errors in capturing the real voice of the customer.
In one focus group study conducted by Microsoft, customers were
asked to use Microsoft software for a few hours and then answer ques-
tions interactively. It went something like this:
Question: Did you like the product?
Answer: Yup!
Question: Any features you do not like or want to add?
Answer: Nope!
Based on these answers, it might appear that Microsoft had a winning
product. But when Microsoft developers began recording keystrokes and
videotaping customers behavior, they discovered a wide range of nega-
tive customer reactionsgrimaces, hesitations, etc. This simple example,
and numerous other cases, demonstrate that traditional VOC collec-
tion methods based on what customers say are not adequate. There are
169
170 Chapter Five
many problems with simply asking customers what they think about
a product:
What customers say is highly dependent upon how they are asked.
Customers usually will not talk about problems that are not related
to the interview or survey questions. However, survey and interview
questions are usually designed by the product developers or their
business partners, so they are usually highly subjective and are deter-
mined by the mindset of product developers.
Customers usually are not technical experts. They generally cannot
express the precise problems they are having with a product just by
talking about it. They often have difficulty describing what type of
product they need or want.
When customers are asked about things that happened in the past,
their answers are subject to many kinds of errors. Customers tend to
remember events in a way that conforms to what they expect; they
are highly influenced by the way questions are asked.
Customers ability to describe a products problems usually depend on
how familiar they are with the product. When they are not familiar
with the product and struggle to learn how to use it, they notice dif-
ficulties in using it. When they have used a product repeatedly, their
usage becomes routine and they have a hard time noticing details.
Customers are often reluctant to state the truth to avoid sounding
ignorant or exposing their perceived weaknesses. They see themselves
as possessing qualities that they do not really possess, so they answer
questions for their ideal selves rather than their real selves. They are
embarrassed to admit that they have problems with products because
they fear those problems might reflect poorly on their abilities.
Customers have all sorts of agendas that interfere with strict accu-
racy. They develop opinions about what interviewers want to hear and
consciously or unconsciously tailor their statements based on these
opinions.
Research indicates that what people say does not exactly reflect what
people do. Wicker (1969) found that there is only a weak correlation
between speech and behavior. Loftus and Wells (1984) showed that
peoples descriptions of past events are highly fallible. Clearly, by using
traditional VOC collection methods alone, we cannot get accurate infor-
mation on what customers really want. We need some other method to
supplement the traditional techniques discussed in Chapter 4, so we can
capture what customers really want rather than just what customers
say. Ethnographic methods can be used to capture this extra informa-
tion, and this chapter will look at them in detail.
Proactive Customer Information GatheringEthnographic Methods 171
5.1 What Are Ethnographic Methods?

Ethnographic methods are research methods used by anthropologists to
understand peoples behavior in context. Ethnographic methods are used
in social science to study peoples social behavior, such as family and com-
munity relationships, cultural beliefs, and values in their native habitat.
Ethnographic methods are used in business and product development
to study peoples behavior when using products in the actual environ-
ments of use, such as using a recreational vehicle (RV) on highways and
in trailer parks, using detergent in coin laundries or at home, and so on.
Even if the product is not yet available, ethnographic researchers will
make every effort to learn how users will use something similar to the
product, such as a prototype of the new product, a competitors product,
or an old generation of that product. It is important to observe peoples
behavior in realistic environments, because each usage environment is
unique. Without knowing the specific usage environment and how this
environment will affect the usage of the product, it is difficult to develop
a top-notch product that works perfectly with consumers.
In Jeff Likers popular book, The Toyota Way (Liker 2004), he tells
the story of the redesign of the 2004 Toyota Sienna. In this redesign
effort, Yuji Yokoya was assigned to be the chief engineer. Yokoya had
worked on Japanese and European projects, but he had never been to
North America, and the North American market is the native habi-
tat for the users of the 2004 Sienna. Yokoya didnt feel he understood
the North American market, so to overcome this deficiency, he flew to
North America and drove an older version of the Toyota Sienna to all
48 contiguous states in the U.S., 13 Canadian provinces, and all parts
of Mexico. On this extended trip, he found quite a few surprises and
made many design changes that would make no sense to a Japanese
engineer living in Japan. For example:
The roads in Canada have a higher crown than in America (they are
more bowed up in the middle), perhaps because of the amount of snow
in Canada. He learned that controlling the drift of the minivan is
very important for drivers in Canada.
When Yokoya was driving on a bridge over the Mississippi River,
a gust of wind blew him very hard, and Yokoya realized that side-
wind stability was very important. Driving through the crosswinds
of Ontario, he was alarmed at how easy it was for trucks to blow the
minivan aside. Based on this observation, the newer Sienna design
improved its ability to handle crosswinds.
When he was driving the narrow streets of Santa Fe, Yokoya found
that it was hard to turn the corner with his older version of the Sienna.
Based on this observation, Toyota improved the turning radius of the
172 Chapter Five
newer Sienna by 3 feet. This is a huge improvement, since the 2004

model was also significantly larger.
By practically living in the Sienna for all his driving trips, Yokoya
learned the value of cup holders. In Japan, driving distances are usu-
ally shorter, and while you can buy drinks in cans, it is more common
to drink them outside the car. In America, distances between stops
can be much longer, and you need space to put drinks in the carnot
just coffee, but other kinds of drinks as well. Based on this experience,
there are 14 sturdy cup holders in the 2004 Sienna, so each passenger
can have 2 to 3 cup holders.
Yokoya also noticed the American custom of eating in cars rather than
taking the time to stop and eat. In Japan, the roads are narrower
and there are many trucks around, so driving is more stressful and
requires high concentrationdriving while eating is not a good idea.
In America, the roads are wider and less crowded and driving is more
relaxing, but trip distances are longer. So it makes sense to eat in the
car. Yokoya learned the value of having a place for hamburgers and
fries, so he added a flip-up tray accessible from the drivers position.
Besides the results of Yokoyas experience, Toyotas global R&D head,

Dr. Akihiko Saito, also found that Americans buy larger items, such as
4' 8' sheets of plywood, and put them into the back of their pickup trucks
and minivans. Dr. Saito noticed that the older version of the Toyota Sienna
did not have this capacity, so he approved the change of design for the new
Sienna so that it could accommodate a 4' 8' sheet of plywood. As this
story demonstrates, it is important in the product development process to
study consumers behavior in the actual product usage environment.
In general, ethnographic research is the direct, first-hand observa-
tion of daily behavior where the researcher may even participate as a
participant observer. It is a research method based entirely on fieldwork,
and it seeks to observe phenomena as they occur in real time. In other
words, ethnographic researchers have to study behavior in natural set-
tings, sometimes getting their hands and pants dirty. Another principle
of ethnography is that you cannot understand peoples behavior fully if
you dont understand the symbolic world of the people you are study-
ingyou have to see the world through their eyes and use their shared
meanings. This involves learning the language in use, the dialect, any
buzzwords or jargon, special words, and so on. Ethnographic research
involves staying in the fieldyou have to walk miles in a customers
shoes to develop a real understanding of their experiences.
Unlike regular customer survey studies, in which the sample size
is usually large enough to make statistically sound inferences, eth-
nographic research uses small but good samples of informants.
Informants are people who may have special knowledge of the area
under study. When ethnographic research is used to study indigenous

tribes, the informants are usually influential and knowledgeable
people in the tribe; when ethnographic research is used in customer
research and product development, the informants are usually lead
users (usually heavy users or state-of-the-art users) of the products.
For example, in an ethnographic research study for computer-aided
design (CAD) software, the lead users could be research and devel-
opment scientists or engineers who use CAD software very heavily in
advanced research, who often take their use of the software to the edge,
and who often develop driver programs or customize the software. These
people really understand the product and they often can provide valu-
able insight on it. Ethnographic researchers can work with informants
to gain in-depth knowledge in the area.
5.1.1 Frequently Used Ethnographic

Methods
There are several frequently used ethnographic methods: participant
observation, nonparticipant observation, formal interviews, informal
interviews and casual conversation, and informant diaries.
5.1.1.1 Participant Observation Participant observation is an approach

in which you are directly involved in community life, observing and talk-
ing with people as you learn from them their view of reality (Agar 1996).
In ethnographic research, there are four levels of involvement in observa-
tion: the complete participant, the participant-as-observer, the observer-
as-participant, and the complete observer (Gold 1958). However, as
Hammersly and Atkinson (1983) pointed out, the quality of information
that can be gathered by people at either of the two endpoints is severely
limited, and the difference between the middle two is highly question-
able. The ideal approach is to minimize the effect of the researcher on the
researched, and to maximize the depth of information that is obtained.
The actual approach can range from a hidden or disguised voyeur to an
active participant who acts as a member (not as a researcher), as long as
these roles will not alter the flow of interaction in an unnatural way.
5.1.1.2 Nonparticipant Observation In some cases, it is undesirable to

conduct customer research as a participant observer because the in-
troduction of an outsider may disturb or destroy the natural customer
behavior. In this circumstance, nonparticipant observation is preferred.
One typical example of this is when gathering data about how parents
and children interact, as described by Rust (1993). In this study, re-
searchers posed as shoppers in supermarkets and toy shops, waiting
in aisles for a parent-child group. On seeing a shopping party enter
the aisle, the researchers would estimate the childs age, record some
174 Chapter Five
basic information about the shoppers, and then take notes on what the
parent and child said and did. As soon as the party left the aisle, the
researchers would finish up the notes and wait for the next party to
appear. In this study, 200 records were gathered and used to develop a
number of marketing strategies.
5.1.1.3 Formal Interviews Formal interviews are important tools for

data collection in ethnographic research. A key feature of the ethno-
graphic interview is the use of nondirective questions. These questions
are designed as triggers that stimulate the interviewee into talking
about a particular broad area. The focus of the interview is not to draw
conclusions on a given set of hypotheses, but to uncover as many un-
known details about the research subjects as possible.
5.1.1.4 Informal Interviews and Casual Conversations Much of the rich

data that ethnography can get comes from the whole realm of informal
talk between researchers and informants. The main feature of informal
interviews is that the researcher does not have a written list of ques-
tions, but rather a set of question-asking strategies to let the conver-
sation flow. These interviews take place in all kinds of circumstances,
such as when working with informants, taking a coffee break, relaxing
in front of the TV, and so on. Informants have all kinds of freedom in
responding to questionsthey can criticize a question, correct it, and
answer in any way they want.
5.1.1.5 Informant Diaries Informants can be asked to keep diaries rel-

evant to the research area. Because these diaries are created by the
informants themselves, they provide some untainted information.
5.1.2 Data Recording Methods

There are a couple of common methods of recording data in ethnographic
research, which include making field notes and taking photographs or
videos.
5.1.2.1Field Notes In the ethnographic tradition, field notes play a

very important role. Field notes are records of social activity written
up by the researcher as soon as possible after the event. According to
Spradley (1980), a field note should provide the following information:
Space The physical place or places

Actor The person or people involved
Object The physical things that are present
Act A single action that the actors perform
Activity A set of related acts that the actors perform

Event A set of related activities that the actors carry out
Goal What the actors are trying to accomplish
Time The sequence of acts, activities, and events over time
Feeling The emotions felt and expressed.
5.1.2.2 Photographs and Videos Visual data can be extremely useful

in interpreting behavior, including the temporal flow of events, cultur-
ally significant moments, and human-object interactions. They can also
be used to stimulate discussion with informants.
5.1.3 Types of Ethnographic Research

Used in Product Development
In product development, there are three kinds of VOC research in which
ethnographic methods play important roles: discovery, definition, and
evaluation.
5.1.3.1 Discovery Research Discovery research is an open-ended ex-

ploratory effort to learn about consumer culture. It is useful for devel-
oping original product and service ideas, or for finding new applications
for existing and emerging technologies. In discovery research, ethno-
graphic researchers collect and analyze a combination of verbal, obser-
vational, and contextual information to identify what people say and
do in their natural environment. The inconsistencies in what customers
say and do, and the work-arounds people use, can help to identify
unarticulated, unrecognized needs and gaps.
The ethnographic methods and other VOC data collection methods
used in discovery research include:
Contextual observations
Participant observations
Video recording
Open-ended interviews
Focus groups
Content analysis (text, audio, and video)
One good example of discovery research is described by Squires (2002).

He describes an ethnographic case study about family morning routines
and breakfast behavior. In this case study, researchers visited families
as early as 6:30 in the morning, and observed details about how people
chose their breakfasts. Overall, they found that people want good quality,
176 Chapter Five
good portions, and nutritional meals; however, the time pressure made
American families start their days earlier and earlier, when many kids
and teenagers were barely over their sleeping cycles. So what was actu-
ally eaten at breakfast time was very far from parents expectations,
ranging from eating very little to skipping breakfast entirely. Based on
these observations, Yoplait developed Go-Gurt, which is squeezed out of
a tube directly into the mouth, instead of being scooped with a spoon.
The sale of Go-Gurt took off quickly, and the key to the success of this
product was that it filled unarticulated needs: kids liked it because it
was fun and mobile, and parents liked it because it was nutritious food
for their children.
5.1.3.2 Definition Research Definition research is used when there is

already a product conceptit helps developers take a product concept
and turn it into a meaningful form. Definition research can be used to
determine what the product should do, how it should be used, and how
it can be promoted to consumers. Therefore, definition research focuses
on identifying the use and use features of a product, and on firming up
what a fully functional product prototype should be like.
The methods used in definition research are different from those of
discovery research because we already know the product concept, the
key parameters of the product, and something about the consumers.
Because of this knowledge, we dont need to start from a clean slate
we can zoom in on the user and the product. This research can compare
existing products in context and determine how the product concept
and design could be changed to fit users daily use. The ethnographi-
cal and other VOC data collection methods used in definition research
include:
Contextual observations
Directed and semistructured interviews
Ratings, rankings, and comparisons of the relevant products
Participant observation
Scenario analysis
Squires (2002) offers a good example of definition research in his dis-

cussion of a project for a personal care product. Personal care products
include tissues, lotions, toothbrushes, and so on. Usually people use
them in their homes, but some professionals, such as front office sec-
retaries or salespeople, need to use them during long working days. In
this case, the product concept exists, but the personal care product needs
to be modified to fit into the front office environment. In this type of
product development, it is very important to know the culture, the work
surroundings, and the user context for potential customersthe office

workers. Ethnographic researchers need to understand the culturally
appreciated practices for using these products. Researchers found that
while it is important to maintain ones personal image by using these
personal care products during workdays, their usage should be kept low
key. In this case, the researcher found that packaging these products to
fit in with the office environment was the most important feature.
5.1.3.3 Evaluation Research Evaluation research deals with situations

where there is already a working product or prototype. Evaluation re-
search helps to validate the product, test customer usability, and refine
or fix feature details. Usually companies perform evaluation research at
the end of product development. The ethnographic methods and other
VOC data collection methods used in evaluation research include:
Surveys
Focus groups
Usability tests
Product simulation
User-experience interviews
5.1.4 Key Winning Factors for

Ethnographic Methods
In recent years, ethnographic-based VOC methods have been on the
rise. So what are the key advantages of ethnographic VOC methods
over traditional VOC methods? The following sections outline the key
winning factors for ethnographic methods that set them apart.
5.1.4.1 Objective and First-Hand Information Just as with any anthro-

pology practices, the ethnographic researchers conduct their investi-
gation in a beginners mind. That is, they should not bring any pre-
existing judgments into the study, but observe and record all the fine
details during the fieldwork as they happen. On the other hand, in the
traditional VOC data collection methods, such as customer survey and
focus group interview, the designing of the survey questions is very
dependent on pre-existing judgments.
5.1.4.2 Detailed and More Accurate Information In ethnographic re-

search, ethnographic researchers observe how customers use the prod-
uct in real time and in the actual usage environment. They can observe
and record what goes right with the product, and what goes wrong
in great detail. These observations could reveal unexpected product
178 Chapter Five
usage, product inconvenience, and failure modes. On the other hand,

the completeness and accuracy of the traditional VOC methods are se-
verely limited by the product manufacturers subjective assumptions
about how the product is used.
5.1.4.3 Audio and Video Documentation Audio and video recording is

a part of standard practices in ethnographic research. Audio and video
recording often captures some information that cannot be captured in
traditional VOC methods, such as facial expression, hand gestures, re-
luctance, and other body language. This information can reveal many
important hidden facts about customers true opinions, doubts, and
true appreciation for the product.
5.1.4.4 Cultural and Contextual Meaning Capture One of the basic

principles in ethnographic research is to understand the customers
worldview, cultural values, and the meaning of the product in their life.
Ethnographic researchers do this by observing and even living with
customers to study their schedules, their lifestyles, their work environ-
ments, and their buzzwords. By putting all these puzzles together, the
ethnographic researchers can gain a very deep understanding of cus-
tomers values and opinions on the product.
5.2 Ethnographic Research

Project Planning
The first step in any ethnographic research project is planning. There
are several steps in the planning:
Research objectives Determine what is to be accomplished with

the project.
Research design Decide what kind of informants, or customers,
the project will work with.
Research methods Determine what kind of ethnographic meth-
ods will be used in the project, such as participant observation, video
recording, and so on.
Research team and rules Develop the ethnographic research
team and set up the ground rules for the project fieldwork.
5.2.1 Determining Research Objectives

Ethnographic methods can be applied to many stages of the product devel-
opment process, as was mentioned in Section 5.1. We can generally clas-
sify ethnographic research projects into three types: discovery research,
definition research, and evaluation research. In an ethnographic research
project, the objective should be specific enough to give direction but broad
enough so that we are not limited to any predetermined narrow frame-
work. In most cases, we can roughly define three types of ethnographic
projects, as illustrated in Figure 5.1.
When we deal with fuzzy front end problems, we dont have a
product concept defined yet, and we are researching product concepts,
discovering market opportunities, and looking for new product develop-
ment opportunities. This situation calls for a discovery research project.
When we have already developed a product concept, but are not sure
about product details, or how to satisfy specific needs from customers,
or market-needs details for the product designs, we should go conduct a
definition research project. When we already have developed a detailed
product design, and we want to launch the product and develop the best
marketing strategy and packaging and test the product, we should set
up an evaluation research project.
Example 5.1. Lipstick Study One lipstick company that wanted to

maintain and enhance their products market position wanted to de-
velop new lipstick products and brand extensions. Ethnography was
chosen as a research method because it would allow researchers to bet-
ter understand lipstick products and usage in context. When determin-
ing research objectives, they decided to focus solely on women in the
context of their lipstick usage. They had two objectives:
To understand how, when, and where women used lipstick (specific
lipstick usage)
To understand the needs and drivers for women who used lipstick
(contextual behavior)
Product Development Process
Fuzzy Front End Concept Design, Product Design Launch

Determine customer needs Learn how customer use product Test new product
Uncover unmet needs Discover functional emotional benefits Determine packaging
Find problem to solve Determine product satisfaction and dissatisfaction
Generate new product Determine product performance targets
Ethnographic Research Project
Discovery Definition Evaluation

Research Project Research Project Research Project
Figure 5.1 Types of ethnographic research projects

180 Chapter Five
Clearly, this project was a definition research project, because the

product conceptlipstickwas not new (Wellner 2002).
5.2.2 Recruiting Informants

After the research objectives have been determined, it is important
to determine who the best informants are, or which key custom-
ers should be interviewed or observed, and how to recruit them. The
research objectives determine what types of informants, or custom-
ers, are needed for the study. For example, for a definition research
project, the product concept has already been developed, and existing
customers can be studied for this kind of product. However, for a dis-
covery research project, the type of informants should be considerably
different.
Example 5.2. Lipstick Study (Continued) A lipstick manufacturer

wants to beat the competition and promote its own product. This calls
for a definition research project. The product concept (the lipstick) had
already been developed, but this project is not benchmarking or evalu-
ating the productit is researching customers needs so that the manu-
facturer could improve the product or the marketing. The ethnographic
team selected the following four types of customers:
Loyal lipstick users

Occasional lipstick users
Those who rejected the usage of lipstick
Those who were aware of lipstick but did not use it
Clearly, this selection includes a very broad base of current customers

and people who have opinions on the product.
Example 5.3 Breakfast Study An ethnographic research project

was conducted to study the breakfast habits of Americans. The spon-
sor of this research, a large breakfast food company, was interested in
discovering new product development ideas. This project is discovery
research. The team didnt have a specific product concept to evaluate,
so the target informants needed to include more than the sponsors
existing customers, or just packaged food consumers. In this study,
the target customers were all the different types of Americans who
consume breakfast, such as men, women, kids of different ages, and
senior citizens, people with different income status, education levels,
and so on.
5.2.2.1 Psychographics, Lifestyles, and Sample Size Besides the basic

informants qualifications, which are based on research objectives,
psychographics is also worth considering when recruiting informants
psychographics focuses on the attitudes people have about themselves,
their lives, and their futures. Psychographics are criteria for segmenting
customers by attitude, personality, value system, behavior, and
experience.
Determining informants psychographics profiles is valuable and
important in the following respects:
It is important to choose informants with psychographics profiles

that match the objectives of the ethnographic research project. For
example, if you are doing discovery research on a high-end product,
you should recruit informants who are very self- confident and have
upbeat personalities and mostly successful experiences.
When you try to make a case based on an informants interviews and
observations, you have to pay attention to the informants psycho-
graphics profile. For example, if the informant is a successful profes-
sional who has a high income, a stable job, a very positive outlook on
the future, and identifies him/herself with prestigious brands, and
who is very positive about the product concept, but the product is
really designed for the low-end mass market, you should be careful
about making inferences about the products market potential.
Lifestylethe way a person lives his/her life and spends his/her

moneyis also an important factor in recruiting informants. For exam-
ple, does he spend most of his time and energy making his home beau-
tiful? Does she put a lot of her money into travel? Does your audience
consist of young people who spend most of their money on clothes?
Lifestyle profile is important when you are dealing with a product or
service that expresses personal identity.
The sample size is the number of informants who should be recruited
in the ethnographic research project. Based on the experience of ethno-
graphic researchers (Abrams 2000), if the sample size is equal or fewer
than 15, it is likely that just a few eccentric people will throw off the
conclusion of the study. If the sample size is 50 or more, the project will
often be too long and costly. A sample size between 20 and 30 is very
common in ethnographic research projects, and if enough attention is
paid to interviewing and screening informants, this sample size is good
enough to extract a rich profile of information.
5.2.2.2 Recruiting and Screening of Informants The recruitment of in-

formants consists of two tasks: generating the initial pool of candidates
182 Chapter Five
and them screening them. The initial pool of candidates can come from
the following sources:
Known customers or friends of friends: This type of informant may

work very well because there is mutual trust between them and eth-
nographic researchers. However, it is necessary to prevent and reduce
the possible bias from this kind of customer.
Customers from past focus groups: Interesting and articulate focus
group participants may be good candidates for ethnographic research
projects. They may have already been screened, and be known to be
articulate and responsive participants.
Ads in newspapers, newsletters, postings, and online bulletin boards:
This recruiting method may provide a large fraction of unsuitable
candidates, so careful screening is essential.
The screening of candidates is usually done by a screening survey.

This survey contains many questions that screen the candidates based
on demographic, psychographic, and lifestyle considerations. The follow-
ing are key ingredients for a good screening survey:
Start with broad questions and move to more specific questions

Ask nonleading, multiple-choice questions
Screen the person for past participation in market research and com-
petitive employment.
Example 5.4 shows a complete screening survey.
Example 5.4 A Screening Survey for an Ethnographic Research

Project An ethnographic research project is launched for a food com-
pany that developed a new product prototype consisting of noodles
and a flavored coating for chicken. Here is the complete screening
survey:
Ethnographic Study Participants Survey Questions

Hello. My name is _____________ from ____________________, a
research company in ____________________. Were conducting a study
on meals and would like to include your opinion. Do you have time to
answer a few questions?
1a. Do you or anyone in your household work for:

( ) a market research company or advertising agency?
(TERMINATE)
( ) a food processor, retailer, or wholesaler or any related busi-

ness? (TERMINATE)
( ) none of the above (CONTINUE)
1b. Have you participated in a market research study, focus
group, or telephone survey concerning any food product
during the last six months?
( ) Yes. (TERMINATE)
( ) No. (CONTINUE)
2. Do you shop and prepare meals for your family?
( ) No. (TERMINATE)
( ) Yes. (CONTINUE)
3. Does your family include at least one child between the
ages of:
( ) 6 and 10? (CONTINUE)
( ) None of the above (TERMINATE)
4. Is your spouse living with you at present?
( ) Yes
( ) No
At least 15 families must include both husband and wife.
5. What category best describes your total annual household
income?
( ) Under $30,000 (TERMINATE)
( ) $30,000$40,000 (CONTINUE)
( ) $40,000$50,000 (CONTINUE)
( ) $50,000$60,000 (CONTINUE)
( ) Over $60,000 (CONTINUE)
Respondents incomes should represent a spread.
6a. Do you sometimes prepare meals with one or more of the
following side dishes?
( ) Prepackaged, flavored rice? (CONTINUE)
( ) Prepackaged, flavored, dried noodles? (CONTINUE)
( ) Prepackaged, flavored, dried pasta? (CONTINUE)
( ) None of the above (TERMINATE)
6b. How frequently do you use these kinds of side dishes?
( ) Less than twice a month (TERMINATE)
( ) At least twice a month (CONTINUE)
184 Chapter Five
7. Im going to read the names of various brands to you.

Please tell me if youve ever used them, if youve used them
in the past three months, or if youve used them in the past
month:
Past Three
Product Ever Months Past Month
Kraft Rice & Cheese ( ) ( ) ( )
Noodle Roni ( ) ( ) ( )
Uncle Bens Long Grain & ( ) ( ) ( )
Wild Rice
Rice-A-Roni ( ) ( ) ( )
Savory Classics ( ) ( ) ( )
Uncle Bens Country Inn ( ) ( ) ( )
Kraft Pasta & Cheese ( ) ( ) ( )
Golden Saute ( ) ( ) ( )
Lipton Rice & Sauce ( ) ( ) ( )
Lipton Noodles & Sauce ( ) ( ) ( )
Lipton Pasta & Sauce ( ) ( ) ( )
If respondent has not used any one of the above brands during
the past month, TERMINATE.
If respondent has used at least one brand during the past month,
CONTINUE.
8a. Have you used Hamburger Helper at any time during the
past six months?
( ) No
( ) Yes
Respondents must include at least five Hamburger Helper
users.
8b. Have you used Shake N Bake at any time during the past
six months?
( ) No
( ) Yes
Respondents must include at least five Shake N Bake users.
9. How frequently do you serve your family chicken?
( ) Less than once a week (TERMINATE)
( ) Once a week or more (CONTINUE)
10a. Have you served pasta or noodles with an alfredo flavor

sometime in the past six months?
( ) No
( ) Yes
10b. Would you consider serving pasta or noodles with an
alfredo flavor?
( ) No
( ) Yes
If answer is No to questions 10a and 10b, TERMINATE.
If one answer is yes, CONTINUE.
A well-known food company would like to invite you to participate in
a very unusual market research study. It would involve an interview in
your home as you prepare and serve a meal to your immediate family.
The interview would be conducted by two women and would last an
hour to an hour and a half. The interview would be videotapedbut
only for market research purposes. Your privacy would be completely
respected. You would receive seventy-five dollars for your time and your
participation, plus the cost of main ingredients other than those we
would supply.
Respondents name: ___________________________________________
Address: ______________________________________________________
City: ____________________________ State: ________ ZIP: __________
Home phone: __________________ Office phone: _______ ___________
5.2.3 Selecting Research and Data

Collection Methods
Based on the specific objectives of the ethnographic research project, the
appropriate research methods and data collection methods have to be
determined. The most frequently used research methods are participant
observation and customer interviews.
Participant observation means to dive deep into the environment

where the product is being used. In a participant observation study,
detailed notes are taken about observations on product users, usage
environment, and all activities associated with the usage. These notes
will be analyzed to discover possible new insights about the product.
Interviewing is the other part of the equation. The research team
explores and probes customers actions and beliefs in real time. This
allows the research team to check their assumptions, which leads to
new and robust interpretations. Interviews are generally one-on-one
and in-depth, but sometimes the interviews are done in groups, if
186 Chapter Five
appropriate; for example, research that is conducted with a teen best-

friend pair, family unit, or an adult group of friends who share a passion
or influence each other. The interviews are open-ended, which allows
customers to cover all the relevant topics based on their own way of
seeing the world. Subsequent interviews are typically more structured,
as knowledge and relevant questions are developed that need testing.
The amount of time spent with the customer and the number of visits
depend on the research objectives, but some general rules are as follows:
Longitudinal versus short study A longitudinal study is defined

as a study with more than one cycle of fieldwork, and a short study
is one cycle of fieldwork. If a subject topic is intimate or of a private
nature, such as feminine hygiene or finances, a longitudinal study
is appropriate. In this case, it is important to build rapport with the
interviewee over time before delving into more personal topics; thus
multiple visits to a site would be required. A short ethnographic study
is appropriate for everyday products that are not private and do not
require a lot of cognitive thought.
Number of site visits or interviews Ethnographic sampling is
very different from sampling methods used in quantitative work. An
ethnographic sample is quite small and broadly representative of
a constituency. The goal is to understand and gain insight, not the
kind of generalization that comes from more quantitative and larger
samples. Like a detective, ethnography looks for patterns and clues
that get closer to the reality of others. As more time is spent with the
customer, more insight is gained. It generally takes about 9 to 20 in-
depth site visits to gain the most insight from the participants.
In ethnographic research projects, there are primarily four data col-

lection methods:
Field journals or field notes A field journal is a written record

of the site visit. The research team takes verbatim shorthand notes
during the site visits and then revisits those notes to fill in any gaps
and highlight the most important points. Notes are critical to the
teams ability to debrief after their field experience and create an archi-
val record. This method should be used in most ethnography studies.
Audio recordings Audio recordings of the site visit allow for an in-
depth objective review of the content. Choose high-quality equipment
that is as unobtrusive as possible.
Photographs Photography provides a visual picture of the envi-
ronment of the site. This method combined with others allows other
people to get a feel for the context of the entire site. Equipment varies
for this method, but again, unobtrusiveness is important. One benefit

of photography is that it can be relatively inexpensive, depending on
the type of camera, yet still provide a visual record of the site visit.
Digital photographs can easily be sent with the research to others.
Videos Video can provide a visual record of conversations, behavior,
and the environment of the site visit. Video combines a visual and
verbal record of the conversation and gives the most realistic perspec-
tive for those not involved in the research. This method can be more
obtrusive and expensive and needs to be weighed against the goals
of the research and comfort of the consumer. Coding, editing, and
transcribing the tapes can be time-consuming, but if done correctly,
it can provide a wonderful overview of the research.
The selection of data collection methods should fit the objectives of the
ethnographic research project and should be practical in the field.
Example 5.5 A Complete Ethnographic Research Project Plan This

is a project plan for a manufacturer of plastic bags for trash and gar-
bage disposal.
Objectives
To understand how consumers store, sort, and dispose of their trash
To uncover problems and frustrations involved in handling trash
particularly those not currently addressed by available products
To help open potential areas for new trash-handling product develop-
ment or current product repositioning.
Methodology
House Calls will videotape and interview consumers in their homes
in two or three separate suburban areas to discover where and how
they store various kinds of trash, how they sort their trash, and how
they dispose of it.
Detail
There are various kinds of common household trash:
Garbage generally developed in the kitchen
Wastebasket trash (usually accumulated in bathrooms, bedrooms,
home offices, and so on)
Newspapers and magazines
Dead leaves and grasses (probably seasonal)
Heavy-duty cleaning trash (possibly seasonal)
188 Chapter Five
In addition, trash can be categorized as wet or dry, organic or nonor-

ganic, cans, bottles, plastic, or metal.
House Calls will seek to understand whether consumers in various
parts of the country make these distinctions and whether they treat
different kinds of trash in different ways.
Half the recruitment will take place in areas with strong recycling
mandates and half in areas with no strong recycling laws. Respondents
will be told not to empty their garbage or wastebaskets before the
interview.
House Calls will perform a visual inventory of all the trash in each
household, encouraging respondents to talk about the origins and
frequency of trash accumulation in each case. Respondents will be
probed on the benefits and problems associated with each trash recep-
tacle, covering the following issues:
General appropriateness for specific trash
Adequacy of size
Weight or thickness of material
Durability/breakage
Leakage
Sanitation
Closure
Odor containment
Aesthetic concerns
Disposability
Cost
Other problem issues
Respondents will then be asked to empty trash containers as they
normally do: into larger bags, garbage cans, or whatever the final
disposal container may be. If respondents usually sort trash in any
way, they will be asked to do so at this point.
The interviewer will probe respondents on these issues as they relate
to final disposal containers.
In addition, the interviewer will probe respondents concerning trash
collection or disposal at the local dump. What are the local require-
ments concerning the receptacles respondents use? Does what they
use meet these requirements adequately? If not, why not?
Interviews will last approximately one hour.
Taped interviews will be sent by express mail each night to House

Calls for logging and analysis and further direction, if required. The
client will be informed continuously as the assignment proceeds about
significant attitudes and practices.
Deliverables for Clients

House Calls will provide a summary of the findings. A transcript of
the interview excerpts and videotape narration will be prepared. The
summary tape will contain a detailed examination of habits, practices,
and attitudes concerning trash disposal. The tape will note problem
areas, product gaps, potential positioning opportunities for current
products, and areas for new product development.
Raw footage of all interviews will be available as well.
5.2.4 Developing the Ethnographic

Research Team and Ground Rules
An ideal ethnographic research team is cross-functional, which means
that the team consists of people from different backgrounds, coming
from different departments. The team members should commit to work
together throughout the project. Demographics, job functions, and expe-
rience in ethnography should be balanced among team members. A
professional ethnographic researcher should play a leading rolehe or
she should act as research designer, coach, trainer, and facilitator. The
team should also include chief product design engineers, and people
from the marketing department.
Because the knowledge gained through the ethnographic process is
exceptionally rich, and only a fraction of it can be communicated to
people who have not been in the field, the actual product development
team should perform the field research and participate in the analysis
whenever possible. This process helps a development team break away
from incremental learning and gain a new perspective on the product by
refocusing on the problem as experienced by the customer. The lifeline
of innovation is dirty field knowledge. First-hand field knowledge of
customers needs and desires brings with it insight into how to develop
new products, improve product designs, or fix problems.
After the research team has been selected, it should be broken down
into smaller field teams consisting of no more than three people who
will conduct the individual site visits. Teams of more than three people
tend to overwhelm customers and make them uncomfortable.
It is imperative that those who participate in ethnographic research
be coached and instructed by a trained ethnographer on what to expect
in the field. An experiential learning format is the best way to train
190 Chapter Five
a team to conduct ethnographic research. Their training should cover

the following:
Basic theories and techniques in ethnographic research methods
How to observe and listen
How to develop open-ended, story-laden, and nonleading questions
and probes
How to take notes, use other methods of data collection, build a field
record, and debrief
How to sustain disciplined subjectivity to understand and manage
personal and corporate biases
After the ethnographic research team is established and team mem-

bers are trained, the ethnographic fieldwork guides must be set up.
These are the guides for interviews, observations, and data collection.
The following rules of thumb should be kept in mind when putting
together those guides:
Allow time at the beginning of the conversation to set expectations

for the site visit. Introduce the field team members and their roles;
review the purpose of the research, how long the site visit will last,
and what the customer will be asked to do. Begin to build rapport with
the customer.
Start with a tour of the entire site, home, office, or factory.
Start the visit by asking about broad topics and then move to more
specific topics. Save the sensitive and personal topics until closer to
the end of the conversation.
Ask open-ended questions that allow customers to give as much detail
as they want and that allow the researcher the opportunity to ask
more specific or probing follow-up questions.
Do not ask questions that lead or direct the respondent toward a
particular answer. Leading questions tend to start with words such
as do, are, can, could, and would.
Allow time at the end of the site visit for the respondent to ask ques-
tions of the field research team. At this time, the research team can
physically stop taking notes but mentally needs to be extremely
aware. This is a cue to the respondent that the research is over, and
often the customer will open up.
At the close of the time together, thank the customers, tell them what
is planned for the information collected, and compensate them for
their time.
Have sample and photo notes in your hand, so they are not forgotten.
After the guide is created, review it for any biases. Ask yourself
what hypotheses and filters have already been created in your mind
about the research outcomes. Then rework the guide to make it more
objective.
5.3 Ethnographic Project Execution

After the ethnographic project planning is complete, the research team
can enter the field and execute the project. As was mentioned in pre-
vious sections, the popular ethnographic methods include interviews,
participant observations, and ethnographic data collection. In this sec-
tion, we will look at how those methods are used in real ethnographic
projects.
5.3.1 Ethnographic Interviews and

Documentation
As in regular VOC data collection, interviews with selected customers,
called informants, are an important tool for getting valuable informa-
tion. However, there are several distinct differences between ethno-
graphic interviews and regular customer interviews:
Ethnographic interviews usually happen in the field, for example,

the customers home, car, office, or shopping places.
Ethnographic interviews do not have a fixed list of interview ques-
tions. The ethnographic researcher is trying to let informants to talk
as much as they can and guide the talk to all possible aspects about
the products or issues to be investigated. The researchers want to get
all the possible information from the informants.
Ethnographic interviews will go side by side with participant observa-
tions. Ethnographic researchers will work and play with informants
using the products and see what happens in real time.
Ethnographic researchers will look for and find problems while cus-
tomers are using the products.
Ethnographic researchers will generate reality-based product ideas,
modifications, and improvements with their field work.
5.3.1.1Guidelines for Conducting Ethnographic Interviews While inter-

viewing the informants, ethnographic researchers must act carefully,
nonintrusively, and positively in order to let informants say whats on
their minds. At the same time, the researchers must very carefully
watch the informants behavior and the circumstances, as well as take
notes.
192 Chapter Five
The following guidelines are helpful in structuring the interview

process:
Do not make the respondent feel self-conscious by calling attention to

a behavior or practice; for example, You just skipped that part of the
program. Try not to probe at all until a behavioral routine is reaching
completion.
Researchers should be quiet, laid-back, and unobtrusive. They should
be good observers and listeners, actively interested in everything that
the informant does or says, but being careful not to impose. The last
thing you want is a researcher behaving like a marketing person,
taking control and directing the interview.
Minimize asking why a respondent is doing or not doing something
in a particular way; for example, Why didnt you clean behind the TV
set? In general, avoid making the respondent react defensively. With
patience and keen observation, you will eventually discover whether
that area was avoided because the respondent believed that it was
dangerous, because no current product was adequate, because the
respondent didnt care about the out-of-sight area, or any other rea-
sons. The immediate challenge is likely to provoke a defensive reac-
tion, regardless of the true reason for the behavior.
Keep the same interviewer throughout the whole project if possible.
There is a cumulative learning process that takes place from interview
to interview that works very much to the advantage of the project.
Avoid asking respondents directly to explain or describe what they are
doing step-by-step, let them to do things in their own way and watch
them carefully instead.
Probe behavior gently and indirectly by asking respondents to describe
their goals or intentions; for example, Please describe what you are
trying to accomplish or Please tell me how you expect this to go?
Another way to probe is to ask respondents: Describe what is going
through your mind.
If equipment is taken to the interview, it should be as compact as pos-
sible. No attempt should be made to hide a camera or tape recorder
when a research team is in a consumers home. The person handling
the equipment should be up-front about showing it to the informants,
so they can quickly get used to its presence.
Respect the rights, property, and privacy of the participants. Avoid
interfering with the respondents other responsibilities and relation-
ships. For example, if a friend normally comes by to visit while dinner
is being prepared, it is to your benefit to preserve that pattern of
interaction. Be patient and tolerant if a nonresearch responsibility
(for example, a telephone call from a family member) suddenly breaks

the respondents attention.
Participants must he thoroughly briefed about how the research will
be implemented. Enough background information should be provided
without informing respondents to such a degree that they lose sponta-
neity during the visits. Try to avoid surprises that are likely to confuse
or disorient respondents.
Try to minimize your own impact on the environment being studied
so that you can observe it as naturally as possible. Do not help par-
ticipants with tasks or provide advice or otherwise influence the way
that the respondent naturally behaves. For example, if the respondent
has been asked to prepare a meal, it is incorrect to accept and eat that
mealno matter how tasty it may appearbecause the researchers
satisfaction, rather than the respondents behavior, becomes the focus
of the interaction. Similarly, if the respondent must carry unwieldy
supplies from one location to another during the course of product
usage, it is inappropriate for any site visitor to help.
Be patient and tolerant about the course of activities during a site
visit. Let the respondent set the pace; the ethnographer should not
rush activities or try to stop something already in motion.
Because a high level of intimacy normally develops between the eth-
nographic researcher and respondent, matters that are entirely extra-
neous to the subject under investigation will come up for discussion:
the news of the day, family composition, social activities, and personal
interests. It is only natural for visiting researchers to share details
about their own children or a recent vacation destination. Innocuous
small talk supports rapport building especially when people find they
share some commonality. Nevertheless, researchers should be careful
about discussing their own opinions about their profession, personal
matters, or the category under investigation, because these would also
be out of role.
Even though the observation guide will be used as a map or a blue-
print to the process or behaviors you are exploring, it is not a strict
questionnaire or program of events. Try to follow the guide as much
as possible; but be ready to stray from the guide as needed to follow,
engage, or understand a respondent.
There should be no more than two or three interviewers or observers
so that they dont become intrusive in the household.
Respondents often incorrectly view ethnography visitors as authori-
ties and specialists on the category being studied. It is only natu-
ral for respondents to ask interviewers to evaluate their behavior or
make suggestions for solving vexing problems they have experienced.
194 Chapter Five
Decline graciously, because playing the authority will diminish the

respondents sense of autonomy and aptitude. A good way to deflect a
request for an evaluation is to make a noncommittal remark about the
respondents effort without commenting on results. For example:
Respondent: Doesnt this product put a great shine on my stovetop?

Researcher: You like the stovetop shining, dont you?
If the respondent is given space to feel like the host and expert and
the ethnographic researcher remains modest and behaves like a privi-
leged guest, the ethnographic encounter will prove to be enriching and
satisfying for both parties.
5.3.1.2 Observing and Documenting Ethnographic Interviews In ethno-

graphic interviews, most of the talk and all of the actions should come
from the respondent. The interviewer should play a secondary role. The
questions should be brief and should serve as triggers to start the flow
of actions and words. During the interview, observing informants fa-
cial expressions, body language, and actions can be very revealing, as
shown in Example 5.6.
Example 5.6: Baking Soda Brand When a baking soda-based un-

derarm deodorant was first introduced to the market, an observational
study was made of potential respondents. Many were women who were
using baking soda as a deodorant for their refrigerators, but it often took
the respondents 5 or 10 minutes to make the connection between odor
protection for food and odor protection under the arms. Their uncom-
prehending facial expressions made the manufacturer aware that they
could not assume consumers would instantly see the value of a baking
soda underarm deodorant. The personal deodorizing power of baking
soda had to be played up both on the package and in the advertising.
During the interview, anything related to the product under consid-
eration should be studied and recorded in the interview, which includes
how the product is stored, how the product is related to the customers
lifestyle, and any inconvenience related to the use of the product. These
detailed observations can lead to substantial new product ideas and
improvements, as shown in Example 5.7.
Example 5.7 Pocket-Sized Camera A manufacturer of single-use

cameras made a research visit to a popular zoo and handed out single-
use cameras to people who hadnt used this particular compact, pocket
model before. When watching respondents take pictures of their fami-
lies, researchers observed that they moved their fingers all around the
camera, sometimes obscuring the lens, before they found the shutter
button. In addition, many respondents with larger hands were unsure
about how and where to grip the camera. Their fingers seemed to over-
whelm the equipment.
A brief interview with each respondent after the pictures had been
taken confirmed the observation. They didnt see where the shutter was
at first. A pervasive concern was that, with such a small camera, their
fingers might get in the way of the lens.
Clearly, the pocket-sized model needed to be more user-friendly.
The result was a redesign of the camera, making the model a little
less compact (but still pocket-sized) and the shutter more obviously
accessible.
Interviewing while observing can also reveal subtle aspects of key
customer requirements. In one ethnographic research study, researchers
watched people cleaning the bathroom, and they heard many use the
word shiny or a variation on it. Informants spoke of shining up the
tub or toilet, or pointed to a shiny sink with pride. Probing what they
meant, researchers found that shine was a synonym for cleanliness. If a
bathroom surface was shiny, it was assumed to be clean. Furthermore,
in many informants minds, a shiny surface also meant a disinfected
surface free of germs. This special meaning of shiny can provide a key
input in the functional design of bathroom detergent, as well as in a
marketing campaign.
In another study for a household pesticide, the principal scientist was
very proud of the effectiveness of a newly developed chemical. It could
kill many kinds of bugs, such as ants, effectively. However, after putting
this product on the market, the sales figures were very disappointing.
When a marketing person asked customers what was wrong with the
product, the customers said: It does not work! The marketing person
was surprised because it had been proved that the pesticide did work
remarkably well. When the marketing person prodded further, he found
that the customers sprayed the product on a bunch of ants and pointed
at them saying: See, these ants are still moving; they are not dead, so
this product doesnt work. The marketing person explained that the
ants would die in a couple of days, but the customer replied: The bugs
come and go. After I use this product, I still see bugs, I cant tell whether
they are the old bugs that I sprayed or new bugs. This example shows
that in customers minds, the effectiveness of the pesticide is measured
by its ability to kill bugs instantly.
Knowledge of lifestyles, especially lifestyle changes, is very important
in designing breakthrough products. One of the trends in lifestyles is
that people are getting busier. More and more women continue to join
the workforce, so the amount of time people have to spend doing house-
hold chores is reduced. Two very successful products, Arm & Hammers
Clean Shower and WD-40s 2000 Flushes (Figure 5.2), are examples of
products that are designed to fit these busier lifestyles.
196 Chapter Five
Figure 5.2 Clean Shower and 2000 Flushes
Clean Shower is a spray that prevents soap scum and mildew from
forming, reducing the need for cleaning. 2000 Flushes is a very durable
clip-on toilet flush cleaner that can last for four months. In the market,
both products are very successful because they relieve customers from
tedious, time-consuming housecleaning work.
5.3.2 Ethnographic Observations in Shops

Another field for ethnographic research is where customers shop, such
as in supermarkets, specialty shops, and fast-food restaurants. There
are several ways to conduct ethnographic research in stores. One is to
recruit shoppers beforehand and follow them around stores while they
select items they want, with the researcher interviewing them through-
out. Another way is to watch shoppers as they shop, and record their
shopping processes. You can select shoppers randomly for interviews,
with incentives.
This kind of ethnographic observation can show you the primary, sec-
ondary, and other motivations that prompt customers to select certain
products at the point of sale. You can also observe the effectiveness of
the packaging, promotion, and advertisement of the products. Based
on this valuable information, you can improve your product design, as
well as the sales and promotion. Ethnographic observation can also
reveal which store traffic patterns and displays attract customers to
specific areas and merchandise, so both the retailer and manufacturer
can benefit.
In ethnographic in-store studies for general merchandise stores, such
as supermarkets, the primary tools are note-taking, videotaping, and
interviewing selected customers. Such a study could have many pur-
poses. One common objective is to identify the primary, secondary, and
other motivations when customers do their shopping: What is more
important, the brand? The quality? The price? The next two examples
illustrate these objectives.
Example 5.8 Difficult-to-Access Goods In an in-store study, shop-

pers were observed hunting high and low for a well-known brand of
single-use cameras in a giant West Coast drugstore. When they finally
found what they were looking for, it was locked up in a glass case, and
they had to ask the clerk to open the case for them.
Example 5.9 Disconnection Between Advertisement and Packaging of

the Goods Shopper after shopper in a supermarket passed by a new,
well-advertised category of concentrated beverages. Interviews at the
shelf revealed that shoppers remembered the advertising but were not
connecting it to the display on the shelf.
Example 5.10 shows how ethnographic research revealed the key
factor in juice shopping.
Example 5.10 Juice Beverage Shoppers The following excerpts are

from actual interviews taken in an ethnographic research project con-
cerning juice shoppers in large supermarkets at various locations. The
material was recorded right at the shelf. First, the shoppers were filmed
with a hidden camera and the same shoppers were then interviewed in
front of an open camera. These excerpts show the kind of information
you can expect to obtain from random in-store interviews. The numbers
represent codes differentiating the interview portions.
Climbs Up Shelf to Get Capri Sun
MIS 5: 23:00
Shows 8-pack boxes of Capri Sun Red Berry and Wild Berry
MIS 5:23:15
My three kids drink them. Ages 15, 11, and 7. They ask for flavors and
make me write them down.
MIT 6:30:40
I know my daughter likes the punch flavor.
Daughter shows boxes.
MIS 1:12:13
The Juicy Juice boxes
MIS 1:12:28
Plan to buy this specific one? Not this specific one.
DISS TO MIS 1:12:38
What made you select this particular one? She grabbed it. Grabs it
all the time.
198 Chapter Five
Father and Son

NJS 2:17:10
I let him decide what juice, because Ill buy juice and he wont drink
it. I know hell drink applehis favorite.
He likes Juicy Juice and Motts.
Mother and Daughter Look Through Boxes for Flavor

MIT 6:04:00
Daughter: I look for flavor.
MIT 6:07:00
Mother: We look for juices. Fun juices. Whatever appeals to them.
MIT 5:25:32.
To Daughter: I saw youre the one who actually selected it. How did
you select it?
Daughter: I dont know, it looked good.
T 5:25:44
I had Motts before, and I decided to try a different flavor.
T 5: 26: 54
To Mother: How do you decide?
Mother: Generally, whatever she likes.
In the above notes, MIS, MIT, NJS, and so on are initials of interview-
ers or note takers. By observing the above notes we can see clearly the
strong influence that kids have in the selection of juice beverages. As
many in-store studies do, this one uncovered proprietary information
that has helped the beverage manufacturer, who sponsored it, enhance
marketing and packaging practices.
Shopping ethnographic research can tell you how and why a customer
selects a brand or a product on a particular day. It will not tell you
how and why the customer chooses that brand or product on an ongo-
ing basis. In order to get a complete picture of purchase motivations
and usage patterns, you also need to observe customers in their homes,
offices, and wherever the products are used.
5.3.3 Ethnographic Observations in

Product Usage Processes
The most important ethnographic observations usually happen where
the products are used. By observing customers using the products, espe-
cially paying attention to where and how customers struggle in using
products, you can find the possible weak points in the current product
design, and thus improve on the design.
This kind of ethnographic research is also called usability research. It is

especially effective for understanding the user interface; that is, the inter-
action between people and technology. Usability research originated in the
military as human factors research, and it seeks to assess how easily users
manage their prospective technological tools, and how much satisfaction
that engagement produces (Fox and Fisher 2002; Nielsen 2000). It can be
applied to such technologies as web sites, computer software, ATM dialogs,
automobile dashboard technologies, and cell phone functionality.
5.3.3.1 Key Issues in Usability Research In usability research, there

are several key issues: whose equipment is being used, what factors
will be analyzed, and what mix of objective and subjective measures
need to be acquired.
Whose Equipment Is Being Used? It is only natural to start by con-

sidering the location of the study: whether within special usability
laboratories sponsored by technology companies or research facilities,
or in the respondents own home or at work. Researchers gain the ad-
vantage of consistency and uniformity when they invite respondents to
their own labs.
What Factors Will Be Analyzed? Specific user-interface metrics that

may be observed include:
Ease of learning How quickly can users acquire skills for negotiat-
ing the technology?
Efficiency of use Are the steps associated with achieving a user-
defined objective, such as sending and receiving instant messaging,
economical, and logical?
Memorability Can users recall how to complete tasks after some time
has elapsed, or, like most of us, do they have to relearn functions?
Error frequency and severity When and why are mistakes made?
Do the errors represent some kind of mistaken assumptions about
how people naturally use the product?
Subjective satisfaction Do users like the product? Once factors asso-
ciated with successful product implementation have been learned, they
become part of the standard knowledge about a category and may be engi-
neered into successive versions and alternative technology designs.
For example, if you were studying the usability of the user interface
of a web site, the following factors should be studied:
Navigation Does the site produce the results that users seek? Do they
know where they are at every moment? Can they find what they want?
200 Chapter Five
Does the search mechanism produce the desired results? Is there a high
level of association between web site terminology and users terminology
so that confusing jargon does not misdirect users?
Structure of the web site Is it organized in a logical and meaning-
ful hierarchy?
Layout Is the page pleasing to the eye? Is there sufficient white
space so that each page can be managed without strain? Do repeat-
ing elements have a clear relationship with changing elements? Are
graphics and images aesthetically pleasing, and do they avoid down-
loading complications?
Error messages Are users empowered to take action following
error messages, or do the messages produce paralysis?
What Mix of Objective and Subjective Measures Need to Be

Acquired? When completing a usability test, respondents must gen-
erally complete tasks as behavioral challenges. Objective measures
may include the time needed to complete tasks and the error rates.
Subjective analysis may involve user comments and behavioral obser-
vations. Sometimes, the think-aloud technique is applied, in which
respondents are encouraged to divulge what is going on in their minds
continuously while completing a task.
5.3.3.2 Conducting Usability Research To get a clear picture of how

ethnographic observations are conducted in product usage research,
well look at some typical ethnographic observation scenarios concern-
ing the use of cars, computers, and workplaces.
Cars In recent years, car manufacturers have learned a lot about

improving car design by putting a video camera in the front seat and
recording the driver on an entire commuting or shopping trip. How do
drivers handle the controls? Are the controls well placed? Are they easy
or difficult to use? Researchers have learned from the expressions on
drivers faces as they manipulate the car through traffic and park. What
frustrates them? What annoys them? Are there blind spots that might
cause safety problems or that make parking harder than it should be?
Do drivers have to take their eyes off the road for any length of time to
read the gauges on the dashboard or use the radio? Do they have to crane
their necks to see if theyre going to fit into that tight space at the curb?
Loading the car, either with people or packages, is a natural subject
for observational research. Is there enough room for the familyin
the front and the back? Is the vehicle easy to get in and out of? Does
the trunk or rear portion of a minivan provide problem-free access for
grocery bags and other supplies?
Just the basic knowledge of where a driver takes his or her car each
daythe kind of traffic, the length of the trips, the frequency of loading
and unloadingcan be highly useful to car makers designing and mar-
keting cars for specific consumer segments. Its a good idea either to have
an interviewer riding along with the driver, or to take the time afterward
to go over the video recording with the driver. The interviewer should
probe in detail why the driver does what he or she does. At the same time,
the interviewer can investigate the relationship between car and driver.
The observational researcher can tap into many kinds of meaning
especially in a one-on-one conversation in the intimate, private space of
a respondents automobile. Questioning both indirectly and directly, a
researcher should be able to explore issues of status, personal identity,
power, freedom, and more as they touch on car ownership. Are these
motivations as much of a driving force for a leased car as they are for a
car thats owned? This is an interesting question for both a car manu-
facturer and a dealers association.
There is also a wealth of meaning in the way a car is displayedor not
displayedwhen its not being used. Is it parked in the driveway, close
to the road so that the neighbors and passersby can see it? Or is it kept
within the protection (or secrecy) of a garage? How often is it washed?
How often is it tuned up? Is the inside clean and relatively free of junk?
More than a simple matter of pride, the display and condition of the car
reveals something of the way the owner feels about himself or herself
and about his or her readiness for the next automobile purchase.
It is also very valuable to observe what goes on in the dealership
during the car-buying process. We can record the whole process of the
negotiation in an auto salesmans cubicle, watching and listening with
one camera trained on the buyer and another on the salesperson this
would point out effective and ineffective selling and negotiating tech-
niques and provide the basis for a training video for dealerships, dealer
associations, and car manufacturers.
Ethnographic researchers can also study car sales on the Internet.
It would be interesting to know how Internet shopping affects the car-
buying process. How much of the buyers decision is made in front of
the computer? How much is made when he or she visits the dealership?
Following an online shopper through the entire search and buying pro-
cess would point out ways the marketing processparticularly on a
local levelcan be made more efficient.
Computers There are a number of computer-related consumer expe-

riences in which observational research can be useful: setting up a new
computer (known as the out-of-the-box experience), installing and learn-
ing new software, using a web site, and buying online. For most people,
setting up a computer can be a full days worth of frustration, with at
202 Chapter Five
least three frantic calls to the service number for help. Observational
research may not be able to find a way completely out of this technical
morass, but it could help.
By observing and recordingmoment by momentthe points of frus-
tration, the inadequate digestion of dubious instructions, the mismatch-
ing of cable to port, the actions and words (even if unspeakable) that
result from a printers refusal to print, the whole process of dialing
for help and being asked to stand by while other customers are being
served, and having to explain the problem to one of those voices with
an ever-so-subtle implication that youre just another non-tech moron, a
company can go a long way toward correcting the out-of-the box experi-
ence and making it user-friendly.
Todays software programs are often even more formidable than the
hardware, and the experience of installing and learning to use them
can take hours and weeks of trial and error. There is an ever-increasing
opportunity for observational researchers to help make the software
world a kinder, gentler place.
Observation in this case may be conducted over a period of weeks,
with the observer present at program installation and then returning at
intervals to record progress (or lack of it) as the respondent begins to use
the new software. The objectives of the research neednt only be oriented
toward picking up on and solving problems. The researcher could enlist
the user to figure out ways the program could be made better, easier to
install, faster to learn, with fewer steps and keystrokes.
Frustration seems built into the process of getting from one web page
to another, exploring links, and finding exactly what you want. Watching
people bumble from one icon to another while following inadequate
directions, revealing their feelings through facial expressions and
grunts, and the obversewatching them use an intelligently designed,
easy-to-follow web sitewill offer clear insights into the ways your web
site can be improved.
The researcher sits with the respondent as he or she attempts to
find and navigate the web site under consideration. Every time the
researcher notices a hesitation or an error, the respondent is probed.
Why the pause? What went through the respondents mind when he or
she clicked on the wrong icon? Where did the respondent really want
to go? When all is said and done, was the web site worth the trouble?
Would respondents return to it in the future? If not, why not? What
would make it more accessible? There is nothing like direct observation
of the experience to find reality-based answers to these questions.
Broader studies of computer usage have revealed that, for some fami-
lies, the computer has replaced the family hearth. When the computer
is not used for homework, E-commerce or business, friends or family
members often gather around it for games, looking up family-related
information such as potential vacation spots, and surfing the net.
Workplaces Observation should not be confined to the home com-

puter user. Office networksfrom the e-mail experience, to the trans-
fer of files, to the control of security and morecan benefit from close
study of flow and process as employees communicate with their cowork-
ers and with distant offices.
Researchers working with systems engineers before a new network
or new hardware is installed can help save organizations months of
frustration and expense. How are employees using the current system?
What are the problem points? What needs to be changed? What parts
should remain unchanged? Day-to-day observation of individual users
throughout the organization will provide valuable background for the
open-minded engineer assigned to upgrade the network.
Web site design and architecture have become more and more in demand
as companies use the Internet to sell, to communicate with customers,
and to communicate between departments. Here, too, close observation
of the stumbling blocks can point the way to improvements.
The computer is not the only potential target of observational research
in the office. Imagine a hidden camera facing the copier. Is the machine
providing angst-free output, or is the office staff frustrated with its
creaky service? What would it take to make the copier more useful for
the organization? Watch a time-coded video of just one days work, and
you will probably learn the answers to these questions.
Other busy places and activities such as the cafeteria, the mail room
and delivery process, the phone system, and the everyday traffic that
flows from one office to another are worthy of observation and study in
the interests of efficiency and organizational morale. It is also useful
for an outsider to observe and characterize what is commonly referred
to as the culture of the company. Its hard for those inside the organiza-
tion to step back and look at the company objectively, just as its difficult
for people to appraise and express their own salient characteristics. A
trained ethnographic researcher familiar with the various components
and values of an office culture can help a company understand and
present itself to the outside world.
5.3.4 Ethnographic Studies of Customer

Cultures
Ethnography represents a combination of several social science disci-
plines, and ethnographic researchers should be good at understand-
ing human behavior. In product and service development, what will be
successful depends on whether these products or services will satisfy
peoples needs in many aspects, such as function, social value, and self
image. By observing customers in real life, we can study the lifestyle
and cultural aspects of customers. This valuable information can help
to uncover customers unstated needs and generate clues for developing
new products.
204 Chapter Five
Culture represents the baseline of our experience as human beings

living in society. It is a broad concept encompassing the individual
worldviews, social rules, and interpersonal dynamics characterizing a
group of people in a particular time and place. Culture works through
religious beliefs, language, social institutions, and other group dynamics
to create a pattern, a taken-for-granted set of ideas and instincts that
delimit and define people in their social settings. Although cultures are
highly mutable, they have a conservative tendency and can sometimes
be changed only by circumstances of military conquest, or rapid tech-
nological change. People tend to defend their cultures and view alterna-
tives as different or even wrong.
Culture operates on both the material and immaterial levels of
human experience. It serves as the foundation for the behaviors, mean-
ings, and tools of all human collectivities. By cultural tools, we mean all
of the physical components of a groups way of life, its technology and
materials, as well as the fundamental rules, codes, and techniques for
accomplishing daily affairs. Cultural behaviors are the totality of activi-
ties associated with membership in a group, whether or not they are
practical, goal-oriented, sensate, or mystical. Cultural meanings refer
to the process of making sense: how we intellectually or emotionally
understand the purposes, implications, and associations that underlie
all of our behaviors and the tools we use in everyday life.
Some additional explanation will help clarify the definitions of these
terms and others that are useful in analyzing culture.
5.3.4.1 Cultural Behaviors The concept of cultural behavior can include

any human action outside of biologically imposed conditions (such as
sleep and digestion). Behaviors can be conscious and purposeful or sub-
conscious and non-goal-directed. Here are several ways of classifying
behaviors.
Rituals Rituals are patterned behaviors, usually performed without

thought, that are repeated by force of habit or belief. Examples of ritu-
als we may observe in the course of ethnographic practice may include
the order in which dishes are served in a meal or ones own particular
pattern of personal care and grooming in the morning.
Roles Through role behaviors, we enact a relationship between our-

selves and another person or a group. Being a boss and being a father are
examples of roles. Status refers to behavior that dramatizes differential
power or prestige in a setting. Thus, being a father implies more than just
particular role-related responsibilities, such as nurturance of children.
Particular brands are often purchased to reinforce the roles played
by consumers or their status in a group. A luxury brand icon on an
automobile or premium cigarette brands may be selected to demonstrate

a certain level of authority and power.
Ethnographic findings can often demonstrate opportunities for role-
related new products. A major manufacturer of paper goods discovered
that paper towels are mostly used by women in kitchens, but there are
similar needs in auto repair shops, machine shops, and so on. A new kind
of paper towel is developed with higher weight and strength, colored in
a blue denim, and it becomes a successful product. This is an example
of linking culture roles to product development.
Practical Activities and Goal-Oriented Behaviors These include

work, shopping, cooking, cleaning, or anything else that is directed to-
ward accomplishing actions necessary in everyday life. Carrying out
these daily tasks involves a range of habits, routines, skills, and styles
that have been learned or otherwise channeled through social influence.
Understanding the details of these practical activities and goal-oriented
behaviors are very important in developing and promoting relevant
products. For example, ethnographic researchers made detailed obser-
vations on how people are doing paperwork on paper, where peoples
eyes are moving, and where it is easy to get stuck. This research work
helped a great deal in developing all-electronic paperwork software.
Performances Performances are patterned behaviors that are

staged for the benefit of an observer. Baking cookies for the family, for
example, may be a response to the homemakers desire to perform the
role of good parent.
Play, Games, and Diversions Although the activities of play, games,

and diversions are believed to induce personal relaxation or bonding
between individuals, they may also serve functions with greater mean-
ing. They can also provide the basis for various social relationships
and either reinforce or negate status differences between individuals.
For example, when the boss invites a junior employee to join in a golf
outing, he or she is encouraging team commitment and inviting the
employee to join an inner circle.
5.3.4.2 Cultural Meanings Meanings are the ideas, emotions, or beliefs

that we attach to an object, a behavior, or another idea. Researchers
make sense of behaviors by trying to understand the meanings behind
them. The purpose or function of a behavior is not always self-evident.
Ethnographers typically have to probe the respondent and deeply un-
derstand the context if they wish to gain insights into the underlying
meanings attached to observed behaviors. The language of cultural pat-
terns that hold or convey meaning includes these concepts:
206 Chapter Five
Symbols Symbols are things that stand for something else, in-
cluding simple shapes or marks, such as the cross, which stands for
Christianity, the crucifixion, Christs suffering, salvation, or other reli-
gious ideas. Alternatively, symbols can be much less ethereal, such as
a corporate logo. Symbols connect an idea or thing to some underlying
system of meaning. Thus, for example, a Coca-Cola logo can stand for
youth, energy, refreshment, or tradition.
Signs Signs or markings point to something in the environment.

A highway sign, for example, directs you to your exit. The McDonalds
arches point consumers to a place where they can enjoy a consistent
and predictable family meal and share each others company.
Language, Jargon, and Slang Words are tools of communication

that have meanings beyond their basic communication value. They
serve as code words to distinguish in groups from out groups; words
provide a cipher or code to communicate meanings within closed groups
such as business organizations and religious or ethnic subcultures.
Marketing ethnographies provide opportunities to study language
in its natural context by listening in on what consumers say to each
other, in addition to what they say in response to a question posed by
a researcher. We can learn a great deal about beer brands or mobile
phones by studying the ways that young consumers talk about these
subjects during a night out with friends.
Beliefs and Values Beliefs and values are meaning filters or stan-
dards of truth for everyday life. They provide legitimacy to personal
behavior and help people tell right from wrong. Broad patterns of val-
uesfor example, the refusal to eat meat or a concern about the en-
vironmentcan be the basis for consumer behavior across a range of
categories.
Attitudes and Opinions Attitudes and opinions are expressions of

a point of view toward people, things, and events, and can range from
positive to neutral to negative. They can be intensely upheld or main-
tained with minimal salience. Attitudes and opinions are occasionally
rooted in deeply held principles and values, but, more likely, they are
situational and less character-based.
Interpretation Through interpretation, we make sense of things that

are communicated to us. Two people may see the same advertisement
or read the same novel but interpret them differently. Interpretation is
an essential component of the marketing communication process; its
what consumers hear when marketers talk.
Emotions and Feelings Inner conscious and unconscious emotions

and feelings normally occur reflexively; sometimes, we actively attach
them to people, behaviors, or ideas. In a marketing context, emotions
are consumers inner response to the external world.
Relationships Socially constructed ties between things or individu-

alsfor example, being part of a family, nation, or clubhave a tremen-
dous impact on our daily life, our work habits, and our beliefs. Many
brands are used to tie people into larger communities. Harley-Davidson
owners, for example, think of themselves as part of a larger community
of somewhat rebellious, adventure-seeking individuals.
5.3.4.3 Cultural Tools Tools are culturally produced devicesboth

objects and ideasthat expand human powers. Tools can help us live
life more comfortably or securely; they can help us perform our work,
defend ourselves, and organize our social ties. To a considerable degree,
the practice of marketing ethnography is directed toward evaluating
the marketplace tools that are available to consumers and, potentially,
creating new tools that expand consumer satisfaction, stimulate pro-
ductive efficiencies, and boost client profitability. In a broad ethno-
graphic context, tools may take a variety of forms.
Physical Space The environment, including our homes, workplaces,

shopping locations, and cities, is the most basic tool.
TechnologyLow to High The conventional meaning of tool is any-

thing that expands human powers, from paper clips to supercomputers.
Rules Culture rules provide a systematic framework for the opera-

tion of social organizations and processes: nation-states, professions,
commercial enterprises, and families. They guarantee that everyday
life proceeds with some degree of predictability and security. Rules can
be both formalwritten, codified, and elaboratedor informal, a set of
understandings that guide everyday behavior such as rules for queuing
for the commuter bus. A simple company policy statement, the rules of
baseball, and the U.S. Constitution are examples of rules that help us
carry on our lives.
An important component of ethnographic practice is decoding the rules
that are operating in the settings under study. In many cases, particu-
larly where informal rules are active, consumers may not be aware that
they are following any set of patterned behaviors until the basis of their
actions is questioned. In studying home barbecue grilling, for example,
we found that knowledgeable consumers generally followed a highly
prescriptive set of rules for the steps in tasks such as how meats should
208 Chapter Five
be prepared and how charcoal fires should be started and maintained.

Less successful chefs were less aware of the rules and, consequently, were
susceptible to switching to gas grills, a competing technology, because of
the larger tolerance for error that gas grilling offered.
Techniques Ways of getting things done are the how-tos of our

daily lives. Techniques work together with technologies and rules to
advance human potential. The techniques that people use to complete
the tasks of daily living in a culture may serve as clues to how well
people are satisfied with the technologies available to them. Many con-
sumers clean floors on their hands and knees in the belief that nothing
else produces their desired state of cleanliness. The ways in which we
complete everyday tasks may be creative and innovative; alternatively,
they may follow conventional rules quite ritualistically. Ethnographic
study proves its value to marketers when it can use findings about
techniques to structure innovations and incremental improvements in
product formulations and delivery.
A Case Study on Ethnographic Research on Customer Culture and

Product Development This case study is about Liptons effort in bring-
ing its brand and product into China. Lipton is the worlds biggest tea
brand, selling both hot and iced tea around the world. It started as a
British tea company, founded by Sir Thomas Lipton. The brand name
is now owned by Unilever. Lipton Tea represents around 10 percent of
the world market for tea.
China is one of the top tea producing and consuming countries of the
world. After the late 1970s, the Chinese market gradually opened to the
outside world. As a top tea company, Lipton wanted a part of this huge
Chinese tea market. Lipton is famous for its tea bags and black tea. In
many Chinese peoples minds, black tea is foreign. Because of Liptons
long history and brand image; it did not take long for Lipton to take
80% of the tea bag market.
However, black tea in bags accounts for less than 2% of the over-
all Chinese tea market. The predominate portion of the tea market
in China is taken by green tea and jasmine tea. In these two market
segments, most people in China think green tea and jasmine tea are
mostly Chinese; Lipton has nothing to do with it. Secondly, Lipton is
well known for its tea bags, and most Chinese people who consume
green tea or jasmine tea serve their tea in the form of loose leaves, as
In the West, most hot tea is made with tea bags (Figure 5.3). Tea bags
do have several clear advantages over loose leaves:
Tea bags are easier to dispose of.

Figure 5.3 Loose tea leaves and tea bags
When tea is served in a cup with loose tea leaves, some tea leaves tend
to float on top, so it is easy for people to end up swallowing tea leaves,
an uncomfortable experience. Tea bags prevent these problems.
Tea cups are easier to clean with tea bags than with loose tea
leaves.
However, in spite of these apparent advantages, Chinese people have

a hard time accepting green tea or jasmine tea in tea bags. One impor-
tant reason is that China has a unique and rich tea culture.
In China, tea is not only a popular drink but also serves many cul-
tural rituals:
As a sign of respect In Chinese society, the younger generation

always shows its respect to the older generation by offering a cup of
tea. Inviting and paying for their elders to go to restaurants for tea is a
traditional activity on holidays. In the past, people of lower rank served
tea to higher-ranking people. Today, as Chinese society becomes more
liberal, sometimes at home parents may pour a cup of tea for their
children, or a boss may even pour tea for subordinates at restaurants.
The lower-ranking person should not expect the higher-ranking person
to serve him or her tea on formal occasions, however.
For a family gathering When sons and daughters leave home
to work and get married, they may seldom visit their parents. As a
result, parents may seldom see their grandchildren. Going to restau-
rants and drinking tea, therefore, becomes an important activity for
family gatherings. Every Sunday, Chinese restaurants are crowded,
especially when people celebrate festivals. This phenomenon reflects
Chinese family values.
To apologize In Chinese culture, people make serious apologies to
others by pouring them tea. That is a sign of regret and submission.
The way Chinese tea is brewed, served, and consumed is another

sophisticated ritual. There are many different ways of brewing Chinese
210 Chapter Five
tea depending on variables like the formality of the occasion, the means
of the people preparing it, and the kind of tea being brewed.
The tea can be brewed in teapots or in cups. The formal way is to brew
tea in pots, and in one form is considered to be a kind of art (Figure 5.4):
1. Wash the teapot.

2. Put tea leaves into teapot.
3. Put hot water into the teapot, and wait about 1 to 2 minutes.
4. Pour out the tea in cups; this is the first infusion.
5. Put hot water in teapot again, and wait about 2 minutes.
6. Pour out the tea into cups; this is the second infusion.
7. Repeat step 3 and keep going for about two more infusions.
8. In Chinese tea culture, the first and second infusions are the best
shots; it is believed that the first infusion is the best in terms of aroma,
and the second infusion is the best in terms of flavor and color.
This process of brewing and drinking tea is an essential part of

Chinese tea culture. In many Chinese peoples minds, drinking tea is
not just for quenching thirst; it is a process for experiencing different
tea tastes, and for sharing the experience with others.
A less informal way of brewing tea is in tea cups. The procedure is
similar to that of teapots, and people enjoy watching the changes in
shapes of the tea leaves. In the beginning, the tea leaves are dry, but
after a couple of infusions, the tea leaves are in full blossom and float
to the top, and finally the tea leaves fall down to the bottom. Reading
tea leaves is a part of the tea-drinking process.
After studying the Chinese tea-drinking culture, Lipton researchers
realized that the tea bags advantages would not be enough to capture
the green tea and jasmine tea market, for the following reasons:
In the West, the tea brewing is very simple. First, you put a tea bag
in the cup; second, you pour hot water in; third, you wait a couple of
minutes, and then you drink. After one infusion, you throw away the
Figure 5.4 Teapots and the tea brewing process

tea bag. In this case, it is desirable to make the tea in such a way that
most of its contents, aroma, and flavor should be getting into the hot
water in one shot. Clearly, this design doesnt match the Chinese tea
culture.
In the West, the tea bag is brewed only once, in China, it will be brewed
many times, as the part of tea culture. So the tea bags used in Chinese
green tea or jasmine tea should be more sturdy and durable.
Tea bags are good for disposal and cleaning, but the tea bag takes
away all the cultural meanings of reading the tea leaves.
Based on this analysis, Lipton came up with three major design
improvements:
Design the teabag in such a way that the tea contents, aroma, and
flavor will be gradually dissolved into hot water, not in one shot.
Make the tea bag of a more sturdy and durable material, so it can
withstand multiple brewings and infusions.
Make the shape of the tea bag like a pyramid, so it is a tea bag, but
you can still read the tea leaves (Figure 5.5).
It is reported that this new design achieved some degree of success.
Figure 5.5 Pyramid tea bags and tea leaves reading

Chapter
VOC Data Processing

6
Chapters 4 and 5 discussed many effective methods of capturing VOC
information, and in a comprehensive VOC study, we might use several
different approaches, such as ethnographic observation, notes, taping,
customer surveys, and lead-user interviews. Regardless of the tech-
niques, though, we will have a lot of data of different kinds. In this
chapter, we will look at how to analyze various types of VOC data and
derive meaningful results.
6.1 Types of VOC Data

The kinds of data you get from VOC-capturing activities will depend on
the methods you use. In general, the ethnographical research methods
described in Chapter 5 will produce the following types of data:
Notes of open-ended interviews

Field observation notes
Video and audio recordings
This kind of data is generally not quantitative in nature.

If you use predesigned surveys (Chapter 4), you will get customers
answers to the survey questions. Most surveys use multiple-choice
questions, with some of the choices involving attribute data (Yes/No
answers) where there is no numerical value associated with the answer,
and some choices being numerical, such as ratings from 1 to 5. If the
survey results are intended to derive product functional requirements
(as discussed in Chapter 3), much more detailed information will need
to be collected, such as What jobs do you want this product to do? and
What outcomes do you want the product to achieve? In this case, we
213
214 Chapter Six
will get some very descriptive, detailed, technical data. In general, there
will be three kinds of data:
Words and notes
Attributes and numerical data
Data that can lead to design specifications
6.2 Analyzing VOC Data

The usual purposes of analyzing VOC data (as discussed in Chapters 13)
are as follows:
Getting sufficient and accurate input for the product devel-

opment process In this case, the purpose is to determine design
requirements for the product based on the VOC data. In Six Sigma
practice, a common outcome is critical-to-quality characteristics (CTQ),
or critical-to-satisfaction characteristics (CTS). CTQ and CTS are sets
of quantitative and actionable key product-quality measures that can
be readily transformed into design specifications. Sometimes only lim-
ited improvements are made to an existing product, so the scope of the
task is smaller than for new product design, but it is still necessary to
derive some CTQs. (The details of deriving CTQs from raw VOC data
will be discussed at the end of this chapter.)
Evaluating a products competitive position, customer value
rating, and so on The output of this kind of VOC data analysis is
usually a traditional statistical analysis report. In this case, we need
to analyze survey data both attributes and quantitative data.
6.2.1 Methods of Analyzing VOC Data

For data in the form of words and notes, the objective is to discover mean-
ings and patterns in the data. The affinity diagram or KJ method is very
good for this. Affinity diagrams will be discussed in the next section.
For attributes and numerical data, the objective is to quantify the
datas statistical measures and possibly conduct statistical inferences. We
will look at statistical analysis tools for this kind of data in Section 6.3.
The procedures for deriving and quantifying CTQs from data that can
lead to design specifications will be covered in Section 6.4.
6.2.2 Affinity DiagramKJ Method

The affinity diagram is also called the KJ Method, named after the
Japanese anthropologist, Jiro Kawakita, who developed a method of
establishing an orderly system from chaotic information (Kawakita
VOC Data Processing 215
1977, 1991). The KJ method is a four-step process for organizing and

summarizing a large amount of data (ideas, issues, solutions, problems)
into logical categories so that it is possible to understand the essence
of a problem or solution.
When using the KJ method, we write all relevant facts and informa-
tion on individual cards, which we collate, shuffle, spread out, and read
carefully. We shuffle the cards because these cards could come from
several sources. We then review, classify, and sort the cards based on
the similarity, affinity, and characteristics of the ideas. The four steps
of the KJ method are illustrated in Figure 6.1.
6.2.2.1 Step 1: Collecting Data and Preparing Notes First, prepare some
index cards or sticky notes for use. Then write each idea from the words
and notes raw data on one of the cards or notesone idea for each card.
6.2.2.2 Step 2: Sorting Ideas into Related Groups Sort the cards into
groups using the following process:
1. Put all the cards in one pile.

2. Draw one card at a time from this pile, and examine its content. If
the idea from this card is related to the idea from the card that you
drew before, put them together; these two cards belong to a group.
Step 1 Step 2
Step 3 Step 4
Figure 6.1 Four steps for building affinity diagrams (KJ method)
216 Chapter Six
3. Keep drawing the cards one at a time. Examine the card to see if the
idea from the card is similar to any existing group; if the answer is
yes, put this card into that group. If the answer is no, this card can
start a new group. Keep doing this step until all the cards are drawn
and all of the ideas are sorted into groups.
4. After all cards are drawn, re-examine all the groups. You can move
some cards around so that each groups ideas look more coherent. If
an idea seems equally applicable to two groups, create a duplicate
of that card and place one in each group.
5. It is possible for one card to stand alone and form a group.
6. The ideal grouping result should have the following features:
a. The ideas within a group should be closely related.
b. There should be significant differences between groups.
6.2.2.3 Step 3: Creating Header Cards for Each Group Create header
cards for each group. A header is a title that captures the essential link
among the ideas contained in a group of cards.
a. The header should be the best word or phrase that describes the
meaning of each group. The meaning of the header should stand
alone and be clear to outside readers without reading the contents
of the cards in the group.
b. During the process of creating headers, it is possible to regroup the
ideas so that the headers will have clear and better meanings. It is
also possible that a big group will be subdivided into several small
groups under different headers.
c. It may take several iterations to finalize the header process in order
to best capture the meaning of each group.
d. Clarify and finalize headers through consensus.
e. It is possible that hierarchical groups or multilevel groups will be
adopted.
Figure 6.2 gives an example of idea groups and headers.
6.2.2.4 Step 4: Writing Reports and Doing Further Analysis Affinity dia-
grams can help to reveal hidden groups and structures in a sea of frag-
mented notes and words, so you can see how everything fits together.
Once you have done this initial analysis, you will need to write a report
outlining the meaning of these groups, and this information can be used
in the product development process. One way to figure out the meaning
of the idea groups is to try to connect different groups with arrows or
lines that indicate relationships among the groups. Figure 6.3 shows
Product Customer Employee

strategy satisfaction development
Innovative Low price Teamwork

product features
Quick delivery
Motivated
Unique product employees
High quality
Mass
customization Personal
Responsive employee
technical growth
Low cost
support
Online help for

customers
Figure 6.2 Groups and headers in the affinity diagram
such an example; it connects different idea groups related to a hand-

held telecom product.
The structure of affinity diagrams can even be used to analyze whether
you have captured enough VOC information in your study. If you were
Portable
Feel close to Stay
iPod playstation
my child connected
Instant
w/my wife
message
Play video Talk on the my friends
Listen to my
games phone
music
Pursue my Communicate
Have fun Reply to
interests
emails
Read my quickly
My handheld
favorite my life...
magazines
Look
Stay organized successful
Never forget
a phone on the road
number Have the
latest Live upto
Make it to gadgets techy
Keep boss
meetings on image
happy
time
Figure 6.3 A connected arrow diagram

218 Chapter Six
expecting some idea groups to show up in the results, but they are absent,
this is an indication that you havent captured enough VOC information.
Figure 6.4 illustrates how this works.
In the real world, we may stick Post-It notes to a whiteboard to con-
struct an affinity diagram, as shown in Figure 6.5.
Example 6.1 (from Ramaswamy 1996) shows an affinity diagram
being used to study VOC data for a restaurant chain.
Something
should be here.
More cards
Figure 6.4 Discovering missing VOC information with the KJ method

Voice of the customer Interpretation Critical to Quality (CTQ)

Wait in the lobby for too
Short wait Waiting time
long
Sit in triage for a long Quick response
time from nurses and Response speed in
doctor in emergency care unit
emergency care
Too many errors in my Percentage of erroneous
Billing accuracy
bill invoices
It takes me several visits Accuracy in Percentage of erroneous
to get it right diagnosis diagnosis
Nurses dont care about
Patient Care Customer satisfaction rating
the patients
It takes so long to see the
Slow Process Expert utilization rate
expert
Figure 6.5 A finished affinity diagram
Example 6.1 Customer Needs for Restaurant Service VOC data can
be used to derive key CTS (Critical to Customer Satisfaction) metrics in
the restaurant business. The VOC data is often disorganized, nonspe-
cific, and nonquantitative, as shown in Table 6.1, which lists customer
needs for restaurant service.
TABLE 6.1 List of Customer Needs for Restaurant Service (Ramaswamy 1996)
1. Food tastes good 16. Dont want noisy atmosphere

2. Unusual items on menu 17. Want smoke-free atmosphere
3. Hot soup, cold ice cream 18. Wide choice of food
4. Feel full after the meal 19. Enough time to read menu
5. Dont feel overfull after meal 20. Can order quickly
6. Food looks appetizing 21. Know how long the wait for a table will be
7. Food courses arrive on table at 22. Food is healthy
right time
8. Dont feel hungry one hour 23. Menu items easy to understand
after meal
9. Clean restrooms 24. Prompt delivery after ordering
10. Clean tables 25. Get what was ordered
11. Clean plates and silverware 26. Get the correct bill
12. Clean, well-dressed employees 27. Billed as soon as meal is over
13. Light not too bright 28. Shouldnt feel rushed out of restaurant
14. Light not too dim 29. Make me feel at home
15. Shouldnt feel too crowded in 30. Order additional items quickly
space
220 Chapter Six
TABLE 6.1 List of Customer Needs for Restaurant Service

(Ramaswamy 1996) (Continued )
31. Errors and problems quickly 35. Waiter should be patient while ordering
resolved
32. Errors and problems 36. Fill water glass promptly without asking
satisfactorily resolved
33. Staff willing to answer questions 37. Polite, friendly staff
34. Greeted immediately on being 38. Short wait for table
seated
The affinity diagram or KJ method (Shigeru 1988) can be used to ana-

lyze and organize the VOC data into a CTS tree. A CTS tree is a refined
multilevel table of attributes that identify critical characteristics for
customer satisfaction. By using an affinity diagram, the two levels of
affinity groups described in Table 6.2 can be discovered.
TABLE 6.2 Two Levels of Affinity Groups for Restaurant Service (Ramaswamy 1996)
First Level Second Level Original Notes

Satisfying Tasty food Food tastes good
food
Balance of flavors
Hot soup, cold ice cream
Food looks appetizing
Food is healthy
Enough food Feel full after meal
Dont feel overfull after meal
Dont feel hungry one hour after meal
A lot of variety Wide choice of food
Unusual items on menu
Clean and Clean facility Clean restrooms

attractive
Clean tables
surroundings
Clean plates and silverware
Clean, well-dressed employees
Comfortable Light not too bright
atmosphere
Light not too dim
Shouldnt feel crowded in space
Dont want noisy atmosphere
Smoke-free atmosphere
TABLE 6.2 Two Levels of Affinity Groups for Restaurant Service

(Ramaswamy 1996) (Continued )
First Level Second Level Original Notes
Good service Friendly and Make me feel at home

knowledgeable staff
Staff willing to answer questions
Polite, friendly staff
Waiter should be patient while ordering
Menu items easy to understand
Shouldnt feel rushed out of restaurant
Fill water glass promptly without asking
Enough time to read menu
Quick and correct Short wait for table
service
Know how long the wait for a table will be
Can order quickly
Greeted immediately on being seated
Prompt delivery after ordering
Get what was ordered
Order additional items quickly
Food courses arrive on table at right time
Accurate billing Get the correct bill
Billed as soon as meal is over
Problems and Problems quickly resolved
complaints addressed
Problems satisfactorily resolved
effectively
6.3 Quantitative VOC Data Analysis

Quantitative VOC data can come from customer surveys, competitive
benchmarking (collecting competitors product specifications and per-
formance data), lead-user interviews, and so on. There are two types of
quantitative VOC data: attribute data and variable data.
Attribute data are either categorical or discrete. Examples of cat-
egorical data include gender, color, social class, and so on. Discrete data
means that the data values can only be integers (that is, no fractions or
decimal points), such as the number of defective units, paint chips per
unit, number of scratches, and so on.
Variable data, or continuous data values, can be any real number,
such as1.238, 78.45, 0.02875, and so on. Examples of continuous data
include length, volume, time, weight, and so on.
222 Chapter Six
Another important characteristic of data is its measurement scale.

Table 6.3 describes four measurement scales with the latter ones being
more useful for statistical analysis.
For data with different measurement scales, the two important statisti-
cal metricsthe central tendency and dispersionare different. Table 6.4
shows the statistical measures for data with different measurement scales.
(Commonly used statistical basics, measures, and analysis methods are
discussed in Chapter 11.)
TABLE 6.3 Four Measurement Scale Levels
Scale Description Example

Nominal Data consists of names or A parking lot has cars of the following
categories only. No ordering colors:
scheme is possible. Red 5
White 4
Blue 7
Black 6
Ordinal Data is arranged in some order A survey question:
(Ranking) but differences between values Ice cream is good for breakfast:
cannot be determined or are 1. Strongly disagree
meaningless. 2. Disagree
3. Neither agree nor disagree
4. Agree
5. Strongly agree
The difference between Strongly agree
(5) and Agree (4) does not have the same
meaning as the difference between
Disagree (2) and Strongly disagree (1).
Interval Data is arranged in order The temperature of three heated metal
o o o
and differences can be found. pieces is 300 C, 600 C, and 900 C
o
However, there is no inherent respectively. Three times 300 C is not
starting point and ratios are the same as 900oC in temperature
meaningless. measurement.
Ratio An extension of the interval Product A costs $400; product B costs
scale that includes an inherent $200.
zero starting point. Both the
difference between values and
the ratios are meaningful.
TABLE 6.4 Statistical Measures for Data with Different Measurement Scales
Measurement Scale Central Tendency Dispersion

Nominal Mode Bar chart
Ordinal Median Percentage distribution
Interval Arithmetic mean Standard deviation or range
Ratio Geometric or harmonic mean Coefficient of variation
Various statistical analyses can be performed on quantitative VOC

data, and it is important to take the measurement scale of the data into
consideration to make sure you are doing the right analysis.
6.4 Critical-to-Quality Characteristics (CTQ)

The ultimate goal of product development is to design a product that
customers really want. That is why it is important to spend so much
effort capturing the VOC information. However, we cannot design a good
product simply by using the raw VOC information. For example, suppose
you want to design a household power saw for yard work, and when you
summarize the findings from the VOC data, you have a list like this:
We want a saw that cuts wood quickly and easily; We want a saw that
does not get stuck easily; We want a saw that is easy to carry. When
you give these findings to design engineers, they will likely ask: What
do you mean by cutting wood quickly?; How quickly is quick enough?;
What kind of wood you are talking about? and so on. Clearly, this raw
VOC information wont give design engineers enough information to set
technical specifications.
The raw VOC data needs to be developed into clear, specific, quantita-
tive requirements in order to be really helpful in product development.
These kind of requirements are called critical-to-quality characteristics
(CTQs) in Six Sigma practice. CTQs are the product or service charac-
teristics that the customer considers important, and they are measur-
able characteristics whose performance standards or specification limits
must be met to satisfy customer requirements.
For example, light in weight may be one of the customers key
requirements for a power saw, but the statement light in weight is
not a CTQ, because it does not give either a performance standard or
specification limit. The statement the weight of power saw should be
no more than 3 kg is a CTQ because weight is a key performance factor
that is important to customers, and this statement gives a very specific
performance specification.
A typical CTQ is illustrated in Figure 6.6. It usually has four compo-
nents: characteristic, measure, target, and specification limits.
Characteristic
Weight of saw
(Y)
Pure weight in kg
Measure (Weight of electrical
wire not included)
Target 2 kg
Figure 6.6 Components of a CTQ and example

Specification LSL = 0 kg
Limit (s) USL = 3 kg
224 Chapter Six
Delivery Speed
Order meal Price Cheap
Figure 6.7 A CTQ tree

Taste Good
For a product or service item, there are usually several CTQs. For
example, a fast-food restaurant meal order process may have three
CTQs: delivery, price, and taste. For customer satisfaction, the delivery
needs to be fast, the cost needs to be low, and the taste needs to be good,
as illustrated in Figure 6.7. This is also called a CTQ tree. Of course,
for a complete CTQ tree, quantitative target and specification limits
are required. The delivery time has a numerical target (in minutes and
seconds), the cost target should be in dollars and cents, and the taste
can be quantified by using customer-rating scores.
The next question is how to derive CTQs from raw VOC data. We can
follow these steps:
1. Listen to statements and comments made by clients.

2. Determine what factors really drive a client to turn on or turn off.
What product features, service features, attributes, dimensions, and
characteristics (reliability, availability, and so on) cause a client to
be excited or angry?
3. Determine the best way to measure performance relative to these
factors and specifications.
4. Determine what targets we would need in order to meet the clients
factors and specifications.
5. Assess current performance relative to these factors and specifica-
tions.Table 6.5 shows the process of translating some raw VOC data
to CTQs. Each row of Table 6.5 shows a three-step transformation
from a raw VOC item to a CTQ. The VOC items in Table 6.5 are all
related to patients expectations when they are in the hospital.
The last question in CTQ development is how to derive target values

and specifications for CTQs. There are several ways to do this:
1. Customers inputs Though customers are usually not experts in

the product or service they are purchasing, you can still find their
specification limits. For the example illustrated in Figure 6.7, when
asked what speed and price they consider good, fast-food restau-
rant customers may answer: Faster than McDonalds, cheaper than
Wendys. In that case, you can determine the appropriate serving
TABLE 6.5 Translating VOC to CTQ for Hospital Patients
Voice of the Customer Interpretation Critical to Quality (CTQ)

Wait in the lobby for too Short Wait Waiting Time
long
Sit in Triage for a long Quick Response from Response Speed in
time Nurses and Doctor in Emergency Care Unit
Emergency Care
Too many errors in my Billing accuracy Percentage of Erroneous
bill Invoices
It takes me several Accuracy in Diagnosis Percentage of Erroneous
visits to get it right Diagnoses
Nurses dont care about Patient Care Customer satisfaction rating
the patients
It takes so long to see Slow Process Expert Utilization Rate
the expert
speed and price by finding out what is being done in McDonalds and
Wendys. If customers give very clear outcome statements, you can
derive specifications from this, for example, if customers who are
buying weed killer say that the weed should die within a day after
application.
2. Benchmarking Benchmarking is comparing your approach to
existing approaches or competitive designs. One easy way to do this
is to check your competitors specifications. You can also look for
existing ways of doing your function or for similar designs. Check
the Internet, the library, and patents.
3. System constraints Government regulations, safety concerns,
and physical limitations will all set limits on specifications.
Example 6.2 shows how to derive CTQs (adapted from Ramaswamy

1996)it is a continuation of Example 6.1.
Example 6.2 Deriving Quantitative CTQs for the Restaurant In

Example 6.1, we derived three groups of customer requirements from
the KJ method:
Satisfying food
Clean and attractive surroundings
Good service
The second-level groups explained the first-level groups and they are
the aggregated categories of original customer statements.
226 Chapter Six
To derive CTQs, however, we still need quantitative measures. For exam-

ple, short wait for a table is a specific customer requirement, but how
short is short? Five minutes or ten minutes? Similarly, just what does food
tastes good mean? Does the food taste good enough now? Also, we need to
know the relative importance of each CTS item. For example, which is more
important to the average customer: the taste of the food, or the nutritional
quality of the food? All this needed information can be found by various
means, such as using specially designed customer surveys, benchmarking
competitors, and using the mystery customer approach.
The mystery customer approach involves hiring either employees or
temporary helpers as mystery customers who are paid to visit com-
petitors facilities, as well as your own facility, and to fill out a number
of specially designed questionnaires after each visit. Some stopwatch
activities may also be included to record such measures as the ser-
vice waiting time, time to deliver the meal, and so on. With the help of
mystery customers, clear, quantitative measures can be developed. For
example, if the waiting time of the best competitor is no longer than five
minutes, and if waiting time is really important to customers, you need
to set a goal of waiting time being less than five minutes. If mystery
customers found that your competitor does offer better-tasting food,
you need to work on making your food taste better.
After processing the original affinity diagram, the following CTQ
measures are determined:
Degree of waiter patience
Degree of waiter responsiveness
Degree of waiter knowledge
Degree of waiter friendliness
Time between seating and menu delivery
Time between menu delivery and order taking
Time between ordering and meal delivery
Percentage of bills produced without errors
Mystery customer studies and competitive bench marking on several
competitors yielded the results in Table 6.6.
This example illustrates how you can develop two different kinds of
measurable performance metrics. One is an evaluation score type, such
as degree of waiter patience, and the other is a measurable perfor-
mance metric, such as the time between ordering and meal delivery.
Benchmarking competitors can help in designing performance speci-
fications, which are listed as our desirable performance. The perfor-
mance gaps on the performance metrics can be used to guide redesign
practice.
TABLE 6.6 Benchmarking Results for Restaurants
Time
Time between Time between between
Degree Seating Menu Ordering Percent of
of Waiter Degree of Degree of Degree of and Menu Delivery and and Meal Bill produced
Patience Responsiveness Knowledge Friendliness Delivery Order Taking Delivery Without Errors
Performance One grade One grade short Two grades One grade Achieved 3-minute gap 2-minute 5% gap
Gap short short short gap
Our Desired Exceptional Exceptional Exceptional Exceptional < 5 minutes < 5 minutes 10 minutes 95%
Performance
Our Restaurant Excellent Excellent Good Excellent < 5 minutes < 8 minutes 12 minutes 90%
Vive la France Good Exceptional Excellent Good < 10 minutes < 10 minutes 20 minutes >90%
Downtown Excellent Excellent Excellent Exceptional < 5 minutes < 8 minutes 15 minutes >90%
Steakhouse
Sarahs Seafood Good Good Good Good < 5 minutes <5 minutes 10.5 >92%
House minutes
VOC Data Processing

227
Chapter
Quality Function Deployment

7
(QFD)
Quality Function Deployment (QFD) is a planning tool that can be used

to translate customer needs and expectations into appropriate design
actions. QFD stresses problem prevention and places its emphasis on
achieving results in customer satisfaction, reducing design cycle time,
optimizing the allocation of resources, and ensuring that minimum
changes are required. Together with other quality tools and concepts,
QFD makes it possible to release products at Six Sigma level. Since the
customer defines quality, QFD develops customer and technical mea-
sures to identify areas for improvement. It translates customer needs
and expectations into design requirements by incorporating the voice
of the customer into all phases of the product development process,
through production and into the marketplace.
In the context of product development, the real value of QFD is its abil-
ity to direct the application of other quality tools to those design tasks
that will have the greatest impact on the teams ability to design a prod-
uct, service, or process that satisfies the needs and expectations of the
customers, both internal and external. QFD is best viewed as a planning
tool that relates a list of VOC parameters to functional design require-
ments. Customers define the product using their own expressions, which
usually do not carry any significant technical terminology. With the
application of QFD, possible relationships between quality character-
istics, as expressed by customers, and substitute quality requirements,
expressed in engineering terms, can be explored (Cohen 1988, 1995;
Clausing 1988). In the context of product development, these require-
ments are called Critical-To characteristics, which include subsets like
Critical-To-Quality (CTQs), Critical-To-Delivery (CTDs), and others.
229
230 Chapter Seven
The VOC items will be developed into a list of needs used later as input
to a relationship diagram, which is called QFDs House of Quality.
Correct market predictions are of little value if the requirements
cannot be incorporated into the product design at the right time.
Wresting market share away from a viable competitor is more difficult
than capturing market share by being the first producer into a market.
One major advantage of QFD is that it shortens the development cycle
by deploying customer expectations into product design in the early
stage, avoiding major design changes or redesigns.
The other significant advantage of QFD is increased customer satis-
faction. The team will employ marketing and product planning inputs
to incorporate customer expectations in the design process, production
planning, and in all functional departments. This will ensure that issues
are resolved, and the design is kept lean and focuses on those innova-
tions that are important to the customer.
Figure 7.1 shows that a company that is using QFD places more
emphasis on responding to problems early in the design cycle; it will con-
sume more resources in the early design stage to make sure the design
concept is sound. A company that does not use QFD will consume fewer
resources in the early design stage, but it will incur more resources later
in the design stage to fund design changes, and it usually will spend a
lot of money to fix problems after design release.
This chapter will explain how Quality Function Deployment works.
Specifically, I will start with QFD notations and the QFD matrix, and
then I will show how to fill a QFD matrix. Finally I will discuss the Kano
model and four phases of QFD.
Resource Usage
Resource Usage without QFD Unplanned
with QFD Resource Usage
after Design
Release
Resource Usage
Time
Figure 7.1 Effect of QFD on project resources

Quality Function Deployment (QFD) 231
7.1 History of QFD

QFD was created by Mitsubishi Heavy Industry at Kobe Shipyards in
the early 1970s. Stringent government requirements for military vessels
coupled with the large capital outlay per ship forced Kobe Shipyards
management to commit to upstream quality assurance. The Kobe engi-
neers drafted a matrix, which related all the government regulations,
critical design requirements, and customer requirements to technical
characteristics the company could control for and achieve. In addition,
the matrix also depicted the relative importance of each entry, making
it possible for important items to be identified and prioritized so they
would receive a greater share of the available resources.
Winning is contagious. Other companies adopted QFD in the mid-
1970s. For example, the Japanese automotive industry first applied QFD
to the rust problem. Since then, QFD usage has become a well-rooted
methodology in many American businesses. It has become familiar
because of its adopted commandment: Design it right the first time.
7.2 QFD Benefits, Requirements, and

Practicalities
The major benefit of QFD is customer satisfaction. QFD gives customers
what they want: Development cycles are shorter, failures and redesign
peaks during pre-launch (shown in Figure 7.1) can be avoided, and cus-
tomer demand knowledge is preserved and transferred to subsequent
design teams.
Certain things must be done before QFD can be implemented.
They include forming a multidisciplinary product development team,
which includes people from the marketing department, engineering
department, and others; spending time beforehand understanding
customer needs and expectations; and defining the product or service
in detail.
Many practical concerns must be addressed in order to implement
QFD successfully. For example, departments represented in the team
often dont tend to talk to one another. Also, problem prevention is not
traditionally rewarded as well as problem solving or fire fighting, so
convincing the management to allocate manpower and resource on a
problem-prevention methodology such as QFD is a challenging task.
Usually, it is easier to use QFD on an incremental design than a brand
new creative design, because both marketing and engineering people
would have a better idea of the voice of the customer for an existing
product. So it is better to develop your first QFD application as an
incremental design task.
232 Chapter Seven
HOWs
CTSs
HOWs
Functional
Attributes
Customer
WHATs
House of Requirements Design HOWs

Quality
#1 Parameters
WHATs House of Process
CTSs
Requirements
Quality Variables
Functional
WHATs
Prioritized House of
#2
CTSs Quality
Parameters
WHATs
House of
Design
Prioritized #3
Quality
Functions
#4
Prioritized
Parameters
Prioritized
Variables
Figure 7.2 Four phases of QFD in the product development process
7.3 QFD Methodology Overview

Quality Function Deployment is accomplished by multidisciplinary
product development teams using a series of charts to deploy critical cus-
tomer attributes throughout the phases of design development. QFD is
usually deployed in multiple phases. Figure 7.2 shows the typical four-
phase deployment in a product development application.
These four phases focus on planning CTS characteristics, functional
requirements, design parameters, and process variables.
In a typical service industry setting, QFD can be deployed in two
phases, focusing on planning CTS characteristics and operation vari-
ables, as illustrated in Figure 7.3.
QFD incorporates many techniques to make it easier to handle the
large numbers of functional requirements that might be encountered.
Applications involving 130 functions multiplied by 100 customer fea-
tures have been recorded (Hauser and Clausing 1988). One typical
HOWS
Operation
CTSs
Variables
Attributes
Customer
WHATS
WHATS
House of House of
CTSs
quality Quality
#1 #2
Prioritized Figure 7.3 Multiple phases of QFD in

Prioritized
Operation a service application
CTSs
Variables
grouping technique that may initially be used in a QFD study is the

affinity diagrama hierarchical grouping technique that consolidates
multiple unstructured ideas generated by the voice of the customer
(affinity diagrams are explained in Chapter 6). Affinity diagrams are
based on intuitive similarities from low-level stand-alone ideas to
arrangements of classes of ideas. This bundling of the customer fea-
tures is critical. It requires a cross-functional team that can brainstorm,
evaluate, and reconsider existing ideas in pursuit of identifying logical
(not necessarily optimum) groupings in order to organize the overall list
of needs into manageable classes.
Another technique is the tree diagram, which goes a step beyond the
affinity diagram. The tree diagram is used mainly to fill the gaps not
previously detected in order to achieve a more complete structure, which
in turn can lead to more ideas. Such expansion of ideas will allow the
structure to grow, but will also provide more insight into the voice of
the customer (Cohen 1988).
The House of Quality (see Figure 7.4) is the most important template
produced in a QFD study. The House of Quality looks like a house;
the center box of the house is a matrix diagram, which quantitatively
relates rows and columns. The rows on the left of the center usually rep-
resent the customer attributes, or what the customer wants (WHATs);
CTS Correlation
CTSs
(Hows)
Direction of Improvement
Customer Desirability
Attributes
Customer
(Whats)
Relationship
Planning Matrix
Matrix
Importance Rating
Competitive
Benchmarks
Targets & Limits
Figure 7.4 House of Quality

234 Chapter Seven
the columns on the top of the center box are usually refined product
requirements (HOWs), or how to design the product to satisfy the cus-
tomer attributes. The quantitative relationships between WHATs and
HOWs are the key information produced in a QFD study. The other
portions in the House of Quality provide other supporting information in
the QFD study. Employing the house will result in improved communica-
tion, planning, and design activity. This benefit also extends beyond the
QFD team to the whole organizationcustomer wants defined through
QFD can be applied to many similar products and form the basis of a
corporate memory on the subject of critical-to-satisfaction requirements
(CTSs). As a direct result of the use of QFD, customer intent, or what the
customer wants, will become the driver of the design process as well as
the catalyst for modifying product design solutions.
In Figure 7.4, we have the following components that constitute the
House of Quality (Cohen 1988):
Customer attributes (WHATs)

CTSs (HOWs)
Relationship matrix
Importance ratings (Customer Desirability Index and Technical
Importance Ratings)
Planning matrix
CTS correlation (HOWs correlation)
Targets and limits (HOW MUCH)
Competitive benchmarks (Customer Competitive Assessment and
Technical Competitive Assessment)
Other optional QFD chart extensions
7.3.1 Customer Attributes (WHATs)

Customer attributes are obtained from the VOC data collected by
surveys, claim data, warranty, and promotion campaigns. Usually
customers use fuzzy expressions in characterizing their needs and refer
simultaneously to many dimensions that need to be satisfiedaffinity
and tree diagrams may be used to complete the list of needs. I use
WHATs here because most QFD literature uses WHATs to represent
what customers want, so WHATs are really customer attributes. Most
of these WHATs are very general ideas that require more detailed
definition. For example, customers often say they want a stylish or
cool look when they purchase a product. Cool may be a very desirable
feature, but since it has different meanings to different people, it cannot
be acted upon directly.
Legal and safety requirements or other internal requirements, such

as company policy, or company internal safety requirements, are consid-
ered extensions of the WHATs. The WHATs can be characterized using
a Kano Model (discussed in Section 7.4).
7.3.2 CTSs (HOWs)

The HOWs are design features derived by the product development
team to answer the WHATs. Each of the initial WHATs needs opera-
tional definitionsthe objective is to determine a set of Critical-To-
Satisfaction requirements (CTSs) that can materialize the WHATs. This
process translates customer expectations into design criteria such as
speed, torque, and time to delivery.
For each WHAT, there should be one or more HOWs that describe a
means of attaining customer satisfaction. For example, a cool car can
be achieved through a new stylish body, improved seat design, more
leg room, and requirements for lower noise, harshness, and vibration
levels.
At this stage only overall requirements that can be measured and con-
trolled need to be determined. These substitute for the customer needs
and expectations and are traditionally known as Substitute Quality
Characteristics.
Teams should define the HOWs in a solution-neutral environment
and not be restricted by listing specific parts and processes. Just itemize
the means (the HOWs) whereby the list of WHATs can be realized. In
addition, each HOW will have some direction of goodness or improve-
ment as shown in the following illustration:
Maximize 1.0
Target 0.0
Minimize 1.0
For example, if one of the WHATs is the fuel efficiency for the automo-
bile, and one of the HOWs is the weight of the car, because less weight
in the car will save fuel consumption, then minimize or down arrow
is the direction of improvement. The circle represents the nominal best
target case.
7.3.3 Relationship Matrix

The process of relating WHATs to HOWs often becomes complicated
by the absence of one-to-one relationships, as some of the HOWs affect
more than one WHAT. In many cases, they adversely affect one another.
HOWs that could have an adverse effect on another customer want are
important. For example, cool and stylish are two of the WHATs that
236 Chapter Seven
a customer would want in a vehicle. The HOWs that support cool are
lower noise, increased roominess, and seat design requirements, among
others. These HOWs will also have some effect on stylish as well.
In the relationship matrix, the HOWs (in columns) and the WHATs
(in rows) form a grid. The relationship in every (WHAT, HOW) cell can
be displayed by placing a symbol representing the cause-and-effect rela-
tionship strength in that cell. When employees at the Kobe Shipyards
developed this matrix in 1972, they used the local horse racing symbols
as relationship matrix symbolssolid circles mean a strong relation-
ship, one circle means medium strength, and a triangle indicates a weak
relationship. Symbols are used instead of numbers because they can be
identified and interpreted easily and quickly. Different symbol notations
have been used, but this one is more common than others:
Standard 9-3-1
Strong 9.0
Moderate 3.0
Weak 1.0
After determining the strength of each (WHAT, HOW) relationship

and marking it in the cells, the product development team should take
the time to review the relationship matrix. For example, blank rows or
columns indicate gaps in either teams understanding or a deficiency
in fulfilling customer attributes. A blank row shows a need to develop
a HOW for the WHAT in that row, indicating a potentially unsatisfied
customer attribute. When a blank column exists, then one of the HOWs
does not impact any of the WHATs. Delivering that HOW may require
a new WHAT that has not been identified, or it might be a waste. The
relationship matrix gives the product development team the oppor-
tunity to revisit their work, leading to better planning and therefore
better results.
What is needed is a way of determining to what extent the CTS in a
column contributes to meeting the customer attribute in the row. This
requires a subjective weighing of the possible cause-effect relationships.
To rank the CTS and customer features in order, multiply the numeri-
cal value of the symbol representing the relationship by the Customer
Desirability Index. The Customer Desirability Index is the numerical
rating of the relative importance of a given customer attribute. For exam-
ple, if a WHAT is about safety, customers usually think it is very impor-
tant, so this will have a very high rating. This product, when summed over
all the customer features in the WHATs array, provides a measure to the
relative importance of such CTSs to the product development team and
is used as a planning index to allocate resources and efforts, comparing
the strength, importance, and interactions of these various relationships.
This importance rating is called the Technical Importance Rating.
7.3.4 Importance Ratings

Importance ratings are a relative measure indicating the importance of
each WHAT or HOW to the design. In QFD, there are two importance
ratings:
Customer Desirability Index This is obtained from VOC activi-

ties such as surveys, clinics, and so on, and is usually rated on a scale
from 1 (not important) to 5 (extremely important) as given in the
following illustration.
Importance
Extremely Important 5.0
Very Important 4.0
Somewhat Important 3.0
A little Important 2.0
Not Important 1.0
Technical Importance Ratings These are calculated as follows:

1. By convention, each symbol in the relationships matrix receives
a value representing the strength in the (WHAT, HOW) cell.
2. These values are multiplied by the Customer Desirability Index,
resulting in a numerical value for each symbol in the matrix.
3. The Technical Importance Rating for each HOW can then be found
by adding together the values of all the relationship symbols in
each column.
The technical importance ratings have no physical interpretation

and their value lies in their ranking relative to one another. They are
utilized to determine which HOWs are priorities and should receive the
most resources allocation.
7.3.5 Planning Matrix

The planning matrix is the area in the House of Quality (see Figure 7.4)
that is located on the right side of the relationship matrix. The planning
matrix is used to make comparisons of competitive performance and
identification of a benchmark in the context of ability to meet specific
customer needs. It is also used as a tool to set goals for improvement
using a ratio of performance (goal rating/current rating).
Hauser and Clausing (1988) view this matrix as a perceptual map
in trying to answer the following question: How can we change the
existing product or develop a new one to reflect customer intent, given
that the customer is more biased toward certain features? The product
of customer value, the targeted improvement ratio for the raw feature,
and the sales point, which is a measure of how the raw feature affects
238 Chapter Seven
sales, will provide a weighted measure of the relative importance of this

customer feature to be considered by the team.
7.3.6 CTS Correlation (HOWs Correlation)

Each cell in the roof is a measure of the possible correlation of two dif-
ferent HOWs. The use of this information improves the teams ability
to develop a systems perspective for the various HOWs under consider-
ation. The correlation matrix is one of the more commonly used optional
extensions over the original QFD developed by Kobe engineers.
Traditionally, the major task of the correlation matrix is to make trade-
off decisions by identifying the qualitative correlations between the vari-
ous HOWs. This is a very important function in the QFD because HOWs
are most often correlated. For example, a matrix contains quality and
cost. The design engineer is looking to decrease cost, but any improve-
ment in this aspect will have a negative effect on the quality. This is
called a negative correlation and it must be identified so that a trade-off
can be addressed.
Trade-offs are usually accomplished by revising the target values
(HOW MUCHs). These revisions are called realistic objectives. Using
the negative correlation example discussed previously, in order to
resolve the conflict between cost and quality, the cost objective would
be changed to a realistic objective.
In the correlation matrix, once again, symbols are used for ease of
reference to indicate the different levels of correlation with the follow-
ing scale:
Trade Offs
Synergy 1.0
Compromise 1.0
If one HOW directly supports another HOW, a positive correlation is

produced.
7.3.7 Targets and Limits (HOW MUCH)

For every HOW shown on the relationship matrix, a HOW MUCH
should be determined. The goal here is to quantify the customers needs
and expectations and create a target for the design team. The HOW
MUCHs also create a basis for assessing success, so HOWs should be
measurable. It is necessary to review the HOWs and develop a means
of quantification.
A target orientation to provide a visual indication of target type is
usually optional. In addition, the tolerance around targets needs to be
identified based on the company marketing strategy and contrasting it
with the best-in-class competitor.
7.3.8 Competitive Benchmarks

Competitive assessments are used to compare the competitions design
with the team design. There are two types of competitive assessments:
The Customer Competitive Assessment is found to the right of the

relationships matrix in the planning matrix. VOC activities (for exam-
ple, surveys) are used to rate the WHATs of the various designs in a
particular segment of the market.
The Technical Competitive Assessment is located below the relation-
ship matrix. It rates HOWs from several competitors from a technical
perspective.
7.4 Kano Model of Quality

In QFD, the voice-of-the-customer activities, such as market research,
provide the array of WHATs that represent the customer attributes.
These WHATs are spoken by the customer and are called performance
quality. For example, gas mileage of a car is a performance quality; it is
also the more, the better. However, more WHATs have to be addressed
than just those directly spoken by the customer. As Figure 7.5 shows,
there are also unspoken WHATs.
Unspoken WHATs are the basic quality features that the customer
automatically assumes will be in the design. Such WHATs are implied
Excitement
Quality
Performance
Satisfaction
Customer
Quality
Wow!
Degree of CTS
Achievement
..
of...
re Basic
Mo Unspoken Wants
ve Quality
Gi
Figure 7.5 Kano model of customer attributes

240 Chapter Seven
in the functional requirements of the design or assumed from historical

experience. For example, customers automatically expect their lawnmower
to cut grass to the specified level, but they wouldnt discuss it on a survey
unless they had trouble with one in the past. Unspoken wants have a
weird propertythey dont increase customer satisfaction, but if they are
not delivered, they have a strong negative effect on customer satisfaction.
Another group of unspoken WHATs can be categorized as excite-
ment qualityinnovations or delighters. These pleasant surprises
increase customer satisfaction in a nonlinear fashion. For example, in
the automotive industry, van owners were delighted by the second van
side door and by baby-seat anchor bolts.
Design features may change position on the Kano model over time. In
the 1990s, the second side door in a van was a pleasant surprise for cus-
tomers. Now, on most models, the second door is standard and is expected
to be installed without a specific request. The ideal product development
plan would include all three types of quality features: excitement qual-
ity (unspoken latent requirements), performance quality (spoken and
one-dimensional requirements), and basic quality (unspoken or assumed
requirements).
7.5 QFD Analysis

The completion of the first QFD House of Quality may give the product
development team a false impression that their job is completed. In real-
ity, they have simply created a tool that will guide future efforts toward
deploying the VOC information into the design. QFD matrix analysis in
every phase will identify design weaknesses, which must be dealt with
by a continuous improvement effort.
Here is a relatively simple procedure for analyzing the House of
Quality:
Blank or weak columns These indicate HOWs that dont strongly
relate to any customer attribute.
Blank or weak rows These are customer attributes that are not
being strongly addressed by a HOW.
Conflicts These are technical competitive assessments that are in
conflict with customer competitive assessments.
Significant points These are HOWs that relate to many customer
attributes, safety/regulatory, and internal company requirements.
Eye Opener opportunities These are areas where the teams
company and competitors are doing poorly. The QFD project team
should seize the opportunity to deliver on these sales points, which
may initially be treated as delighters in the Kano model.
Benchmarking This points out opportunities to incorporate the

competitors highly rated HOWs. The team can modify and incorpo-
rate these using benchmarking and not resort to creation.
7.6 Example 7.1 Information System Design

This example (from Yoruk 2003) deals with a QFD guided information
system design for a trucking company, Migros Inc. The key function of
this information system is to create management reports that capture
all important aspects of trucking company operation.
7.6.1 Ranking Customer Input

The first step in this QFD project was to survey the important aspects
of trucking company operation relevant to this information system.
The QFD project leader found three groups of people in the trucking
company relevant to this information system: partners (the trucking
companys customers), site managers, and data workers. A mini-QFD
project was done to determine the relative importance of these three
groups of people in giving customer input, and the findings are shown
in Figure 7.6.
The precise QFD scores calculation is shown in Table 7.1, where X1,
X2,..., Xm are the items in the QFD rows; x1, x2,, xm are the correspond-
ing scores for X1, X2,..., Xm; Y1, Y2,..., Yn are the items in the QFD rows;
and y1,y2,,ym are the corresponding scores for Y1, Y2,..., Yn.
Customers
HOWs
Data Workers
Weight
Site Manager
Partners
WHATs
Intensity of Use 5
CRITERIA
It Knowledge 1 MATRIX WEIGHTS

Influence on Use 5 Strong 9
Number of People 3 Medium 3

Weak 1
Influence on Project 5
2 34. 101
1 49. 147
3 15. 47
ABSOLUTE IMPORTANCE
RELATIVE IMPORTANCE
RANK
Figure 7.6 QFD table for weighting key customers

242 Chapter Seven
TABLE 7.1 QFD Scores Calculation
Items Y1 Y2 . Yj . Yn
Scores y1 y2 . yj . yn
X1 x1 p11 p12 p1j P1n
X2 x2 p21 p22 p2j P2n
.
Xi xi pi1 pi2 pij pin
.
Xm xm pm1 pm2 pmj pmn
The relationships between yjs and xis are given in the following equa-
tion (7-2):
m
y j = pij xi (7-2)
i=1
In Figure 7.6, the column Weight gives the scores for the relative
importance of customer selection criteria: 1 = lowest importance, 5=
highest importance. The items in weight are Xis, and the scores in the
weight column are corresponding xis. The three key customers, (part-
ners, site managers, and data workers) correspond to Yis. The row abso-
lute importance corresponds to the scores of key customers, that is, yis,
and they are calculated based on equation (7.2), as follows:
Partners absolute importance:
5
y1 = pi1 xi = 1 5 + 3 1 + 9 5 + 1 3 + 9 5 = 101
i=1
Site managers absolute importance:
5
y2 = pi 2 xi = 9 5 + 3 1 + 9 5 + 3 3 + 9 5 = 147
i=1
Data workers absolute importance:
5
y3 = pi3 xi = 3 5 + 1 5 + 9 3 + 9 5 = 47
i=1
The row relative importance is the normalized importance scores

as follows:
101
Partners relative importance = = 0.34 = 34%
101 + 147 + 47
147
Site managers relative importance = = 0.50 = 50%
101 + 147 + 47
47
Data workers relative importance = = 0.16 = 16%
101 + 147 + 47
Finally, based on the importance scores, the site manager is ranked
number 1, the partner is ranked number 2, and the data worker is
ranked number 3.
7.6.2 Ranking the Functional Requirements

In the next step, the QFD team rated the relative importance scores
of the following functional requirements by weighting the inputs from
these three key customers, that is, partners, site managers, and data
workers. The results are summarized in Table 7.2.
In Table 7.2, the relative importance score calculation is still based
on equation (7-2), for example:
The relative importance score of Fair truck run distribution =

0.34 3 + 0.5 9 + 0.16 0 = 5.5
The relative importance score of Prevent fuel theft =
0.34 9 + 0.5 9 + 0.16 0 = 7.6
and so on.
TABLE 7.2 Relative Importance Scores for Functional Requirements
Weighted
Partner Site Manager Data Workers Importance
Functional Requirements 34% 50% 16%
Fair truck run distribution 3 9 5.5
Prevent fuel theft 9 9 7.6
Prepare and print waybills 1 3 9 3.3
Minimize truck idle time 9 9 7.6
Calculate payment to drivers 1 3 3 2.3
Eliminate unreliable trucks 3 9 5.5
Prepare invoice data for 9 3 1 4.7
Migros
Prepare confirmation report 9 3 4.6
to Migros
Minimize fuel consumption 9 3 4.6
Follows runs to the market 3 9 1 5.7
Follow actual costs of runs 9 3 4.6
244 Chapter Seven
At the next step, the QFD team developed a QFD House of Quality
that linked the functional requirements of the information system with
the management report. This QFD is illustrated in Figure 7.7.
Again, in Figure 7.7, the importance score calculation is also based
on equation (7-2).
7.7 QFD Case Study 1: Global Commercial

Process Design
Project Objectives:
Design a global commercial process with high performance.
Project Problem Statement:

Sales cycle time (lead generation to full customer setup) exceeds 182
business days. Internal and external customer specifications range
from 1 to 72 business days.
Only 54% of customer service requests are closed by the commitment
date. The customers expect 100% of their service requests to be com-
pleted on time.
The current commercial processes are not standardized, and none of
the current processes are capable of standardization.
MANAGEMENT REPORTS
IMPORTANCE
Fuel Consumption
Distance Traveled
HOWs
Run Distribution
Hour Statistics
Market Visits
Expenditures
Truck Usage
Payments
Failures
WHATs
FUNCTIONAL REQUIREMENTS
Fair truck run distribution 5.5

Prepare & print waybills 7.6
Minimize truck idle time 3.3
TRUCK
Calculate payments to drivers 7.6
Eliminate unreliable trucks 2.3
Follow actual cost of runs 5.5
Prevent fuel theft 4.7
FUEL
Minimize fuel consumption 4.6
Prepare invoice data 4.6
CUSTOMER Aid in confirmation with Migros 5.7
Follow destinations of runs 4.6
12.4% 121
24.9% 243
15.0% 146
10.2% 100
14.0% 137
ABSOLUTE IMPORTANCE
3.6% 36
3.0% 29
9.1% 89
7.3% 72
RELATIVE IMPORTANCE
IMPORTANCE RANK
8
4
1
2
9
6
5
7
3
Figure 7.7 QFD house that links functional requirements to management report
Business Case:
There is no consistent, global process for selling to, setting up, and
servicing accounts.
Current sales and customer service information management systems
do not enable measurement of accuracy and timeliness on a global
basis.
Enterprise-wide customer care is a must be requirementfailure to
improve the process threatens growth and retention of the portfolio.
Project Goals:
Reduce prospecting cycle time from 16 to 5 business days.
Reduce discovery cycle time from 34 to 10 business days.
Reduce close-the-deal cycle time from 81 to 45 business days (all sales
metrics net of customer wait time).
Reduce setup cycle time from 51 to 12 business days.
Increase the percentage of service requests closed by commitment
date from 54% to 99.97%.
7.7.1 QFD Steps

The following are the QFD steps:
1. Identify the WHATs & HOWs and their relationship

2. Identify the HOWs and relationship matrix
7.7.1.1 Identify the WHATs & HOWs and Their Relationship The QFD proj-
ect team identifies customers and establishes customer wants, needs,
delights, and usage profiles. In addition, corporate, regulatory, and so-
cial requirements should be identified also. The value of this step is to
greatly improve the understanding and appreciation that team mem-
bers have for customer, corporate, regulatory, and social requirements.
The QFD project team, at this stage, should be expanded to include
market research. A market research professional might help the black
belt assume leadership during startup activities and perhaps later
remain an active participant as the team gains knowledge about cus-
tomer engagement methods. The team leader should put plans in place
to collaborate with identified organizations and/or employee relations
to define tasks and plans in support of the project, and to train team
members in customer processes, that is, forward-thinking methods such
as brainstorming, visioning, and conceptualizing.
The QFD project team should focus on the key customers in order to
optimize decisions around them, and try to include as many additional
customers as possible. The team should establish customer environmen-
tal conditions, customer usage and operating conditions; study customer
246 Chapter Seven
demographics and profiles; conduct customer performance evaluations;

and understand the performance of the competition. In addition, the
team should
Establish a rough definition of an ideal service.
Listen to the customer and capture wants and needs through inter-
views, focus groups, customer councils, field trials, field observations,
surveys, and so on.
Analyze customer complaints and assign satisfaction performance
ratings to attributes.
Acquire and rank these ratings with the Quality Function Deployment
(QFD) process.
Study all available information about the service including marketing
plans.
Create innovative ideas/delights and new wants by investigating
improved functions and cost of ownership, and by benchmarking the
competition to improve weak areas. Create new delights by matching
service functions with needs, experience, and customer beliefs.
The following WHATs are used:
Available Products
Professional Staff
Flexible Processes
Knowledgeable Staff
Easy to Use Products
Speedy Processes
Cost-Effective Products
Accuracy
7.7.1.2 Identify the HOWs and Relationship Matrix The purpose of this
step is to define a good product/process in terms of customer expecta-
tions; to benchmark projections, institutional knowledge, and interface
requirements; and to translate this information into CTSs. These will
then be used to plan an effective and efficient QFD project.
One of the major reasons for customer dissatisfaction and warranty
costs is that the design specifications do not adequately reflect customer
use of the product or process. Too many times the specification is written
after the design is completed, or it is simply a reflection of an old speci-
fication that was also inadequate. In addition, poorly planned designs
commonly do not allocate activities/resources in areas of importance
to customers and waste engineering resources by spending too much
time in activities that provide marginal value. Because missed cus-
tomer requirements are not targeted or checked in the design process,
procedures for handling field complaints for these items are likely to
be incomplete. Spending time overdesigning and overtesting items not
important to customers is a waste. Similarly, not spending development
time on areas important to customers is not only a missed opportunity,
but significant warranty costs are sure to follow.
In good design practice, time is spent up front understanding cus-
tomer wants, needs, and delights together with corporate and regulatory
requirements. This understanding is then translated into CTSs, which
then drive product and process design. The CTSs (HOWs) are given
in the following illustration, as well as, the relationship matrix to the
WHATs, shown in Figure 7.8.
Importance to the Customer
Meet Time Expectations
Know My Business & Offers
Save Money / Enhance Productivity
Do It Right the First Time
Consultative
Know Our Products & Processes
Talk to One Person
Answer Questions
Courteous
Adequate Follow-Up
A mapping begins by considering the high-level requirements for the

product or process. These are the true CTSs, which define what the
customer would like if the product or process were ideal. This consider-
ation of a product or process from a customer perspective must address
the requirements of higher-level systems, internal customers (such as
manufacturing, assembly, service, packaging, and safety), external cus-
tomers, and regulatory legislation. This diagram, which relates true
248 Chapter Seven
quality characteristics to substitute quality characteristics, is called a

relationship matrix.
The mapping of customer characteristics to CTS characteristics is
extremely valuable when done by the QFD team. A team typically begins
differing in opinion and sharing stories/experiences. When a QFD study
is completed, the entire team understands how product and process
characteristics that are detailed on drawings relate to functions that
are important to customers.
The full Phase I and II QFDs are given in Figures 7.87.10. An analy-
sis of Phase I follows in Section 7.7.3, and you are encouraged to perform
such an analysis on the other phases as well.
7.7.2 The HOWs Importance Calculation

Importance ratings are a relative comparison of the importance of each
WHAT or HOW to the quality of the design. The weight scales of 9, 3,
and 1 are used in the relationship matrix. Theses values are multiplied

Importance to the Customer

Adequate Follow-Up
Talk to One Person
Answer Questions
Maximize 1.0
Target 0.0
1.0
Consultative
Minimize
Courteous
10
1
8
9
Direction of Improvement 1
Available Products 1 2.0
Professional Staff 2 3.0
Flexible Processes 3 4.0
Knowledgeable Staff 4 4.0
Easy to Use Products 5 4.0
Speedy Processes 6 5.0
Cost-Effective Products 7 5.0
Accuracy 8 5.0
Figure 7.8 The WHATs, the HOWs, and relationship matrix

by the customer importance rating obtained from customer engage-

ment activities (like surveys) resulting in a numerical value. The HOWs
importance rating is summed by adding all values of all relationships.
For example, in Figure 7.8, the first HOW importance rating is cal-
culated as: 2.0 3.0 + 4.0 3.0 + 4.0 3.0 + 4.0 3.0 + 5.0 9.0 + 5.0
3.0 = 102. Other HOWs importance ratings can be calculated similarly.
7.7.3 Phase I QFD Diagnostics

From the Phase I QFD diagram illustrated in Figure 7.9, the following
problems are identified:
Weak WHATs The black belt needs to identify WHATs with only
weak or no relationships. Such situations represent a failure to
address a customer attribute. When this occurs, the company should
try to develop CTS(s) to address this WHAT, although the team may
sometimes discover that present technology cant satisfy the WHAT.
The QFD project team should resort to customer survey and assess-
ment for review and further understanding.
No such WHAT exists in our example. The closest to this situation
is Available Products in row #1 and Easy to Use Products in Row
#5. Row 1 and Row 5 were highlighted as the weakest WHATs, but
not weak enough to warrant the analysis in Figure 7.9. However, the
team is encouraged to strengthen this situation by a CTS with strong
relationship.
Weak HOWs The team needs to look for blank or weak HOWs
(all entries are inverted deltas). This situation occurs when CTSs
are included that dont really reflect the customer attributes being
addressed by the QFD. The QFD project team leader and his team may
consider eliminating the CTSs from further deployment if they do not
relate to basic quality or performance attributes in the Kano model.
In our example, The CTS Adequate Follow-Up is weak (rated 13 on
importance rating). However, the WHAT Easy to Use Products has
no strong relationship with any CTSs, and eliminating Adequate
Follow-Up may weaken the delivery of this WHAT even further.
Conflicts The DFSS team needs to look for cases where technical
benchmarking rates their product or service high but the customer
assessment is low. Misconception of customer attributes is the major root
cause. The team together with marketing can remedy the situation.
In our example, the Cost-Effective Products, a WHAT, is addressed
by many CTSs, including Save Money/Enhance Productivity. The cus-
tomer rates our design as Weak (rating 2), while the technical assess-
ment is rated the highest (rating is 4). Who is right? Conflicts may be
250
Target
Accuracy
Minimize
Maximize
Our Product
Target Values
Direction of
Improvement
Speedy Processes
Professional Staff
Flexible Processes
Available Products
1.0
0.0
1.0
Knowledgeable Staff
Easy to Use Products
Competitor 2: Dealers
Cost-Effective Products
Competitor 1: Rental Car

Competitor 4: Competitor B
Competitor 3: Competitor A
Importance of Product Attributes
Chapter Seven
Figure 7.9 Phase I QFD

8
7
6
5
4
3
Relative Importance of Product Attributes 2
1
9
8
7
6
5
4
3
2
1
1
1 Importance to the Customer
5.0
5.0
5.0
4.0
4.0
4.0
3.0
2.0
1 = 80% First Call Resolution 3 3 4 4 2 11.8 102.0 1 Meet Time Expectations
2 Discovery Document 100% Complete 4 3.5 2 3 3 14.9 129.0 2 Know My Business & Offers
3 Deliver 100% of Productivity Promises 3 3 3 2.5 4 13.9 120.0 3 Save Money / Enhance Productivity
4 = 80% First Call Resolution 3 3 3 3 3 14.6 126.0 4 Do It Right the First Time
5 80% of AMs / CSAs Score >= 2 on Competency Model 3 3 2 3 4 10.6 92.0 5 Consultative
6 80% of AMs / CSAs Score >=2 on Competency Model 0 0 0 0 4 15.4 133.0 6 Know Our Products & Processes
7 >= 80% First Call Resolution 3.5 3.5 4 4 3.5 7.6 66.0 7 Talk to One Person
8 >= 80% First Call Resolution 3.5 3.5 3.5 3.5 3.5 6.6 57.0 8 Answer Questions
9 80% of AMs / CSAs Score >= 2 on Competency Model 3 3 3 3 3 3.1 27.0 9 Courteous
10 100% Service Requests Closed by Promise Date 3 3 2.5 4 2 1.5 13.0 10 Adequate Follow-Up
Max = 5.0
Our Product
Synergy
Competitor 1: Rental
Compromise
Trade off
1.0
1.0
Min = 1.5
9
8
7
6
5
4
3
2
1
a result of a failure to understand the customer and must be resolved

prior to further progress.
Strengths By identifying the CTSs that contain the most 9 ratings,
the DFSS team can pinpoint which CTSs have a significant impact on
the total design. Changes in these characteristics will greatly affect
the design, and such effects propagate via the correlation matrix to
other CTSs, causing positive and negative implications.
The following CTSs are significant as implied by their importance rat-

ings and number of 9 ratings in their relationships to WHATs: Meet
the Expectations, Know My Business & Offers, Save Money/Enhance
Productivity, Do It Right the First Time, and Know Our Products &
Processes. By examining the Correlation Matrix in Figure 7.10, it is
clear that we have positive correlations throughtout except in the Do
It Right the First Time and Meet Time Expectations cells.
Eye Openers The DFSS team should look at customer attributes

where
1. Their design as well as their competitors design is performing
poorly.
2. The WHATs are performing poorly compared with their competi-
tors based on competitive benchmarking.
3. CTSs need further development in Phase II.
We can pencil Flexible Processes in the first category and Accuracy

and Easy To Use Products in the second category. The CTSs that
deliver these WHATs should receive the greatest attention as they rep-
resent potential pay-offs. For the WHATs where the competitors are
rated significantly higher than in our own design, it is recommended
that we should learn our competitors design and adopt a similar design
strategy. This saves design and research time.
The highest CTSs with the largest importance ratings are the most
important. For example, Know Our Products & Processes has the high-
est rating at 133. This rating is so high because it has three strong
relationships to the WHATs. The degree of difficulty is medium (rating
equal to 3) in the technical benchmarking. In addition, any CTS that
has negative or strong relationships with this CTS in the correlation
matrix should proceed to Phase II.
7.8 QFD Case Study 2: Yaesu Book Center

The Yaesu Book Center is a bookstore in Japan. When it first opened, the
store had few employees experienced in bookselling. Most of the business
was conducted by employees who had recently graduated from school.
252 Chapter Seven
Third House of Quality: Process Planning Matrix
Use of Standardized Documents and Tools
Relative Importance of Part Attributes

Updating of Customer Account Data
Importance of the Part Attributes

Close the Deal Cycle Time
% of Employees Trained
Prospecting Cycle Time

Discovery Cycle Time
Direction of
Setup Cycle Time

Improvement
Systems Uptime
Target Values
Maximize 1.0
Target 0.0
Minimize 1.0
1
3
First Call Resolution % 1 5103.3 15.8 = 80% First Call Resolution 1
% Svc Req Res by Promise Date 2 5004.0 15.5 100% of Service Requests Resolved by Promise Date 2
% Total Portfolio Reviewed / Year 3 4266.0 13.2 10% 3
% Discovery Document Complete 4 3618.0 11.2 100% 4
Sales Cycle Time 5 1911.0 5.9 60 Days 5
Customer Satisfaction Rating 6 3927.0 12.1 6
% AIWCs as >= 2 Competency Model 7 3159.0 9.8 80% 7
Average Speed of Answer 8 1278.0 4.0 80% of Calls Answered in < 24 Seconds 8
Losses Due to Price 9 1356.0 4.2 <10% 9
% CSAs >= 27 Call Coaching 10 2718.0 8.4 80% 10

24.9 647.7
2 Used 90% of the Time 22.7 590.3
18.6 483.3
17.1 443.7
7.8 202.9
89.6
89.6
53.2
Importance of Process Attributes 1

3.4
3.4
2.0
Relative Importance of Process Attribute 2

Standard 9-3-1
Target Values 3
4 95% System Uptime
Strong 9.0
Moderate 3.0
3 Nightly Update
Weak 1.0
5 10 Days
6 45 Days
7 12 Days
8 5 Days
1 100%
Figure 7.10 Phase II QFD
In spite of that, Yaesu Book Center attracted a great deal of attention

and was highly regarded by book lovers.
Yaesu Book Center has its own quality control (QC) circle. In the QC
circle, the area managers are also group leaders. The QC circle deter-
mined that the following three things are essential to satisfy customers
needs:
To have enough books available

To have enough product information
To provide enough service
The QC circle also found that they did not have enough information
to figure out how to accomplish these objectives. They did not know the
following:
Specific customer demands were not clear.

There were no specific quantitative measurements for customer
demands.
Product Trade-offs
Relative Importance of Product Attributes

Importance of Product Attributes

Competitor 1: Rental car

Direction of
Adequate Follow-Up
Improvement
Talk to One Person
Answer Questions
Target Values
Maximize 1.0
Consultative
Our Product
Courteous
Target 0.0
Minimize 1.0
10
1
8
Meet Time Expectations 1 102.0 11.8 2 4 4 3 3 = 80% First Call Resolution 1
Know My Business & Offers 2 129.0 14.9 3 3 2 3.5 4 Discovery Document 100% Complete 2
Save Money / Enhance Productivity 3 120.0 13.9 4 2.5 3 3 3 Deliver 100% of Productivity Promises 3
Do It Right First Time 4 126.0 14.6 3 3 3 3 3 = 80% First Call Resolution 4
80% of AMs / CSAs Score >= 2 on

Consultative 5 92.0 10.6 4 3 2 3 3 5
Competency Model

Know Our Products & Processes 6 133.0 15.4 3 0 0 0 0 6
Competency Model
Talk to One Person 7 66.0 7.6 3.5 4 4 3.5 3.5 >= 80% First Call Resolution 7
Answer Questions 8 57.0 6.6 3.5 3.5 3.5 3.5 3.5 >= 80% First Call Resolution 8

Courteous 9 27.0 3.1 3 3 3 3 3 9
Competency Model
Adequate Follow-Up 10 13.0 1.5 2 4 2.5 3 3 100% Service Requests Closed by Promise Date 10
2 11.8 102.0
3 14.9 129.0
4 13.9 120.0
3 14.6 126.0
3 15.4 133.0
92.0
66.0
57.0
27.0
13.0
Importance of Product Attributes 1

4 10.6
Relative Importance of Product Attributes 2

3.5 7.6
3.5 3.5 3.5 3.5 3.5 6.6
3.1
1.5
Trade Offs
Our Product 3
3
Synergy 1.0
2.5
Competitor 1: Rental Car 4

4
Compromise 1.0
2.5
Competitor 2: Dealers 5
4
3
3.5
3.5 3.5
Competitor 3: Competitor A 6
3
Competitor 4: Competitor B 7
3
Target Values 8
80% of AMs / CSAs Score >= 2 on Competency Model
100% Service Requests Closed by Promise Date

Deliver 100% of Productivity Promises
Discovery Document 100% Complete
>= 80% First Call Resolution
>= 80% First Call Resolution

= 80% First Call Resolution
= 80% First Call Resolution
10
1
6
7
Figure 7.11 Correlation
The relationship between the customers demands and Yaesu Book

Centers service product was not clear.
To solve these problems, VOC data was collected and a two-phase

QFD was conducted by the Yaesu Book Center to improve the bookstore
operation. This QFD study was conducted by performing the following
steps:
1. Determine customer attributes (WHATs)

2. Determine quality characteristics (HOWs)
3. Assign degree of importance to customer attributes
254 Chapter Seven
4. Determine operations items

5. Two-phase QFD analysis for Yaesu Book Center
7.8.1 Determine Customer Attributes

(WHATs)
First, a lot of information about customers demands was collected by
customer surveys and interviews. The raw customers demands were in
their own words. In a brainstorming session, this raw customer informa-
tion was translated into a set of better-defined customer attributes. The
following procedures were used in this translation process:
Vague comments were changed into precise expressions.

Comments expressed in negative conditions were changed into posi-
tive comments.
Comments were grouped into subcategories and similar subcategories
were combined.
Customer attributes were fitted into tree diagrams.
Figure 7.12 shows a partial tree diagram that organizes the customer
attributes in this case.
Customer
Satisfaction
Pleasantness Fast
Service Goods Information
Special Event Location, Facility Available Clean Publisher Market

Convenience
Information Atmosphere Inventory Books Information Information
Information board
Outside sign, show window
Reference corner
Search for books
Can get available book quickly

Offer rapid special order service
Free delivery
Fast delivery
Heat and air conditioning

Has lobby for reading, resting, and coffee shop
Has escalator and bathroom
Bright store interior
Has collections of books and queries

Has books not available in other stores
Has related books
Has best sellers
Has the books customers want
All books are displayed
Clean books
Figure 7.12 Customer attributes in Yaesu Book Center

First Pleasant
Fast
Level Atmosphere
Second Product Product

Service
Level Arrangement Knowledge
Time required to deliver

books within the store
Number of master
Degree of looking
for missing books
Number of books
Compliance rate
Progress in job
that got dirty
training plan
fire drills
Third
Level
Figure 7.13 Quality characteristics (HOWs)
7.8.2 Determine Quality Characteristics (HOWs)

Quality characteristics are extremely important, so the Yaesu Book
Center QC circle made an effort to identify them. A partial list of qual-
ity characteristics is shown in Figure 7.13. A three-level hierarchy of
quality characteristics was used.
7.8.3 Assign Degree of Importance to

Customer Attributes
The survey questionnaire in Table 7.3 was used to collect customer
importance ratings for each customer attribute, and it served as a basis
for determining importance ratings in the QFD study.
7.8.4 Determine Operations Items

In this case, operations items are what bookstore management and
employees are actually doing in their work. The QFD study should provide
guidelines for the best ways to do this work, and indicate what they didnt
do enough before, and how much they should do it now. These operation
items are organized in a tree diagram and illustrated in Figure 7.14.
7.8.5 2 Two-Phase QFD Analysis for Yaesu

Book Center
Based on the work in the first four steps, two QFD House of Quality
charts were developed for the Yaesu Book Center. The first relates cus-
tomer attributes to quality characteristics, and the second relates quality
characteristics to operation items. The partial listings of these two houses
of quality are illustrated in Figure 7.15 and Figure 7.16.
256
TABLE 7.3 Survey Questionnaire
How important are the following items in a bookstore? Please rank from 1 to 5.
Chapter Seven
1 4
Not important 2 3 Somewhat 5
Questions at all Not important Neither important Important
1. Has a good variety of best sellers 1 2 3 4 5
2. Scheduled date for availability of 1 2 3 4 5
out-of-stock books is clear
3. Has a good variety of art books 1 2 3 4 5
4. Has a variety of books on sociology, 1 2 3 4 5
literature, science, and history
.. 1 2 3 4 5
13. The store clerks look hard for books 1 2 3 4 5
for the customers
14. Can easily find books you want 1 2 3 4 5
15. Book classifications are easy to 1 2 3 4 5
understand
16. Attractive, easy-to-find book 1 2 3 4 5
displays
17. Books are always clean 1 2 3 4 5
After Sales During Sales Before Sales
Customer
Deliver Special Additional Merchandise Inspection
Data Analysis Service and
Orders Orders Control and Display
Sales
Delivery date table
Delivery
Checking special order status
Placing customer special orders
Ordering
Investigation on book storage
Looking for new resolution
Follow up investigation
Check stock
Analyzes results of special events
Special slip order analysis
Scheduls analysis
Sales slip analysis
Stocking
Fill in
Arranging shelves
Stocking shelves
Gives in-house information

Gets orders
Checks for cause of out of order
Searches
Answers questions
Rearranges stacks and fills the shelmes

Corrects wrong classifications
Checks fixtures
Inspects the facilities
Makes rounds in the store
Figure 7.14 Operation items
Quality characteristics
Goods
Service
Reception and response Environ-
Availability Orders
Importance Rating
to customers ment
Degree of satisfaction
Immediate response
Ratio of availability
Customer reception
Service satisfaction
Product knowledge
Clear classification
Prouerement time
Order lead time
Delivery sale
Waiting time
Damage rate
Easy to find
Cleanliness
Can tell the book is in stock 2.9 Lighting
Can tell why not in stock 6.0

Can tell if book is available
Can give date of availability 0.1
Can tell detailed book description
Can find related book 1.8
Offer information on the contents
Information list of book available

4.0
Has large variety and volume of books
Has book not available in other stores 1.3
Has newly published books 0.2

Customer Attributes
Has many speciality books
Easy to find books 4.3
Classification is easy to understand 5.8
Signs are easy to see 5.0
Display is easy to see 2.5
His clean books 2.5
Has product knowledge 3.4
Kind and polite 3.8
0.1
Assistance in looking for books 3.4
Figure 7.15 House of Quality Phase 1

258 Chapter Seven
Operation Items
Before sales During sales After sales
Analyzes result of special orders

Correcting wrong classifications
Rearrange stacks/fill shelves
Gives in-house information
Special order slip analysis

Sales slip analysis
Schedule analysis
Make rounds in ?
Checks fixtures
Inspects the ?
Get orders
Searches
?
?
Clear classification
Immediate response
Rate of availability
Degree of satisfaction
Easy to find
Quality Characteristics
Service satisfaction
Waiting time
Product knowledge
Customer reception
Order lead time
Delivery rate
Procurement time
Cleanliness
Lighting
Damage rate
Figure 7.16 House of Quality Phase 2
After constructing the two QFD charts, the QC circle found that there
was one major problem in the book center: The customer attribute book
classification is easy to understand was rated very high in customer
survey results, but by QFD analysis, the rating for this was not high
with current operation items. The following corrections were made:
1. If a book could fall into more than one category, the book would be
displayed in all these categories.
2. Point-of-purchase clerks were placed in the boundary areas between
book sections.
Chapter
Customer Value Creation

8
by Brand Development
Famous brand names make a big difference in the marketplace. T-shirts

can be made of exactly the same fabric, in the same style, and with the
same workmanship, but because they have different brand names, the
retail prices of these T-shirts are vastly different.
A good brand name brings extra value to the product and to the com-
pany that makes the product. McDonalds, Coca-Cola, Disney, Kodak,
and Sony are among the most globally recognized names in the world
(Kochan et al. 1997). The name recognition of these brands brings tre-
mendous marketplace success and high profitability. Brand development
has become one of the key sources of competitive advantage for compa-
nies worldwide. Brands are regarded as among the most valuable assets
owned by a company (Batra 1993, Davis 2000). Some brands are valued
so highly that companies have paid huge amounts to acquire the rights
to them. For example, in 1988 Philip Morris bought Kraft, the maker of
cheese products, for $12.9 billion, a sum that was four times the value
of the assets of the company (Murphy 1989). Companies that have good
brand names can sometimes defend their market positions for a long
time (Arnold 1992), as illustrated by Table 8.1.
What is a brand? Why do brand names have such power? What is the
importance of a brand name in developing a service product? These are
some of the questions that well look at in this chapter.
Merriam-Websters dictionary defines a brand as a mark made to
attest manufacture or quality or to designate ownership or a char-
acteristic or distinctive kind. Philip Koltler (1991) defines a brand as
a name, term, symbol, or design, or a combination of them, which is
intended to signify the goods or services of one seller or group of sellers
259
260 Chapter Eight
and to differentiate them from those of their competitors. Davis (2000)

defines a brand as an intangible but critical component of what a com-
pany stands for, and, as a set of promises, it implies trust, consistency,
and a defined set of expectations. The strongest brands in the world own
a place in the consumers mind, and when they are mentioned almost
everyone thinks of the same things. Mercedes-Benz stands for prestige
and the ultimate driving experience; Ralph Lauren stands for classic
looks, high status, and pride.
The strongest brands usually stand for superior functions, benefits,
and quality; without these, a brand will not be among the strongest
in the world. However, superior functions, benefits, and quality alone
will not make a strong brand. Does McDonalds offer much better
food than Burger King? Does Starbucks offer much better coffee than
Caribou Coffee? Probably notthe difference between the top brand
and second-tier brands is mostly psychological. Research in psychol-
ogy has shown that name recognition alone can result in more positive
feelings toward nearly everything, whether it is music, people, words,
or brands. In a study, respondents were asked to taste each of three
samples of peanut butter. One of these samples contained an unnamed
superior peanut butter (preferred in blind tests 70 percent of the time).
Another contained an inferior peanut butter (not preferred in taste
tests) that was labeled with a brand name known to the respondents
but neither purchased nor used by them before. Surprisingly, 73 per-
cent of respondents selected the inferior brand name option as being
the best-tasting peanut butter. These results clearly show the power
TABLE 8.1 Leading U.S. Brands from 1933 to 1990
Brand Market
Eastman Kodak Cameras/film
Del Monte Canned fruit
Wrigley Chewing gum
Nabisco Baked goods
Gillette Razors
Coca-Cola Soft drinks
Campbells Soup
Ivory Soap
Goodyear Tires
Customer Value Creation by Brand Development 261
of brand name recognition. Mere name recognition can make people

think that an inferior peanut butter tastes better than a better-tasting
peanut butter. Consumers psychology plays a very important role in
brand name strength, and Davis (2000) calls this psychological reaction
to brand names PATH, which is an acronym for promise, acceptance,
trust, and hope. A strong brand makes promise, acceptance, trust, and
hope tangible.
The benefits from strong brands are numerous; Davis (2000) lists the
following benefits:
Seventy-two percent of customers say that they will pay a 20 percent

premium for their brand of choice, relative to the closest competitive
brand. Fifty percent of customers will pay a 25 percent premium.
Forty percent of customers will pay up to 30 percent premium.
Twenty-five percent of customers state that price does not matter if
they are buying a brand that owns their loyalty.
Over 70 percent of customers want to use a brand to guide their purchase
decision and over 50 percent of purchases are actually brand-driven.
Peer recommendations influence almost 30 percent of all purchases
made today, so a good experience by one customer with your brand
may influence anothers purchase decision.
More than 50 percent of consumers believe a strong brand allows for
more successful new product introductions, and they are more willing
to try a new product from a preferred brand because of the implied
endorsement.
These benefits clearly indicate that strong brands do create tremen-

dous value for the companies that own the brands, so creating strong
brands should be an integral part of the product development strategy.
The making of a strong brand, usually called brand development, is a
very elaborate process; it involves the coordinated efforts of product
development, marketing, promotion, customer service, and corporate
leadership. Since the strength of the brand is directly related to the
value of service products, the people who work on product development
must understand the basics of brand development.
This chapter will cover the important aspects of brand development.
Section 8.1 will dive deep into the question of what is a brand?
Section 8.2 will discuss the brand development process. Section 8.4
will discuss the role of brand development in product development
practice.
262 Chapter Eight
8.1 The Anatomy of Brands

Strong brands have an almost magic power to add value to products
and inspire customer loyalty. It is very important to understand how
strong brands influence consumers opinions and what the essential
components of a strong brand are. In this section we will look at all
important aspects of brands.
8.1.1 Peoples Buying Behavior and Brands

According to Arnold (1992), the power of brands can be explained by
how people make buying decisions:
Most customers, especially consumers in mass markets, do not under-

stand a product or service as well as the company selling it. Most
customers have only superficial knowledge of the product or service,
and many are not even interested in the details.
Customers will perceive a product or service in their own terms. Since
customers usually have imperfect knowledge of a product or service,
they will base their opinions on the attributes most obvious to them.
For example, airline customers may rate the airline based on what
they see. If they see stains on the flip-down table, or washrooms that
arent perfectly clean, they may doubt the operation of the entire air-
line, including airplane engine maintenance. Customers may judge a
detergent by its smell, not by how well it cleans. Different customers
will notice different attributes; every customer has a personal view.
Customers perception often focuses on the benefits of the product
or servicewhat the product or service can do for the customer.
Customers will see different benefits. Some may see some functional
benefits, and some may be more interested in emotional benefits. For
example, some kids will want to buy cereal because there is a sports
star on the box (emotional benefit), and not because of the taste of
cereal itself (functional benefit).
Customer perceptions are not always conscious. If you ask a customer
why they chose a particular product or service, you will sometimes
get a rational answer, and sometimes not. Even if there is a rational
answer, it may not tell the whole story. Feelings about a product or
service may not be easily articulated, because these feelings are com-
plex and hard to explain, and they are sometimes subconscious and
irrational.
Because the relationship between customers and the things they buy
is complex, brand names become a shortcut for customers when they
are choosing products or services. When customers gradually develop
positive perceptions about a product or service, the thing they remember

is the brand name. These feelings and perceptions are often contagious,
and customers will spread them to friends, family members, and other
people whom they can easily influence, and this will create a snowball
effect. Watkins (1986) used the diagram in Figure 8.1 to illustrate the
model of customer choice.
Customers buying decisions involve a complex set of perceptions and
demands. Therefore, a brand should also address many elements of cus-
tomer perception and demand in order to be successful. The following
criteria for a successful brand are adapted from Arnold (1992):
The products or services associated with a brand must deliver the

functional benefits to meet market needs at least as well as the com-
petition. No product or service will survive in the long run if it does
not perform. A brand is not merely the creation of advertising and
packaging.
A first-of-its-kind product or service is a strong basis on which to build
a brand. However, the brand will not be successful in the long run if it
cannot satisfy customers. When competing in a crowded market with
similar products from existing brands, a newcomer has to provide
significant advantages in some area, whether functional, emotional,
or price, in order to compete effectively.
Besides the functional benefits, a brand has to offer intangible ben-
efits, such as emotional, belonging, prestige, style, and so on, in order
to shine. People will pay a high price for a top brand T-shirt for the
sake of pride, belonging, and prestige.
The benefits offered by a brand should be consistent with each other
and present a unified character or personality. If the benefits offered
from a brand are too confusing, or change from time to time, it will
drive the customer away because customers will often come to quick
and superficial conclusions when purchasing a product. Customers
form stereotypes about brands quickly, and if they like a brand, they
will stick with it. For example, both McDonalds and Chinese restau-
rants provide food and both have loyal customers based on customers
perception of the food. If a McDonalds restaurant started offering
Regular
Knowledge Brand Brand
Un- Brand Brand
Awareness Preference Repurchase Repeat
awareness Trial
of Features
Purchase
Figure 8.1 Model of customer choice

264 Chapter Eight
good Chinese food, it would send a very confusing signal to custom-

ers, and many customers might turn away. To maintain a brand, a
company must actively manage the personality of the brand to keep
it clear and consistent over time.
The benefits offered by a brand must be wanted by the customer. No
brand image, however clear and consistent, is of any use unless it
meets customer wants. If peoples wants change, the benefits offered
by the brand will have to change.
Brands have a magic power, and brand building is an important ele-

ment in customer value creation. To build a strong brand, we first need
to know what the essential elements of a brand are, and how these ele-
ments are related to each other. In the next two sections, we will look at
two essential elements of a brand: brand identity and brand equity.
8.1.2 Brand Identity

Customers perception about brands is much like peoples perceptions of
other people. A persons name is simply a symbol, and we form opinions
about a person based on our perceptions of what the person is good at,
what his or her personality is like, what the person looks like, what he
or she stands for, what his or her core values are, and so on. These char-
acteristics form a persons identity. Brand identity, according to Aaker
(1996), provides direction, purpose and meaning for the brand:
Brand identity is a unique set of brand associations that the brand strate-
gist aspires to create or maintain. These associations represent what the
brand stands for and imply a promise to customers from the organization
members. Brand identity should help establish a relationship between
the brand and the customer by generating a value proposition involving
functional, emotional, or self-expressive benefits.
There are several models that describe what brand identity is. Aaker
(1996) proposed a brand identity model based on four perspectives:
brand as product, brand as organization, brand as person, and brand
as symbol. Davis (2000) created the brand image model, which involves
two components: brand association and brand persona.
8.1.2.1 Aakers Brand Identity Model The Brand Identity section in

Figure 8.2 illustrates the framework of Aakers brand identity model.
This model describes brands from four perspectives: brand as prod-
uct, brand as organization, brand as person, and brand as symbol. A
brand need not actually employ all these four perspectives, however.
For brands that relate to larger corporations and its products, it is very
likely that all four perspectives will be employed.
The perception of brand identity in customers minds is the result

of the customers total experience with the products and services of
that brand. What perception customers have depends on the products/
services themselves, how customers are treated by the company, and
advertisements and promotions; ultimately, it depends on the companys
business strategy and business operation. To build a strong brand, the
company should carefully design a good brand identity for their prod-
ucts/services and make this ideal brand identity a reality.
The first step in creating a brand is to perform a strategic brand
analysis (as shown in Figure 8.3). A strategic brand analysis consists
of three components: customer analysis, competitor analysis, and self-
analysis. The purpose of this analysis is to provide a basis on which to
design an appropriate brand identity for the product/service offered by
the company. (The details of the strategic brand analysis will be dis-
cussed in Section 8.2.3.)
To design a good brand identity, we need to consider all four perspec-
tives, though we may not deploy all of them. The following sections
describe these four perspectives in detail.
Strategic Brand Analysis

Customer Analysis Competitor Analysis Self-Analysis
Trend Existing brand image
Brand image/identity
Motivation Brand heritage
Strength, strategies
Unmet needs Strengths/capabilities
Vulnerabilities
Segmentation Organization values
Brand Identity
Brand as Product Brand as Organization Brand as Person Brand as Symbol
1. Product scope 7. Organization attributes 9. Personality 11. Visual image
2. Product functions (Innovation, consumer 10. Brand-customer 12 Brand heritage
3. Quality/performance concern, trustworthiness, relationship
4. Uses etc.)
5. Users 8. Local vs. global
6. Country of Origin
Value Proposition Credibility

Functional Emotional Self-expressive
Support other brands
Benefits Benefits Benefits
BrandCustomer Relationship
Figure 8.2 Brand identity model (adapted from Aaker 1996)

266 Chapter Eight
The Brand as a Product: Product-Related Associations Product-

related associations are always an important part of brand identity,
because customers are buying the product. Product-related associations
have the following aspects:
The product scope This aspect deals with the product class the
brand is associated with. For example, McDonalds is associated with
the fast-food product class; VISA is associated with credit cards; Hertz
is associated with rental cars. A brand is unsuccessful if, when the
brand name is mentioned, most people dont know what product class
it is related to. When a nondominant brand name is mentioned, people
know what product class it is related to. In the case of a dominant
brand, when the product class is mentioned, most people will recall
the name of the brand. For example, when soft drink is mentioned,
the name Coca-Cola will at least be thought about. On the other hand,
when Faygo, a nondominant brand, is mentioned, some people will
recognize that it is a soft drink brand; however, when the soft drink
product class is mentioned, many people will not think of Faygo.
Product functions This aspect deals with what functional benefits
(and some emotional benefits) the product or product class can provide
to customers. How well this aspect of the product performs depends
on how well the customers needs (identified and unidentified) are
met. For example, McDonalds functional benefits include all their
breakfast and ordinary meal items, hamburgers, fries, soft drinks; fast
purchase cycle time (time from ordering to getting the food); happy
meals, toys, and playgrounds; unrivaled worldwide product consis-
tency; clean washrooms; and so on. McDonalds emotional benefits
include friendly service, being child-friendly, and so on.
Quality/performance This aspect deals with how well and how
consistently the functional and emotional benefits are provided.
For example, fast purchase cycle time is a key functional benefit for
McDonaldsthe speed and consistency of this cycle time is a matter of
quality and performance. McDonalds is famous for its unrivaled world-
wide product consistency, which is also a matter of quality. For products
with similar functional benefits, the performance depends on how well
these functional benefits are delivered. For example, Mercedes, Buick,
and Kia are all cars, but their performance levels are quite different.
Uses This aspect deals with the particular use or application associ-
ated with a brand. For example, Gatorade specializes in providing a
drink for athletes who want to maintain a high level of performance.
Users This aspect deals with the type of users targeted by a par-
ticular brand. For example, Motherhood Maternity targets pregnant
women and its products are for pregnant women.
Country of Origin Associating a brand with a country of origin

will add credibility if the country that the brand relates to has a good
reputation in this product area. For example, French fashion products
are better regarded than French electronic goods.
Product-related association is an important part of brand identity.

After all, people are buying products and the benefits are related to
product functions. However, if the brand identity is only associated with
product attributes, the brand will have serious limitations. Here are
some specific examples (Aaker 1996):
Failure in brand differentiation A product attribute can be
extremely important to customers, but if all brands are perceived
to be adequate on this attribute, it does not differentiate the brand.
For example, in the hotel business, cleanliness is always rated as
one of the most important attributes to customers. Thus it would be
appropriate for cleanliness to be a part of Hilton Hotels brand iden-
tity. However, because all hotels are expected to be clean, cleanliness
will not be a brand differentiator. Without brand differentiation, the
brand name will not stand out in customers minds when a product
is needed.
Easy to copy Product functional benefits are easy to copy. A brand
that relies on superior performance in functional attributes will even-
tually be beaten, because functional attributes are transparent, fixed
targets. If the brand name does not have a psychological dimension,
a low-cost competitor could easily nudge another brand out of the
marketplace.
Limitation on brand extension An overly strong association with
particular product attributes may limit the ability of a brand name
to extend to other fields. Both GM and Ford have strong financial
arms and they make a profit from them. However, they are mostly
related to automobile financial operations, due to the overwhelming
brand association with the auto industry. It is really hard for these
companies to expand into non-auto-related financial operations in
a big way.
Limitation on business strategy change An overly strong associ-
ation with particular product attributes will also limit a brands ability
to respond to changing markets. The Atkins brand is closely associated
with Atkins diet theory, and it is doing fine today. However, if the Atkins
diet theory goes out of favor, this brand will have a big problem.
Therefore, product-related association alone will not be adequate for

the constantly changing marketplaceit is important for a brand to
address other perspectives of brand identity.
268 Chapter Eight
The Brand as Organization The brand as an organization consists

of the brand attributes that feature the organization that manufac-
tures the brand as opposed to the products or services related to the
brand. Specifically, the brand-as-organization attributes should portray
the desirable and appropriate organization images that organization
leaders want the general public to see, such as corporate visions, beliefs,
core values, innovation, and so on.
Some organization-related attributes can also be related to the prod-
uct. For example, innovation and quality could also be related to product
design. However, when these attributes are related to an organization,
they usually mean different things, such as the culture and values that
create the innovation and quality. For example, Toyotas lean manu-
facturing principles are easy to copy from a procedure point of view;
however, it is Toyotas culture that is really difficult to imitate and copy,
which is why so many companies want to implement lean manufactur-
ing but fail to reach its full benefit (Liker 2004).
Organizational attributes are more enduring and more resistant to com-
petitive claims than are product attributes. First, it is much easier to copy
a product than to duplicate an organization with unique people, value, and
culture. Second, organizational attributes usually apply to a set of product
classes, and a competitor with a single product class is difficult to match.
For example, a low-end cloth manufacturer can imitate a Ralph Lauren
T-shirt easily; however, it is very difficult to actually create a brand with
equivalent value because Ralph Lauren has provided so many kinds of
products, as well as sponsorships for sporting events over many years.
Third, organizational attributes, such as a culture of innovation and qual-
ity, are difficult to measure and communicate, so it is difficult for competi-
tors to convince consumers that they have closed the perceived gap.
The Brand as Person The brand as a person is the brand attribute

that adds personality components, such as uplifting, youthful, energetic,
upscale, trustworthy, casual, and so on. This brand attribute is more
lively and personal than those related to products and organization.
The most important concept here is the brand personality, which can
be defined as the set of human characteristics associated with a given
brand. Thus it includes such characteristics as gender, age, and socioeco-
nomic class, as well as such classic human personality traits as warmth,
concern, and sentimentality.
A brand personality can create a strong brand in several ways. First,
it can help customers to express their own personality. For example, a
rich person may want to drive a Mercedes-Benz to show his affluence
and pride. Second, brand personality can be the basis of a relationship
between the customer and brand. For example, Harley-Davidson has
a personality of rugged, free-spirited, outdoor-oriented men, which the

buyer can use as an identifier for his own personality. Third, a brand
personality may help communicate product attributes.
The brand personality perceived by customers is created by many
factors, some of which are related to the product, and others that are
non-product-related. Table 8.2 summarizes these factors.
Product-related characteristics could be the primary drivers of a
brand personality. These characteristics include
Product class The product class can affect the personality. For
example, a bank or insurance company tends to assume a banker
personality (competent, serious, male, older, upper class).
Package The Huggies package always features healthy, happy
children.
Price Price is a complex factor in affecting brand personalityif
the price is low, it may attract low-end buyers and thus increase sales,
but it also gives the brand a cheap image. If the products related to
a brand have high performance and top quality, a higher price may
actually give an image of a prestigious brand.
Attributes Product attributes also affect brand personality. For
example, the strong flavor in Marlboro brand cigarettes gives them a
rugged male personality.
Non-product-related characteristics can also affect brand personality.

User image User image refers to either the profile of typical users
(the people who use the brand) or idealized users (as portrayed in
advertising and elsewhere). User image can be a powerful driver for
TABLE 8.2 Brand Personality Drivers (Aaker 1996)
Product-Related Characteristics Non-Product-Related Characteristics

Product class User image
Package Sponsorship
Price Symbol
Attributes Age
Ad style
Country of origin
Company image
CEO
Celebrity endorsers
270 Chapter Eight
brand image; for example, the Marlboro Man is the defining image of
Marlboros brand personality, a free-spirited, rugged man.
Sponsorship Sponsorship of particular events will influence a
brands personality. For example, Nauticas sponsorship of Olympic
swimming events gives its brand and swimwear a personality of
world-class swimmer.
Symbol Brand symbols can have a powerful influence on brand
personality because people see them every time they see an advertise-
ment or see the product. The Intel Inside symbol created a very strong
psychological impact for buyerswithout the Intel CPU, it wouldnt
be perceived as a good computer. Marlboro Country and the Maytag
repairman are among the most successful brand symbols that give
their brands desirable stereotypes in the customers minds.
Age How long a brand has been in the market (age) can affect its
personality. An old brand tends to have a traditional or reliable but
old-fashioned brand personality. A newcomer tends to have a younger
brand personality.
Ad style Advertisement style will affect brand personality as well.
The selection of actors/actresses and the selection of themes should
reflect the brand personality.
Country of origin Country of origin is also a very powerful opinion-
shaping factor; for example, a German brand might capture some
perceived characteristics of the German people (precise, serious, hard-
working, and so on).
Company image Company image will affect brand personality. The
brand personality should be consistent with company image.
CEO The personality of the CEO, such as Bill Gates of Microsoft,
influences peoples perception of the company and its products.
Celebrity endorsement Celebrity endorsements, such as
Michael Jordans endorsement of Gatorade, give the brand a similar
personalitya strong, thirsty athlete, in this case.
Value Proposition The purpose of products or product classes un-

der a brand name is to provide customers with benefits, and there are
many kinds of benefits. In Aakers brand identity model, illustrated
in Figure 8.3, three kinds of benefits are listed: functional benefits,
emotional benefits, and self-expressive benefits. The benefits that are
offered by each brand will be different, and Aaker (1996) call this the
value proposition. Specifically, a brand value proposition is a statement
of the functional, emotional, and self-expressive benefits offered by the
brand. An attractive brand value proposition should lead to market
share increase.
The concepts of functional, emotional, and self-expressive benefits

are explained as follows:
Functional benefits Functional benefits are the aggregated prod-

uct functions that a product provides to customers. The functional
benefits of a car include moving from point A to B, changing direction
and speed, providing a nice driving environment, providing styling,
and so on. Besides some must-have functions, a brand often pro-
vides some functional benefits that are special features. For example,
Volvo provides safety and durability; 7-Eleven provides convenience;
Nordstrom provides customer service.
Functional benefits are importantif a brand can dominate a key func-
tional benefit that customers really care about, it can dominate its prod-
uct class. The challenge is to select functional benefits that will ring a
bell with customers. However, just delivering this functional benefit is
not enough, because customers buy products based on perceived quality
and perceived functional superiority. Convincing customers that you
are truly the leader in a key functional area might be more challenging
than simply delivering these key functional benefits.
Emotional benefits When purchasing or using a particular brand
gives the customer a positive feeling, that brand is providing an emo-
tional benefit to customers. For example, you feel safe when you drive
a Volvo; you feel important when you shop at Nordstrom. A strong
brand value proposition often includes emotional benefits, on top of
functional benefits.
If a brand only has functional benefits, it is vulnerable because a
low-cost producer could duplicate the same functional benefits with
a lower price. Emotional benefits are more complex and much more
difficult to copy. Emotional benefits are intertwined with functional
benefits, and it is important to study the relationship between func-
tional attributes and emotional benefit.
Self-expressive benefits Customers can also use a brand to show
off. We call this the self-expressive benefit. For examples, some young-
sters want to buy fashions from the Gap to show off; successful busi-
ness people drive Lincoln, Lexus, or Mercedes-Benz.
Example 8.1 McDonalds Brand Identity McDonalds is one of the

most successful global brands. Its brand identity can be summarized
as follows:
Brand as product
Product scope: Fast food, childrens entertainment, eating spaces
272 Chapter Eight
Product functions:
Variety of fast-food items: hamburgers, Big Mac, Happy Meals, Egg
McMuffin, and so on
Service: Fast, accurate, friendly, and hassle-free
Cleanliness: Spotless in eating spaces, restrooms, and counter
Low prices
Quality/Performance: Consistent in temperature, taste, portion, layout,
and decoration; cleanliness in all its worldwide locations
User: Family and children are the focus, but the restaurant serves a
wide clientele.
Country of Origin: USA
Brand as Organization
Convenience: McDonalds is the most convenient quick-service restau-
rant; it is located close to where people live, work, and travel; it features
efficient, time-saving service; and it serves food that is easy to eat.
Brand Personality
Family-oriented, all-American, genuine, wholesome, cheerful, fun
Brand as Symbol
Logo: Golden Arches
Characters: Ronald McDonald; McDonalds doll and toys
Value Proposition
Functional benefits: Good-tasting burgers, fries, and drinks; extras such
as playgrounds, prizes, and games
Emotional benefits: Childrenfun via excitement of birthday parties;
joy from toys and playgrounds; the feeling of special family times
8.1.2.2 Daviss Brand Image Model Davis (2000) also developed a

brand image model. The brand image has two components: the brand
associations and the brand persona. Brand associations describe the
kinds of benefits the brand delivers to customers and the role it plays
in their lives. Brand persona is a description of the brand in terms of
human characteristics. The brand image model is a concise model for
the brand identity.
Brand Associations Brand associations relate to the product,

service, and organization aspects of the brand. Brand associations
describe a hierarchy of benefits that a brand provides to its customers.
Brand associations are best described by the brand value pyramid,

The Features and Attributes layer is at the bottom of this pyramid.
Here the features and attributes are the most essential product func-
tions, features, and quality levels that must be delivered to customers in
order to survive in the marketplace. The Benefits layer is in the middle
and includes the additional functional and/or emotional benefits that
the brand provides to its customers, beyond the features and attributes
covered in the bottom layer. The Beliefs and Values layer is at the top
of the pyramid, and this layer represents the emotional, spiritual, and
cultural values that are addressed by the brand.
Many brands will not be able to fill all the layers of the brand value
pyramid. If a brand cannot fill the bottom layer, it cannot even deliver
the most basic benefits to its customers for this kind of product, and
this brand will fail in the long run. If a brand can only fill the bottom
layer, it is a very marginal brandit is essentially a commodity, such
as raw cotton, or raw sugar. It is simply the equivalent of the no-brand
T-shirt in Figure 8.1, and its market survival is mostly dependent upon
low price. If a brand fills or somewhat fills the middle layer, it becomes
a surviving brand. It is better than just a commodity. The most powerful
brands should fill all the layers in the brand value pyramid.
Figure 8.4 shows the Ralph Lauren brand value pyramid. Ralph
Lauren has achieved the strongest brand status in its product class.
Many brands can deliver the features and attributes illustrated in
The emotional, spiritual Most meaningful and most

cultural values being difficult to imitate, but
addressed hardest to deliver
Beliefs
and
values
The functional or
emotional
benefits Benefits
provided to
customers
Easiest to
Features and/or deliver, but
functions that least meaningful
must be Features and attributes and most easily
delivered to imitated
customers
Figure 8.3 Brand value pyramid (Davis 2000)

274 Chapter Eight
Beliefs Proud
and Self-esteem
values Confident
Happiness
Fulfilled
Associated with
Benefits most respected designer
High status customers
feel like they are in style
Wide variety of clothes and

Features accessories
and Highest quality and durability
attributes Classic looking
Competitively priced
Serves multiple types of customers
Figure 8.4 Ralph Laurens brand value pyramid (Davis 2000)
Figure 8.4, that is, offering high-quality, durable, and classic-looking

clothes. But few of them can say their clothes allow their customers
to make a statement. Wearing Ralph Lauren clothes is like driving a
Mercedes-Benz in its appeal to social status. The psychological benefits
of this kind of brand name usually take years to evolve; they are difficult
to explain, and they are even more difficult to duplicate.
Brand Persona According to Davis (2000), brand persona is the set

of human characteristics that consumers associate with the brand,
such as personality, appearance, values, likes and dislikes, gender, size,
shape, ethnicity, intelligence, socioeconomic class, and education. The
brand persona brings the brand to life and customers subconsciously
decide if they want to be associated with this brand, just as they do
about whether they want to be associated with other people. If a brand
persona is unpopular or unattractive, that will affect the sale.
The brand persona here is very similar to the brand personality in
Aakers brand identity model.
Example 8.2 Brand Personas for Mail Services Table 8.3 lists the
brand personas of three major mail carriers: Federal Express (FedEx),
the U.S. Postal Service, and United Parcel Service (UPS).
TABLE 8.3 Brand Personas for Three Mail Carriers
FedEx U.S. Postal Service UPS

Male or Female Male Male
Young Old Middle-aged
Friendly Grumpy Friendly
Dependable Unreliable Inconsistent
High technology Low technology Unionized
Athletic Sluggish Evolving
Prompt Slow Brown uniforms
Energetic Unsophisticated Mediocre service
Professional Rigid Professional
Motivated Complacent International
Problem solvers Problem makers Problem solvers
8.1.3 Brand Equity

Brand equity is the set of assets (and liabilities) that is linked to a
brand name and symbol. The brand equity adds (or subtracts) the value
provided by a product or service to a firm and/or that firms customers
(Aaker 1996).
During the 1980s, a lot of research was done to define and estimate
the true value of brands to the competitive position of enterprises (Keller
1993, Aaker 1991, Farquhar 1989, Tauber 1988). There were two reasons
for this: The first reason was an accounting one, and it was to estimate
the value of brands more precisely for the balance sheet, especially in
cases of mergers, acquisitions, and divestitures. The second reason was
a strategy-based motivation to improve marketing productivity (Keller
1993).
Brand equity provides a mechanism for capturing the marketing
effects that are uniquely attributable to the brand (Keller 1993). Aakers
brand equity model (Aaker 1991: 269) is one of the best-known models
of brand equity. It is a tool for understanding the linkage between the
brand and the value it provides to the firm and its customers beyond
what is inherent in the functional attributes of the products and services.
The brand equity model defines five dimensions of value that the brand
provides the firm: brand loyalty, brand awareness, perceived quality,
brand associations, and other proprietary brand assets (see Figure 8.5).
Each of these dimensions is important in influencing the decision of
the customer during their purchasing process, and thus is a contributor
towards the viability of the enterprise. The strategic role of each of these
brand equity dimensions is described in the following sections.
276 Chapter Eight
Reduced marketing costs

Trade leverage
Attracting new customers
Brand Provides value to customer
Create awareness
Loyalty by enhancing customers:
Reassurance
Interpretation/processing
Time to respond to
of information
competitive threats
Confidence in the
Anchor to which other purchase decision
associations can be attached User satisfaction
Brand Familiarity-liking
Awareness Signal of
substance/commitment
Brand to be considered
Reason to buy
Brand Differentiate/position Provides value to firm by
Perceived enhancing:
Equity Price
Quality Efficiency and
Channel member interest
Extension effectiveness of marketing
programs
Help process/retrieve Brand loyalty
information Prices/Margins
Differentiate/position Brand extensions
Brand
Reason to buy Trade leverage
Associations
Creat positive Competitive advantage
attitude/feelings
Extensions
Other
Proprietary
Competitive advantage
Brand
Assets
Figure 8.5 Aakers brand equity model
8.1.3.1 Brand Loyalty In a competitive environment the ability of a

company to retain its existing customers is of key importance. Brand
loyalty is first a key factor influencing the repeat-buying behavior of
customers (Arnold 1986, Keller 1993), and it reduces vulnerability to
competitive actions in the marketplace. Secondly, it reduces the compa-
nys cost of doing business because it is more expensive for a business to
acquire new customers than to retain existing ones, especially when the
existing ones are satisfied with the brand (Aaker 1991). Thirdly, brand
loyalty can be powerful leverage for negotiating more favorable terms
in the distribution channels (Aaker 1991).
8.1.3.2 Brand Name Awareness Brand name awareness relates to the

likelihood that a brand name will come to mind, and the ease with
which it does so. Brand name awareness consists of two dimensions:
brand recall and brand recognition.
Brand recognition reflects a familiarity gained from past experience
with the brand (Aaker 1996). Studies have shown that people will
often buy a brand because they are familiar with it (Aaker 1991).
Brand recall refers to how strongly the brand comes to mind when the
consumer thinks about that product category, or the needs fulfilled by
that product category (Keller 1993).
Brand name awareness plays an important role in consumer decision

making because it allows the brand to be included in the consideration
set, which is a prerequisite for its eventual choice.
8.1.3.3 Perceived Quality Perceived quality is part of the human ex-

perience, and it is developed entirely from the perspective of the con-
sumer based on those product attributes that are important to them.
As a result, perceived quality may be different from the actual quality
of the product. Studies have shown that the customers perception of
quality has one of the greatest impacts on the financial performance of
a company (Buzzell et. al. 1987: 7, Jacobson 1987, Anderson et. al. 1994,
Aaker 1994). A study of 33 publicly traded companies over a four-year
period demonstrated that perceived quality had an impact on stock
return (Aaker 1996). Perceived quality is a key strategic variable for
many companies (Aaker 1996), and it is a key positioning dimension
for corporate brands.
8.1.3.4 Brand Associations Brand associations can be anything that

connects the customer to the brand (Aaker 2000). These associations
help determine the brand image with the customer and marketplace.
Brand associations can be hard attributes, related to specific percep-
tions of tangible functional attributes such as, speed, user-friendliness,
taste, and price. Brand associations can also be soft emotional attributes
like excitement, fun, trustworthiness, and ingenuity (Biel 1993). Apple
is an example of a brand with values that have resonated with those of
its target audience. Emphasizing values such as fun, excitement, inno-
vation, and humor (Kochan 1997), the company has succeeded in carv-
ing out a niche for itself in the highly competitive personal computer
marketplace (Levine 2003).
8.1.3.5 Other Proprietary Brand Assets Beyond their use as a tool in

achieving a competitive advantage, brands are also a financial asset to
a company. Successful brands can be traded, or can be used to increase
the valuation of a company during a corporate acquisition.
8.2 Brand Development

Strong brands can create tremendous value for the companies who
own them. Developing strong brands is one of the most important goals
for companies who strive to excel in the marketplace. The objective of
a brand development process is to create a brand that achieves and
maintains the intended position in the minds of customers within the
targeted market group. In other words, the brand development process
must create a brand image that reflects the brand identity defined by
the company, in the mind of customers.
278 Chapter Eight
Sender Media Receiver
Brand Identity
Signal Brand
Transmitted Image
Organization
Factors
(controllable) Market Factors
(Competition
and Noise,
Uncontrollable)
Figure 8.6 Brand identity transmission process
Creating this brand position in the mind of the customer involves

creating the brand identity, transmitting the brand image to custom-
ers, and having customers receive and accept this image, as illustrated
by Figure 8.6.
This process is influenced by a variety of factors, which include the
nature of the brand identity, organization factors, the communications
media, and market forces. In this section, we will first look at the factors
that influence brand development, and then at the several key steps in
the brand development process.
8.2.1 Key Factors in Brand Development

There are many factors that affect the brand development process. These
factors can be categorized into two classes: controllable factors and uncon-
trollable factors. The controllable factors are those over which the company
that owns the brand has some degree of control. The uncontrollable factors
are ones over which the company has little or no control. These control-
lable and uncontrollable factors are also illustrated in Figure 8.7.
8.2.1.1 Controllable Factors There are three classes of controllable

factors:
Brand identity
Marketing mix factors, or the four Ps: product, price, promotion, and
place
Time to market
In brand development, these factors can be used to shape a desirable

brand image and transmit this image to customers.
Brand Identity Brand identity has been thoroughly discussed in

Section 8.1.2. If we use Aakers brand identity model, there are four
perspectives: brand as product, brand as organization, brand as person-

ality, and brand as symbol. Clearly, the company that owns the brand
has full control of the brand symbol, a good degree of control in product
development, and relatively good control over organization behavior.
Brand personality takes years to form; it is more difficult to change.
In brand development, the company should design a desirable brand
identity based on thorough analysis of the marketplace, competitors,
and the relative strengths and weaknesses of the company itself, in
order to achieve the best brand image possible.
Product When customers are making their purchasing decisions,

they are not buying the brand symbol; they are buying the product.
Initially, customers may be influenced by advertisements or friends
advice to try a product from a certain brand. However, the product has
to perform to customers expectations. If the product performs equal
to or better than customers expectations, the perceived brand image
will be confirmed by their experience. This will trigger word-of-mouth
recommendations, sales will grow, and the positive brand image will
spread among more and more customers.
In order to accomplish this, it is very important that the actual prod-
uct characteristics, such as functions, performance, and quality levels,
be consistent with the brand identity. Therefore, product development
has to go hand-in-hand with brand identity design.
Price The role of price in brand development is quite interesting.

The brand price is related to the benefits that the brand provides, as
illustrated by Figure 8.7.
If a price is too high relative to the benefits that the brand provides,
the perceived value in customers minds will be low. Customers will
think the brand is overpriced. However, if the benefits that the brand
provides are high, but the price is low relative to the benefits, the cus-
tomers reaction can be quite complex. Theoretically, customers will be
happy to get more and spend less, but the perception that it is a cheapo
product may creep in, and this may undercut the brand image.
Usually, benefits are the main focus for brand identity creation. If you
have more benefits and customers are happy with them, it is always
Functional Emotional Self-Expressive Relative

Benefit Benefit Benefit Price
Value Proposition
Figure 8.7 The relationship between price and benefits of

a brand (Aaker 1996)
280 Chapter Eight
easy to raise the price to match the benefits. If the benefits are really
low in comparison with other brands, the price will have nowhere to go
but lower.
Promotion (Communication) The effectiveness of the communica-

tions campaign is a critical factor in creating the desired brand position
and image (Aaker 1996). The purpose of marketing programs, such as
advertising, is to transmit the brand image in order to increase brand
awareness and create a strong, favorable, and unique brand identity
in the customers mind. The strength of a brand image is greatly influ-
enced by communications in the marketing programs, particularly by
the effectiveness with which the brand identity is integrated into the
marketing programs (Keller 1993). Given the many choices of media
available, selecting the correct media mix to reach the targeted audi-
ence and ensuring that the message is integrated is a key factor in
creating a distinct brand image.
Place (Distribution) The distribution channels used to deliver the

products to customers are also very important factors for brand devel-
opment. It is important that the distribution channels deliver the goods
to the right customers. The key variables in the selection of distribution
channels include the type of channels, the number of outlets, the locations
of outlets and stock levels, and so on. The distribution process should be
synchronized with the promotion activities so that the desired customers
will know what the products are and where they are available. An efficient
distribution process will give customers the service benefits and reduce
the hassles in obtaining the products, thus enhancing the brand image.
Time to Market Brands that are able to position themselves first in

the minds of their customers have the best chance of achieving the high-
est brand awareness (Reis 1981). Being first in the marketplace will affect
the companys ability to create a strong brand. However, this factor is not
totally controllable, because it also depends on competitors activities.
8.2.1.2 Uncontrollable Factors The uncontrollable factors in brand

development are mostly marketing factors, as illustrated in Figure 8.7.
The demographics of the target population for the brand is an uncon-
trollable factor; it includes age group, income level, sex, marital status,
area of residence, location, social group membership, stage in life, and so
on. The variation in demographics makes the targeted customers very
nonhomogeneous, so the brand image that the company tries to com-
municate to customers will be perceived in different ways. For example,
a particular brand image might be very attractive for one age group, but
it might be unattractive for another age group.
Cultural factors are also important uncontrollable factors. People from

different cultural backgrounds, such as business clients from different
corporate cultures, will see things in different ways. These cultural fac-
tors may affect how customers perceive the brand image.
Competitive activity is clearly a very important uncontrollable factor.
The number of competitors and their relative strength will have a tre-
mendous impact on brand image. Therefore, it is critical to conduct a com-
petitors analysis at an early stage of the brand development process.
The customer is another uncontrollable factor, but the company can
try to understand the customers needs. The ability of a product or
service to meet the needs of its customers is a critical factor in creat-
ing the brand loyalty that will determine the success or failure of the
brand.
8.2.2 The Brand Development Process

Different processes have been proposed for brand development, but they
are fundamentally based on two paradigms. The first views the brand
development process as being closely associated with the development
and marketing of new products (Watkins 1986). In this paradigm, the
brand development process follows the classical brand management
model (Aaker 2000), and it consists of the following steps:
1. Market exploration
2. Preliminary financial analysis and screening
3. Formal business analysis and planning
4. Product and brand development
5. Product testing
6. Product launch
More recently, the second paradigm has elevated brand management

to a more strategic position within the organization, as defined by the
brand leadership model (Aaker 2000) and brand asset management
process (Davis 2000). The differences between the two brand develop-
ment paradigms are shown in Table 8.4.
The new paradigm for brand development and management is more
focused on managing the brand as a strategic asset of the company
(Arnold 1992, Davis 2000, Aaker 1996). The brand development process
defined by this paradigm generally consists of the following phases
(shown in Figure 8.8):
1. Brand strategy analysis phase

2. Brand strategy development phase
282 Chapter Eight
TABLE 8.4 Paradigms of the Brand Development Process (Adapted from Aaker 2000)
Classical Brand
Features Management Model Brand Leadership Model
Perspective Tactical and reactive Strategic and visionary
Brand Manager Status Less experienced, short Higher in the organization,
time horizon longer time horizon
Conceptual Model Brand image Brand equity
Focus Short-term financials Brand equity measures
ProductMarket Scope Single products and Multiple products and
markets markets
Brand Structures Simple Complex
Number of Brands Focus on single brands Category focusmultiple
brands
Country Scope Single country Global perspectives
Brand Managers Coordinator of limited Team leader of multiple
Communication Role options communication options
Communication Focus External/customer External and internal
Driver of Strategy Sales and share Brand identity
Brand Strategy Analysis Phase

Brand Vision Customer Self Competitor
Development Analysis Analysis Analysis
Brand Strategy Development Phase

Communications Develop Determine
Brand Product-Brand
& Marketing Channel Pricing
Positioning Coordination
Strategy Strategy Strategy
Brand Implementation Phase

Communication &
Public
Marketing the
Relations
Brand
Evaluation Phase
Measuring the
Return on Brand
Investment
Figure 8.8 Brand development process

3. Brand implementation phase

4. Brand evaluation phase
In the following sections, we will look at these four phases in detail.
8.2.3 Strategic Brand Analysis

The strategy analysis phase of brand development focuses on under-
standing the brands competitive position in the marketplace. It also
looks at the ability of the company to influence this positioning through
its capabilities and the attributes of its products and services. The steps
of the strategy analysis phase are brand vision development, customer
analysis, self-analysis, and competitor analysis.
8.2.3.1 Brand Vision Development

The brand vision is a short, succinct statement of what the brand intends
to become and to achieve at some point in the future, often stated in
competitive terms. Brand vision refers to intentions that are broad, all-
inclusive, and forward-thinking. It is the image that a business must
have of its goals before it sets out to reach them. It describes aspirations
for the future, without specifying the means that will be used to achieve
those desired ends.
This is IBMs brand vision:
At IBM, we strive to lead in the creation, development, and manufactur-
ing of the industrys most advanced information technologies, including
computer systems, software, networking systems, storage devices, and
microelectronics. We translate these advanced technologies into value
for our customers through professional solutions and service businesses
throughout the world.
According to Davis (Davis 2000), a good brand vision should have

four components: a statement of the overall goal of the brand, the target
market that the brand will pursue, the points of differentiation that
the brand will strive for, and the overall financial goals for which the
brand will be accountable. Davis gives an example of such a brand vision
statement:
Around the world, our eye care brand will stand for leadership in visual
care. Consumers and the professional channel will recognize us as the
industry leader in visual care solutions, including the best service, follow-
up, expertise, and product innovation. Our brand will help us fill one-third
of our stated financial growth gap through price premiums, better relation-
ships with the channel, and close-in brand extensions.
Developing a brand vision links the brand development process to the

strategic objectives of the company, which is an important step in ensur-
ing the necessary top management and financial commitment to the
284 Chapter Eight
brand (Davis 2000). During this step the strategic and financial goals
of the brand are defined, and the commitment of senior management
to the goals and objectives of the brand are obtained.
8.2.3.2 Customer Analysis The activities in the customer analysis

step are focused on understanding the trends, motivation, and unmet
needs of the various segments of the customer market (Aaker 1996).
This step creates an understanding of how the customer thinks and
acts and why and how they make a purchasing decision (Davis 2000).
The objectives of this step are fourfold (Aaker 1996):
1. To determine the functional, emotional, and self-expressive benefits

that the customer seeks when they buy and use the brand.
Customer surveys are usually needed to determine these benefits as
well as the relative importance of these benefits. The following three
questions are used to assess the functional benefits:
What functional benefits are relevant to customers?
What is the relative importance of each functional benefit?
Can benefit segments be identified?
Emotional and self-expressive benefits are subtle and thus more
difficult to detect than functional benefits.
2. To understand the different segments of the customer market and
their different needs, wants, and behaviors.
We need to develop a deep understanding of the needs of various
market segments. Because different market segments may have dif-
ferent needs in functional, emotional, and self-expressive benefits,
they may respond differently to a brand promotion program. There
are many possible segmentation schemes, but in the brand devel-
opment process, the major task is to find which segments are the
most attractive targets for the brand and are most relevant to brand
identity development. Therefore, the commonly used segmentation
schemes include segmentation by benefits sought, segmentation by
price sensitivity, and so on.
3. To understand trends occurring in the customer markets, so that the
current and future positioning of the brand can be better assessed.
By trends we mean the dynamics of the market, and how the demand
pattern will change. Analysis of market data, such as sales volume
trends and profitability prospects of the submarkets, may help to
understand market trends. Understanding the market trends provides
insight into changing motivations and emerging segments with stra-
tegic importance. For example, in the coffee market, from 1962 to 1993,
the sales of regular supermarket brands have been declining, but

gourmet coffee and coffeehouse sales are increasing. If a company is
in the coffee business, this information will certainly help to position
its brand in the future market.
4. To identify customer needs that current products do not meet.
Unmet needs are customer needs that are not met by existing prod-
ucts in the market. Unmet needs are strategically important because
they can represent opportunities for the company to make beneficial
moves in the market.
For example, Black & Decker organized a focus group of 50 power
tool owners. The executives of Black & Decker visited focus group
members homes and found several major unmet needs. One of the
problems with a cordless drill is that it can run out of battery power
before the job is done. Black & Decker responded by offering detach-
able battery packs that can be recharged quickly. Tool owners can
have several battery packs charged and when one battery runs out,
they can replace it with another one in no time and recharge the
run-out battery while they are working. Several design changes of
this kind gave Black & Deckers Quantum brand a core identity and
competitive advantage.
8.2.3.3 Self-Analysis The objective of the self-analysis step is for man-

agement to examine the strengths and weaknesses of their brands cur-
rent situation, so as to understand how the brand is positioned in the
marketplace and what circumstances contributed to its current posi-
tion (Arnold 1992). The areas analyzed during this step include
The current brand image, that is, what is the perception of the brand
in the marketplace?
This analysis can be done by using customer surveys (see Chapter 4).
The following types of questions should be included in the survey:
How is the companys brand perceived?
What associations are linked with the brand?
Why do customers like the brand? Why do customers not like the
brand?
How is the companys brand different from competing brands?
What benefits do customers get from the brand?
Has the companys brand changed over time? If yes, how?
For different market segments, does the companys brand image
differ? If yes, then how does it differ?
286 Chapter Eight
Does the companys brand have a personality? If yes, what is it?

What are the intangible attributes and benefits of the brand?
In assessing the current brand image, it is important that the cus-
tomer research should study not only the product-related attributes
but also non-product-related attributes, such as organizational associ-
ation, brand personality, brand customer relationships, and emotional
and self-expressive benefits.
The fundamental values of the brand, that is, what does the brand
stand for? Is it for fun, luxury, or an active lifestyle? What is the heri-
tage of the brand?
Besides studying the current image of the brand, it is important to
understand the heritage of the brand. Any surviving brand has some
reasons why it survived; it must have done something right. Many
brands get into trouble because they deviate from their heritage.
Arbitrary changes in brand identity may hurt a brand more than
they help. The answers to such questions as Who were the early
pioneers of the brand?; How did it originate?; and What was the
brand image when it first started? can help in understanding the
brand heritage.
Links and associations to other brands.
Some companies offer several brands of products or services. In this
case, a change in one brand position may affect other brands offered by
the company. Brand position decisions shouldnt be made in isolation
each brand should have a well-defined role, and all brands offered by
the company should work together in a synergistic manner.
The strengths and weaknesses of product and service offerings, and
the capabilities of the organization, that is, what is the organization
good at?
For a realistic brand strategy, the desired brand identity should be
supported by the organizational strength of the company. It is neces-
sary to find out the companys and the products strengths and weak-
nesses. We need to find out what the company is and is not good at,
and how and to what degree these weaknesses can be changed. If a
company is pursuing a goal that cannot be substantiated due to the
companys weakness, the goal will not be achieved.
8.2.3.4 Competitor Analysis The purpose of the competitor analysis

is to understand the current image, positioning, strengths, and weak-
nesses of competitive brands in the marketplace, as well as the pos-
sible future trajectories of these competitive brands (Aaker 1996). This
analysis will develop an understanding of the following aspects:
1. The customers perception of competitive brands.
This information is a fundamental input for brand identity determi-

nation. It indicates how customers perceive competitors brands, so it
provides a basis for what our brand has to do to differentiate itself
from competitors. It is important to find out what the functional
benefits to customers are, how these benefits compare with those of
the companys brand, the brand-customer relationship, and brand
personality.
There are two sources for this information. One is a customer survey
study on competitors brands. The other is competitors own informa-
tion; for example, competitors advertisements and advertisement
plans can provide clues to what kind of brand image they want cus-
tomers to perceive.
2. Previous changes in competitive brand positioning and the future
market positioning (strategic brand objectives) of competitors.
In the brand development process, it is important to consider not only
the current images of competing brands but also past changes and
possible future changes in these images. A thorough examination of
such changes can provide useful information about the reasons for
such changes and the reality of the competitive environment.
3. Strengths and vulnerabilities of the competitors.
Information on the strengths and vulnerabilities of competitors is
valuable input for a companys brand position. It is difficult and
costly to compete head to head with the strong points of competi-
tors. It is much easier to compete in areas where competitors are not
strong.
8.2.4 Brand Strategy Development

The purpose of the development phase is to develop the brand strategy.
During this phase, the positioning of the brand is developed. Also the
channel strategy, pricing strategy, and future extensions to the brand
are developed or aligned. The following sections describe these steps.
8.2.4.1 Brand Positioning Brand position is the part of the brand

identity and value proposition that is to be actively communicated to
the target audience and that demonstrates an advantage over compet-
ing brands (Aaker 1996). The effect of the positioning is to create the
necessary associations the customers will think of when they recall
the brand. Some examples of the associations of some well-positioned
brands are shown in Table 8.5.
288 Chapter Eight
TABLE 8.5 Examples of Top Brands and Their Attributes (Davis 2000)
Brand Attributes
Disney Family fun entertainment
Nordstrom Highest level of retail service
Saturn Your car company
FedEx Guaranteed overnight delivery
Wal-Mart Low prices and good values
Hallmark Caring shared
A brands position should be updated every three to five years, or

as often as needed to update the companys growth strategy. Senior
management has to lead in developing, updating, and implementing
the brand position.
Brand positioning defines the following aspects of the brands position
in the marketplace:
The target market segment
The business it provides to the market segment
Key benefits and points of differentiation between the companys
products and other brands in the marketplace
The contract of the brand with the market, which defines the brands
promises to its customers
Target Market Segment The company that owns the brand needs to
know who the intended customers for its brand are. Customer surveys
and self-study can be used to determine the target market segment.
The following types of questions should be asked:
Is the target market both identifiable and reachable?

Would the current customers be part of our target market segment?
Will the target market be attracted to our distinct brand identity?
If we have never served this market segment before, why do we want
to serve them now?
Business Provided to the Market Segment The company that owns

the brand needs to know what kinds of businesses they will and will
not operate in the target market segment. Customer surveys and self-
study can be used to determine this. The following types of questions
should be asked:
What is the category, industry, or business that we compete in?

How has this changed over time?
Will the marketplace value and believe our participation in this
business?
Key Benefits and Points of Differentiation The company that owns

the brand needs to know what differentiates its brand from other
brands in the target market segment, as well as what the key benefits
of its brand are to customers. Customer surveys and self-study can
be used to determine this. The following types of questions should be
asked:
Are our key benefits important to customers?

Can we deliver these benefits satisfactorily?
What are our key points of difference from other brands?
Can we own these key points of difference over time?
Are we competing at the Features and Attributes layer, at the Benefits
layer, or at the Beliefs and Values layer (see Figure 8.4)?
The Contract of the Brand with the Market The contract of the
brand with the market is also called a brand contract (Davis 2000). It
is a list of all promises the brand makes to customers, in terms of its
products and service quality. Such a contract is executed internally, but
it is defined and validated externally by the marketplace.
The brand contract is derived by analyzing several things:
The current promises of the brand made to the marketplace

Positive and negative feedback from customers regarding current
promises
The results of the brand position analysis
Example 8.3 Starbucks Implicit Brand Contract Starbucks prom-

ises to do the following:
1. Provide the highest quality coffee available on the market today

2. Offer customers a wide variety of coffee options as well as comple-
mentary food and beverage items
3. Have an atmosphere that is warm, friendly, homelike, and appropri-
ate for having conversation with a good friend or reading a book
290 Chapter Eight
4. Recognize that visiting Starbucks is as much about the experience

of drinking coffee as it is about coffee itself
5. Have employees who are friendly, courteous, outgoing, helpful, knowl-
edgeable, and quick to fill customer orders
6. Provide customers the same experience at any one of the several
thousand Starbucks worldwide
7. Stay current with the times, meet customer needs, and help custom-
ers create the Starbucks experience on their own terms
8. Provide customers with an environmentally friendly establishment
9. Educate customers on the different types of coffee offered
8.2.4.2 Five Principles of Effective Brand Positioning Davis (2000) pro-

posed five principles of effective brand positioning:
Value The proposed brand position should provide the targeted cus-
tomers with values that are superior to those of competitors. It should
provide functional, emotional, and self-expressive benefits that a wide
range of customers will appreciate and be willing to pay premium
prices for.
Uniqueness The proposed brand position should have some unique
attributes that no other competitors can deliver. These unique attri-
butes should be important to customers and be appreciated by them.
The uniqueness should make the companys brand stand out in the
crowd.
Credibility The proposed brand position should be implementable
in a credible manner, and the companys effort should be able to make
customers believe that all the promises will be met. The brand position
should be in line with customers perception of the companys ability.
Sustainability Once a proposed brand position is implemented, it
is desirable that this brand position last as long as it can. Changing
brand positioning involves a lot of investment and frequent change
reduces the credibility of the brand. A good brand position should be
difficult for competitors to copy and should meet customers changing
needs for a long time.
Fit The brand position should fit the companys objectives and
culture.
Example 8.4 Brand Positions of Several Bookstores Borders, Crown

Books, Barnes & Noble, and Amazon.com are four dominant booksell-
ers in the United States. The following table lists their unique brand
positions.
Target Market
Company Segment Business Provided Point of Difference
Borders Individuals looking Books, music, Fun place to go
for a community multimedia, and
meeting place online
Crown Books Price-sensitive Bookstores Discount pricing
individuals, strip
mall shoppers
Barnes & Noble Individuals looking Books, music, Library-like setting
for a quiet gathering multimedia, and
place online
Amazon.com Individuals who are Online books, music, Personalized online
active on the Internet and many other service, huge
and shop online items variety
8.2.4.3 Product-Brand Coordination Customers are buying products

or services, and a brand is a symbol and shortcut that helps customers
select the products they need. The customers total experience with the
products, including purchasing, consuming, and servicing, has to be in
tune with the brand position. For example, if a hospitals brand position
is to be the premier hospital that provides you with an attentive team
of caring experts working together for the highest level of professional
care, every word, such as premier, attentive, team, caring, and so on,
should have concrete actions behind it. Table 8.7 provides an example
of a well-coordinated brand and product combination.
These are the key issues in ensuring product brand coordination:
The brand development team should include product development

people.
Key product development professionals should learn the basics in
brand development and management.
Brand positioning and product development should go hand in hand.
8.2.4.4 Communications and Marketing Strategy Marketing deter-

mines what in the brands positioning will be communicated, and how
it will be communicated to the marketplace (Levine 2003). There are
numerous vehicles for communicating the brand to the marketplace,
and these include
Advertising
Internet
Public relationships
Trade and sales promotions
292 Chapter Eight
TABLE 8.6 Product-Brand Coordination of a Hospital
Brand Position Product (Patient Treatment) Attributes

Premier hospital Excellence in all performance metrics
Staffed with first-class doctors, administrators, nurses
Excellent infrastructure, first-class equipment
Modern appearance, spotless, well-organized
Attentive team Reduction of patients waiting time to industrys best
Reduction of unneeded paperwork, tests
Clearly explained treatment plan, hospital protocols, discharge
procedures
Quick feedback to patients requests
Caring Reduction of patients waiting time to industrys best
Caring nurses
Prompt response to all patients care issues
Excellent in-patient facility
Expert Competent doctors
State-of-the-art medical equipment and first-class technical
support
Highest level of Reduction of treatment error/diagnostic error to minimum
professional care Reduction of unneeded treatment to minimum
Consumer promotions
Direct marketing
Event marketing
Product placement
Internal employee communications
In order for the brand to achieve its intended positioning in the mar-
ketplace, it is very important that the brand image is communicated
through various vehicles using an integrated marketing communica-
tions strategy (Davis 2000). The message delivered through all these
vehicles must be consistent and must be related back to the brand image.
The communications strategy determines the best mix of vehicles for
communicating the brand image.
8.2.4.5 Determine the Channel Strategy The objective of this step is to

determine the appropriate distribution channel strategy to enhance the
brand image, and in the case of existing brands, to leverage the strength
of the brand. The selection of the appropriate distribution channel is very
important because of the association that is created between the image of
the channel and that of the brand. Also, because a strong brand can create
a draw to a distribution channel, it is necessary during this step to lever-
age the power of the brand to create the best distribution arrangements
and ensure more control over the distribution of products and services.
8.2.4.6 Determine the Pricing Strategy This step focuses on determining

the correct pricing policy for the brand. A brands price must be related to
the benefits the brand provides (Aaker 1996). An overpriced brand will
not be rewarded in the marketplace, and an underpriced brand can ne-
gate certain associations in the brands image. Also, the ability to charge
premium prices is one of the benefits of developing a strong brand (Davis
2000), so this must be leveraged in determining the pricing strategy.
8.2.5 Brand Implementation

During the brand implementation phase the plans developed in the
brand strategy phase are executed.
8.2.5.1 Communicating and Marketing the Brand This objective in this

step is to communicate the brand to the marketplace using the integrated
communication and marketing strategy developed in the previous phase.
During this step care is taken to use all the selected communication
vehicles effectively to achieve the optimal sales per dollar spent. Also, in
order to improve the effectiveness of the communications, it is important
to track all marketing expenditures, by product, promotional tool, stage
of the life cycle, and observed effects in order to establish a baseline for
improving the use of these tools.
8.5.2.2 Public Relations The purpose of public relations in the develop-

ment of the brand image is to encourage positive feelings about the brand
among the public. However, unlike marketing, which is very visible in
the development of the brand, a well-executed public relations campaign
is not visible (Levine 2003: 17). The public relations activity achieves
its objectives by encouraging third parties to deliver positive messages
about the brand. These messages are usually delivered through news or-
ganizations, and print journalists, in the form of news and press releases
(Levine 2003: 17). Because the company does not have any control over
the news outlet, a challenge during this step is ensuring that the mes-
sage that is delivered is true to the brand image.
8.2.6 Brand Evaluation

This purpose of the brand evaluation step is to measure the performance
of the brand in the marketplace. The classical brand management pro-
cess emphasizes two metrics: recall and awareness of the brand (Davis
294 Chapter Eight
2000). However, these measures alone are not suited for measuring
brand performance, as determined by the equity value of the brand. In
order to provide information for managing the brand as an asset, brand
performance measures should do the following (Davis 2000):
Provide an understanding of how the brand is performing internally
and externally
Provide information about return on investment of marketing and
branding strategies
Assist the organization in its resource allocation decisions
Provide information for rewards and incentive systems
Some of the additional brand performance measures include

Acquired customers
Lost customers
Customer satisfaction
Purchase frequency
Market share
Return on advertising
Price premium
Example 8.5 New Marlboro Cigarette Brand Development The new

Marlboro cigarette brand development process is a brand development
legend. This example illustrates a successful brand development.
The Marlboro cigarette brand is now the best-selling packaged ciga-
rette in the world. However, as recently as the late 1950s, it was an old,
dying tobacco brand in the United States. In 1954, after careful analysis
of the trends in the tobacco market, the management of Philip Morris
Old Marlboro Marlboro

Mild tar blend Stronger blend
Less flavor More flavor
No filter Filter
White pack design Red and white pack design
Older image More modern image
Aimed at women Aimed at men and women
Product-based advertising Imagery advertising
made a number of key decisions on the changes of brand position, as

contrasted in the following table:
To match the newly designed brand image, the product was totally
redesigned. At that time, 90 percent of U.S. smokers used unfiltered
cigarettes, the company realized that the coming trend would be filters,
and this could also help to modernize the image of the brand. To change
the perception that the Marlboro cigarette is a mild cigarette for women,
the flavor of the cigarette was made stronger, and the filter was covered
in tobacco-brown paper, indicating strength and flavor. To shape the
new Marlboro brand identity, a new advertisement agency, Leo Burnett,
was contracted by Phillip Morris to develop a campaign to relaunch the
brand using male role models in tough, rugged jobs, in order to project a
new Marlboro brand personality. At the beginning, the images of pilot,
deep-sea fisherman, cowboy, and engineer were tried. In 1963, market
research indicated that Marlboro needed a more clear-cut identity. The
Marlboro Man, symbolized by a cowboy, was established.
Campaign guidelines were laid down as follows:
The cowboy must symbolize the type of man that other men would
like to be, and women would like to be with.
He must be believable.
Marlboro Country must always be magnificent, never ordinary.
Every ad in the campaign must be candid and have impact.
Variety will be achieved by rotation of cowboy portraits, smoking
moments, and magnificent country material.
To the present day these guidelines have been maintained through

all media. To ensure projecting a consistent brand image, the Marlboro
advertisement and campaign style is highly consistent in worldwide.
After all these efforts, the sales of Marlboro brand cigarette steadily
increased, and by 1975, Marlboro had grown to U.S. brand leadership,
and is now an international brand leader.
Chapter
Value Engineering
9
Any products customer value and satisfaction can be improved by
increasing customer benefits and reducing cost. Among the customer
benefits, functional benefits are of key importance. People pay for functions,
not for hardware or paperwork. For example, people go to a fast-food
restaurant to buy such functions as relieving hunger, getting nutrition,
getting taste, and so on. People go to hospitals, not to buy the doctors
time, or surgery or hospital beds, but to buy such functions as curing
disease and relieving symptoms. Value engineering is a systematic,
team-oriented, creative approach that seeks to deliver functions the
customer wants with lower cost.
The Society of American Value Engineers (SAVE) defines value engi-
neering as follows: Value engineering is the systematic application
of recognized techniques which identify the functions of a product or
service, establish a monetary value for that function, and provide the
function at the lowest cost.
However, value engineering is not merely a cost-cutting program. It
only cuts unnecessary costs, which are the costs that can be removed
without affecting the functional performance of the product or service.
The new design coming out of a value-engineering project should have
the same or better functional performance than the old design. It has
been estimated that for the average product or service, 30 percent of
its cost is unnecessary. This unintentional cost is the result of habits,
attitudes, and all other human factors.
In this chapter, we will look at the six phases in value-engineering
projects, and then at two value-engineering case studies in the service
industry. First, though, well look at how value engineering works.
297
298 Chapter Nine
9.1 An Overview of Value Engineering

Value engineering originated at General Electric Company in 1947. Harry
Erlicher, Vice-President of Purchases, noted that during wartime, it was
frequently necessary to make substitutions for critical materials that not
only satisfied the required functions but also had better performance and
lower cost. He reasoned that if it was possible to do this in wartime, it might
be possible to develop a system that could be applied to normal operations
to increase the companys efficiency and profit. L. D. Miles was assigned to
study the possibility, and the result was a systematic approach to problem
solving based on functional performance, which he called value analysis.
Value analysis, value engineering, value management, value assur-
ance, and value control are all the same in that they make use of the
same set of techniques developed by Miles in 1947. In many cases, the
title tends to describe how the system is being applied. Value analysis
is applied to remove cost from a product. Value engineering and value
assurance are applied in the development phase to keep cost out of a
product. Value management and value control are overall programs that
apply value techniques in business operations.
Value engineering was first applied in product development, manufac-
turing, and the construction industry, and in the 1970s, value engineer-
ing began to be applied in the service industry. David Reeves case study
(1974, 1978) on youth service bureaus was among the first successful
case studies in a service organization. Since then, successful value-
engineering service case studies have been reported in retail, finance,
health care, photo shops, and many other areas.
9.1.1 Collecting Information and Creating

Design Alternatives
Value engineering achieves results by following a well-organized
approach. It identifies unnecessary cost and applies creative problem-
solving techniques to remove it. The three basic steps in this planned
approach are
1. Identify the functions (what does the product or service do for

customers?).
2. Evaluate the functions (what is the lowest cost to create these
functions?).
3. Develop alternatives (what else will do the job?).
9.1.1.1 Identifying Functions Function is the very foundation of value

engineering. The concern is not with the part or the act itself but with
what it does; what is its function? It may be said that function is the
objective of the action being performed by the product or system.
Value Engineering 299
Function is the property that makes something work or sell. We pay

for a function, not for paperwork, or hardware. Hardware has no value
in itself. Similarly we pay to retrieve information, not for the papers.
Focusing on function tends to break down barriers to visualization by
concentrating on what must be accomplished rather than how a task
is presently being done. Concentrating on functions opens the way to
innovative approaches.
Defining functions is not always easyit takes practice and experi-
ence. Functions must be defined in the broadest possible manner so that
the greatest number of potential alternatives can be developed to satisfy
the function. A function must also be defined in two words, a verb and a
noun. If the function has not been defined in two words, the problem has
probably not been properly defined. These are some examples of simple
functions: create design, evaluate information, determine needs, grow
wealth, and enclose space.
There are two types of functions: basic and secondary. The basic func-
tion describes the most important action performed. The secondary
function supports the basic function and almost always adds cost.
9.1.1.2 Evaluating Functions After the functions have been defined and
identified as basic or secondary, they must be evaluated to determine if
they are worth their cost. This step is usually done by comparing them
with something that is known to have a best cost. The best cost is the
lowest overall cost for reliably providing a function.
9.1.1.3 Developing Alternatives Function has been defined as the prop-

erty that makes something work or sell, and the best cost is the low-
est overall cost to reliably provide the function. In value-engineering
analysis, if we find that the current cost for providing a function is
significantly higher than the best cost, we need to ask: what else will
do the job? That is, we must try to develop alternative ways to perform
this function.
In order to develop alternatives, we need to make maximum use of
imagination and creativity. This is where team activities make a major
contribution. The basic tool is brainstorming. In brainstorming, we follow
a rigid procedure in which alternatives are developed and listed with
no attempt to evaluate them. Evaluation comes later. At this stage, the
important thing is to develop revolutionary solutions to the problem.
Free use of imagination means freedom from the constraints of past
habits and attitudes. One persons seemingly wild idea may trigger the
best solution to the problem from someone else. Without a free exchange
of ideas, the best solution may never be developed. A skilled leader can
produce outstanding results through brainstorming and by providing
simple thought stimulation at the proper time.
300 Chapter Nine
9.1.2 Evaluating, Planning, Reporting,

and Implementing
The creative phase does not usually result in concrete ideas that can be
developed into outstanding products. The creative phase is an attempt
to develop the maximum number of possible alternatives to satisfy a
function. These ideas or concepts must be screened, evaluated, com-
bined, and developed to produce a practical recommendation. It requires
flexibility, tenacity, visualization, and frequently the application of spe-
cial methods designed to aid in the selection process.
The final recommendations must be accepted as part of a design or plan
to be successful. In short, they must be sold. They must show the benefits
to be gained, how these benefits will be obtained, and finally, proof that
the ideas will work. This takes time, persistence, and enthusiasm.
9.1.3 The Job Plan

A step-by-step approach makes value engineering an effective tool. The
approach is called the job plan, and it consists of six phases:
1. Information phase
2. Creative phase
3. Evaluation phase
4. Planning phase
5. Reporting phase
6. Implementation phase
Each phase is designed to lead to a solution to the problem after all

of the factors are considered.
9.2 Information Phase

The first phase of the value-engineering job plan is the information phase.
This is the most time-consuming and the most important phase. In the
information phase, we will collect all the necessary raw information for
the project, including product descriptions, process flowcharts and layouts,
and all relevant cost information. Based on the information collected in
this phase, we will produce three important documents for the project:
Function list This is a complete list of all functions that are

required in order for the product to work properly. In this list each
function is defined and classified.
Cost-function worksheet This is a complete cost breakdown cal-
culation for all the product elements (subtasks, items, or components),
as well as for all the functions. The worksheet also lists the actual cost
and best cost for each function.
FAST (Function Analysis System Technique) diagram This
is a very important diagram that provides the exact logical linkage
among all functions. The actual cost and the best cost for each function
are also recorded in the FAST diagram.
The information phase consists of three separate parts:
1. Information development
Collect information
Determine cost visibility
Set a goal
2. Function determination
Define functions
Eliminate duplication
3. Function analysis and evaluation
Construct FAST diagram
Function/cost analysis
Function evaluation
Identify problem areas
Compare potential benefit to the goal
The work done in the information phase is the basis for developing
alternative low-cost methods to perform the required functions. If the
functions have not been properly defined and evaluated, the analysis
will not be correct and the most satisfactory solution is not likely to be
developed. Similarly, if the cost figures are incorrect or incomplete, the
low-cost solution will not be identified.
9.2.1 Information Development

The first step in the information phase is the collection of all available
information concerning the project. This includes drawings, process
sheets, flow charts, procedures, and any other available material. It
is important to discuss the project with people who are in a position
to provide reliable information, and to verify that honest but wrong
impressions are not being collected; that is, information that may have
been true at one time but is no longer valid.
It is very important that good human relations be used during this
information-collecting phase. Get the person originally responsible for
302 Chapter Nine
TABLE 9.1 Project Identification Checklist
1. Flow charts, organization charts

2. Detailed transaction data
3. Facility layout
4. Service product profile
5. Cost data (labor, overhead, material)
6. Work instructions
the project or development to help by showing him how he will be able

to profit from a successfully completed study.
Table 9.1 shows an example of a project identification checklist, which
details all of the information required for a study. It should be filled
out as a first step to identify the project. If the information listed in the
chart is not on hand, it will be necessary to obtain it.
9.2.1.1 Cost Visibility The next step is to complete the cost-visibility

section of the cost-function worksheet, as illustrated in Table 9.2, where
the items in the leftmost column refer to parts of a product or steps of
a service process.
This cost-function worksheet is an important document that should be
produced in the information phase. The left side of this worksheet is the
cost-visibility portion. This is where all costs are listed in a very detailed
fashion, allowing no ambiguity or misunderstanding. Cost visibility does
TABLE 9.2 Cost-Function Worksheet
Cost Visibility
Total Cost $________ Cost Elements Cost Function Analysis

Item No. Name Material $ Labor $ Burden $ F1 Function 1 F2 . . .
1
2
3
.
.
.
Cost Total
Best Cost
not tell us where unnecessary costs are; it tells us where high costs are.
This is important because it identifies a starting point. The right side of
the worksheet is about analyzing the functions delivered by each item
and it will be discussed later.
The following definitions are commonly used in cost-visibility analysis.
Cost The amount of money, time, labor, and so on, required to obtain
anything. In business, it is the cost of making or producing a product
or providing a service.
Fixed costs Cost elements that do not vary with the level of activity
(insurance, taxes, plant, and depreciation).
Actual costs Costs actually incurred during the performance of
a process. They include labor, materials, and costs related to local
ground rules.
Incremental costs Not all variable costs vary in direct proportion
to the change in the level of activity. Some costs remain the same over
a given number of production units or transactions, but rise sharply
to new plateaus at certain increments. The costs thus affected are
incremental costs.
Materials All hardware, raw materials, and purchased items con-
sumed in producing a product.
Labor Manpower expended in producing a product or performing
a service.
Burden (overhead) All costs incurred by the company that cannot
be traced directly to specific products. The accounting department
determines burden rates, which are assigned to individual opera-
tions on a formula basis. Burden consists of both fixed and variable
categories, and separate rates are often established for each.
The method of assigning burden differs from industry to industry and
even from one company to another within an industry. Any quantifi-
able product factor may serve as a basis for assigning burden as long
as consistent use of the factor across the entire product line results
in full and equitable burden distribution.
Fixed burden Includes all continuing costs regardless of the pro-
duction volume for a given item, such as salaries, building rent, real
estate taxes, and insurance.
Variable burden Includes costs that increase or decrease as the
volume rises or falls. Indirect materials, indirect labor, electricity used
to operate equipment, water, and certain perishable tooling are also
included in this classification.
304 Chapter Nine
Allowance Costs other than material, labor, and burden that must
be included in the total cost of a product, such as packaging materials,
scrap, inventory losses, inventory costs, and so on.
Total cost includes production costs plus profit and other expenses.
The following expenses are usually added to production costs by sales

and/or accounting departments to make up the total cost:
Administrative and commercial costs Costs incurred in admin-

istering the company, and researching and selling the product. They
are usually represented as a percentage of production cost.
Freight cost Shipping and handling costs.
Profit Amount earned in producing a product or a service. It is usu-
ally applied as a percentage of production cost.
9.2.1.2 Sources of Cost Information The application of cost-visibility

techniques begins with an analysis of total cost, and progresses through
analyses of cost elements to component or process costs. To perform
these analyses, the best available cost information is required. This
information will be available from sources such as
Accounting Current and historical costs (actual costs)

Purchasing Cost of purchased items
Suppliers Estimates and/or quotations, costs, process information,
and material prices
In the service industry, labor is usually a big portion of cost. In order to

figure out the exact labor cost component in each item, traditional motion-
time studies have to be performed. For example, in the health care industry,
a doctors time is an important source of cost because it is very expensive.
If we conduct a value-engineering study on emergency care, we may have
to use a stopwatch to track the doctors time spent on patient visits. After
recording a sufficient number of patient visits, we can calculate the average
doctor time and use that as a basis in computing the doctors cost.
Review this cost data collected so far and make a preliminary judg-
ment of the potential profit improvement. Consider the factors involved
and set a goal that will provide a profitable position. The target should
indicate a 30100 percent cost reduction. It may seem improbable that
this can be achieved, but it is a target to work toward. A comparison to
this target will be made at the completion of the information phase.
Example 9.1: A Cost-Visibility Worksheet of a Youth Assistance

Program (Reeve 1974) Reeve (1974) did a value-engineering study on
the youth assistance program for Oakland County, Michigan. This is
one of the very first case studies of value engineering for a government/
service organization. The purpose of the youth assistance program is
to help troubled teenagers so they will not become problems for society.
There are two major activities in the youth assistance program: pre-
vention and rehabilitation. Each activity is to be accomplished through
various meetings, contacts, field visits, and office activities.
Tables 9.3 and 9.4 show the cost-visibility sections of the cost-function
worksheets for rehabilitation and prevention, respectively.
In these cost-visibility calculations, the labor cost is computed based on
labor hours multiplied by the labor rate. The labor hours are based on the
historical records of meeting lengths, interview duration, and so on.
Example 9.2 A Cost-Visibility Worksheet for an Automobile Hood

Latch Table 9.5 shows a hardware cost-visibility worksheet for an au-
tomobile hood latch.
This is an example of a cost-visibility worksheet for a product devel-
opment case.
9.2.1.3 Project Scope Once the cost-visibility worksheet has been

completed, it is possible to make a preliminary determination of the
project scope. By considering the project as outlined on the project
TABLE 9.3 Cost Visibility of Rehabilitation per Case
Cost Visibility
Total Cost = $109.64/case Cost Elements
Item No. Name Material Labor Burden Total Cost ($)
1 Client Contact 27.19 27.19
2 Organization Contact 11.04 11.04
3 Secretarial Center Office 6.07 6.07
4 Secretarial Field Office 31.12 31.12
5 Case Management 11.59 11.59
6 General Administration 2.21 2.21
7 Grant Administration 1.38 1.38
8 Others 4.42 4.42
9 Travel Time 5.11 5.11
10 Administration Meetings 1.17 1.17
11 Supervisory Meetings 4.69 4.69
12 Training Meetings 2.35 2.35
13 Statistical Meetings 0.97 0.97
14 Evaluation Meetings 0.34 0.34
306 Chapter Nine
TABLE 9.4 Cost Visibility of Prevention per Case
Cost Visibility
Total Cost = $41.75/case Cost Elements
Item No. Name Material Labor Burden Total Cost ($)
1 Client Contact 1.17 1.17
2 Organization Contact 2.90 2.90
3 Secretarial Center Office 3.45 3.45
4 Secretarial Field Office 8.56 8.56
5 Case Management 0.28 0.28
6 General Administration 2.07 2.07
7 Grant Administration 0.28 0.28
8 Others 2.55 2.55
9 Travel Time 2.90 2.90
10 Administration Meetings 1.86 1.86
11 Supervisory Meetings 4.90 4.90
12 Training Meetings 0.41 0.41
13 Statistical Meetings 0.14 0.14
14 Evaluation Meetings 0.21 0.21
15 Advisory Council 2.76 2.76
Meetings
16 Citizen Committee 1.24 1.24
Meetings
17 Citizen Subcommittee 6.07 6.07
Meetings
identification sheet, the present cost and target for improvement,

and the time available for the study, we can define the scope of the
project.
In a value-engineering project, the analysis of function should first
be performed upon the system level or the whole process level. If the
objectives of the value-engineering study are not achieved at that level,
the second level should be studied, and so on down to the lowest possible
level of detail. The lower the level of system hierarchy it reaches, the
more detailed and complex the study might become.
9.2.2 Function Determination

Once you have defined the initial scope of the project, it is possible to
start defining the functions to be performed or that are being performed
by the product or service.
TABLE 9.5 Cost-Visibility Worksheet for a Car Hood Latch
Cost Visibility for Hood Latch

Cost Elements
Total Cost = $2.616 Material Labor Burden
(Total (Total (Total Total Cost ($)
Item No. Part Name $1.545) $0.713) $0.358) (Total: 2.616)
1 Primary Spring 0.219 0 0.035 0.254
2 Detent Spring 0.09 0 0.015 0.1046
3 Hook Spring 0.09 0 0.015 0.1046
4 Pivot Rivets 0.09 0.005 0.015 0.1104
5 Hook Pivot 0.08 0.005 0.015 0.0988
6 Fork Bolt 0.04 0.096 0.014 0.158
7 Mounting Bracket 0.426 0.198 0.101 0.7253
8 Back Plate 0.08 0.149 0.037 0.2661
9 Secondary Hook 0.26 0.151 0.067 0.4776
10 Detent Level 0.11 0.099 0.034 0.2429
11 Grease 0.02 0.005 0.004 0.0291
12 Sleeve 0.04 0.005 0.007 0.0523
9.2.2.1 What Is a Function? As has been mentioned, a function is the

property that makes something work or sell. The function is the end
result desired by the consumer; it is what consumers pay for. Function
is a requirement, a goal, or an objective.
A function is not an action; it is the objective of an action. For example,
file papers is an action, but what is the purpose of the file papers
action? We file papers not because we enjoy putting papers in folders
or cabinets, but because we want to keep a record that we can use later.
Therefore, the objective of file papers is actually to store information,
so in this case, the function should be store information. The function
is the desired result accomplished by an actionthe action is just one
method of accomplishing the objective.
Functions should be defined in two words, a verb and a noun. The
resultant definition should not define a method for achieving the
result. Abstract definitions offer more opportunity for creative ques-
tions that may produce a number of alternative solutions. For example,
using file papers as the function might limit our thoughts to using
papers, folders, cabinets. By using a more generic function definition,
such as store information, we are free to think about other solutions
involving computers, CDs, and so on to read, retrieve, and catalog
information.
308 Chapter Nine
It is also important that the function be measurable in some way

weight, cost, volume, time, space, etc. In some cases, satisfaction, desire,
or some other abstract measure may require more subjective analysis,
but it can still be measured by comparative techniques.
The following are some examples of function definitions and their
measures:
Verb Noun Unit of Measure

Create Design Time
Confirm Design Time
Authorize Program Cost
Measure Performance Manhours
In his book Techniques of Value Analysis, Miles recognized the difficulty

of applying this technically simple concept: While the naming of functions
may appear simple, the exact opposite is the rule. In fact, naming them
articulately is so difficult and requires such precision in thinking that real
care must be taken to prevent the abandonment of the task before it is
accomplished. He also said, Intense concentration, even what appears to
be over concentration of mental work on these functions, forms the basis
for unexpected steps of advancement of value in the product or service.
Defining functions in two words is the most difficult naming method.
There is some feeling among value-engineering practitioners that
it is unnecessary to struggle for two-word definitions; three-word
definitions or short statement definitions should do the job. However,
value-engineering practitioners also found that if the goal of a value-
engineering study is to generate creative design solutions, two-word
functions are imperative. If the function cannot be defined in two words,
the function needs to be better understood. It is a struggle to define good
functions, but the result is worth the effort.
Defining functions in two words is a forcing technique that requires
consensus among team members, eliminates confusion, creates in-depth
understanding of the requirement, clarifies overall knowledge of the
project, and ultimately breaks down barriers to visualization, which is
necessary to help define the creative questions that will lead to new,
outstanding solutions to the project.
9.2.2.2 Types of Functions There are two types of functions: work

functions and sell functions.
Work functions are those that do the job the customer wants. Work func-
tions provide use value. They are always expressed in action verbs and
measurable nouns, which result in quantitative definitions. This pro-
vides us with a quantitative means of measuring the work functions.
Sell functions are functions that add appeal for the customer and make
them want to buy. Sell functions usually provide prestige value to cus-
tomers. They are always expressed in passive verbs and nonmeasur-
able nouns, which create qualitative definitions. Their measurement
is extremely difficult.
For example, these are some work functions:
Verb Noun Unit

Collect Payment Dollar
Remove Kidney Time and/or cost
Transfer Fund Dollar/Time
Sell Assets Dollar/Cost
These are sell functions:
Verb Noun
Increase Beauty
Improve Style
Increase Prestige
Separating work and sell functions will help us to define the func-
tions more precisely. It will also help us to identify the proportion of cost
allocated to use value and prestige value.
All functions can also be divided into two levels of importance: basic
functions and secondary functions. The basic functions are those that
fulfill the primary purpose of a product or service. Secondary functions
do not directly fulfill the primary purpose of the product or service, but
support the primary purposes.
The result of the function determination step should be a completed
function list as illustrated in Table 9.6.
TABLE 9.6 Function List
Scope includes:
Project Name Scope does not include:
List All Functions Function Types
Verb Noun Basic Second Work Sell Remarks
1.
2.
3.
4.
310 Chapter Nine
Figure 9.1 The parts of a pencil

Lead Body Paint Band Eraser
Example 9.3 Pencil Function Determination Figure 9.1 shows a

pencil. A pencil has five parts: lead, body, paint, band, and eraser. Even
a very simple product such as a pencil will have quite a few functions,
and it is easier to figure out a pencils functions by looking at its five
parts. Table 9.7 summarizes the functions performed by the parts of a
pencil. Table 9.8 presents the function list for the pencil.
Example 9.4 Function List of Oakland County Youth Assistance

Program This example is a continuation of Example 9.1, which listed
all the organizational activities that support youth assistance programs
and their cost calculation. Reeve (1975) identified 41 functions in this
program, some of which are listed in Table 9.9.
Reeve also provided a glossary of functions, which provides detailed
definitions for each function. In a value-engineering project, it is highly
recommended that such a glossary be developed and that consensus be
secured from group members on the definitions of each function. That
way, in later discussions, every team member will be on the same page
when each function is discussed. Here are some of the function defini-
tions for this youth assistance program project:
Identify need Time spent in written and oral communication, that
is, conferences, letters, interviews, etc., with school personnel and/or
other referral sources regarding potential referrals in order to deter-
mine the need for the referral process.
TABLE 9.7 Functions Performed by Pencil Parts
Function
Part Verb Noun
Lead Make Marks
Eraser Remove Marks
Band Secure Eraser
Improve Appearance
Body Support Lead
Transmit Force
Accommodate Grip
Paint Display Information
Protect Wood
Improve Appearance
TABLE 9.8 Function List of a Pencil
Scope includes:
Project Name: Pencil Scope does not include:
List All Functions Function Types
Verb Noun Basic Second Work Sell Remarks
1. Make Marks
2. Remove Marks
3. Secure Eraser
4. Improve Appearance
5. Support Lead
6. Transmit Force
7. Accommodate Grip
8. Display Information
9. Protect Wood
10. Improve Appearance
TABLE 9.9 A Partial List of Functions for a Youth Assistance Program
List All Functions

Verb Noun Basic Second Remarks
1. Identify Need
2. Define Problem
3. Plan Treatment
4. Diagnose Problems
5. Obtain Information
6. Involve Client
7. Identify Client
8. Utilize Resource
9. Assist Client
10. Improve Process
11. Indicate Trend
12. Maintain Record
13. Establish Standard
14. Analyze Data
15. Terminate Contact
16. Evaluate Process
17. Eliminate Deviancy
312 Chapter Nine
TABLE 9.9 A Partial List of Functions for a Youth Assistance Program (continued)
List All Functions
18. Plan Activities
19. Determine Needs
20. Set Goals
21. Secure Action
22. Provide Alternatives
23. Develop Programs
24. Establish Trust
25. Exhibit Concern
26. Improve Programs
27. Evaluate Programs
Assist client This action includes counseling, offering alternatives,

providing a referral service, indicating community programs, helping
kids get to camp, talking to teachers or police or other authorities on
clients behalf, aiding parents and children.
Eliminate deviancy (Measured by time spent) To return to a
homeostatic position and develop modification techniques to reach
normative behavior patterns.
Define problems All communication with client, parents, and
referral sources for the purpose of describing the clients problem
behavior.
9.2.2.3 Creativity and Function Definitions The ultimate objective

of value engineering is to create a better product or service design.
Creativity is very important in coming up with new and better designs.
But what makes people more creative? What is the ingredient that
allows some people to break the barriers to visualization, to be able
to look at something and immediately think of new exciting possibili-
ties for new products, services, methods, or other useful or satisfying
subjects?
The consensus seems to be that to be creative, one must be able to see
beyond the obvious, the existing, but there is no clear-cut formula for
producing creative people. It is known that creative people are some-
what different and that they exhibit certain characteristics. However,
given two people with the same characteristics, one may prove to be
creative, and another may not. Many people feel that the seeds for cre-
ativity exist in everyone.
In all probability, at least one of the ingredients of creativity is the

ability to visualize, to detach oneself from reality and see beyond the
stated problem, the object, or the material facts. A creative person
must be able to create concepts, broaden and develop them, analyze
and examine them, and out of all this select a new idea, new approach,
or new solution to a requirement or problem.
According to L. D. Miles, creative thinking is constrained by the physi-
cal shape or concept of existing products and services. Concentrating
on function helps to break down the barriers to visualization and offers
outstanding opportunities for creativity.
The conventional approach to improving a product or process is to try
to make the existing product work better, cost less, or meet some other
objective. In this situation, creativity is stifled because the existing form
constrains thinking. The function approach is truly different. It breaks
the project into requirements called functions. The process of defining
functions becomes a method for removing the barriers to visualization,
making entirely new solutions possible.
The concept is disarmingly simple. It is easy to understand without
learning complex systems or studying complex technology. However, the
ability to use the system comes only from a thorough understanding of
the principles and the determination and discipline to use them.
Function analysis is basic to value engineering and it starts with an
understanding of the term function and of how to define functions
that will offer creative opportunities. Defining and analyzing functions
can help a person or group visualize alternative solutions. In fact, the
struggle necessary to define functions properly makes it possible to see
new and different things in subjects you may have seen many times
before. It can help you imagine solutions that go beyond the stated
problem, as outstanding people have done throughout the ages.
This means that not only are functions the basic ingredient of value
engineering, but they provide the opportunity for people to break down
barriers and create new things, to eliminate prejudices and allow
insights never seen before.
9.2.2.4Defining Functions for Creativity In defining a function, it is

important that these key questions be kept in mind at all times:
What are we really trying to do when we perform this action? Why is

it necessary to do this?
Why is this part or action necessary?
Specific answers to these questions will aid in zeroing in on a useful

function definition.
314 Chapter Nine
In function definition, it is helpful to think as if you are the product

or service, and play their roles. The idea is to let the job be the boss, as
Kettering said. Be the crankshaft. What do you do? Act the part of the cus-
tomer. How do you feel? What do you see? What does it do for you? If you
were the plant manager, what would you want? How would you get it?
This system helps to eliminate bias in that functions can be defined
from all viewpoints and sorted out in the FAST diagram through its
cause-and-effect relationship for maximum understanding and subject
evaluation.
9.2.2.5 Process for Defining Functions In defining functions, start

with the total project, whether it is an assembly, a complete process,
a program, or an organization. Define the functions. Dont haggle over
whether the function has been properly defined at this stage; it can be
redefined later. Write every thought down so it will not be forgotten.
After all functions of the assembly have been defined, take each part
or segment of the system and define the function of each. There will be
some duplication, but this will be screened out later.
After all functions have been defined, screen the list to eliminate
duplicate functions and redefine functions for clarity. Screen the list
again to define the basic functions. The basic functions are the functions
upon which all other system functions depend. If the basic function were
not needed, none of the other functions would be needed either.
In many cases, a number of functions beyond the system scope will be
defined. These are called higher-order functions, and they are the func-
tions that cause the basic functions to be performed. A detailed discussion
of system scope and high- and low-order functions is beyond the intent of
this orientation; the scope will become clear during the construction of a
FAST diagram, and it may cause a team to reconsider and redefine the
scope because of the new understanding of the overall project.
The end result of applying these function definition principles will be
clearly understandable, measurable functions for use in cost-function
analysis and function evaluation, and will lead to outstanding opportu-
nities in the creative phase.
9.2.3 Function Analysis and Evaluation

After the functions have been determined, identify the basic function
or functions, as well as supporting functions. It is now time to create a
Functional Analysis System Technique (FAST) diagram. The Functional
Analysis System Technique was developed by Charles Bethway in 1964,
and was first presented and published as a paper at the Society of
American Value Engineers Conference in 1965. FAST contributed sig-
nificantly to the most important activity in value-engineering projects,
the function analysis and evaluation.
9.2.3.1 FAST Diagrams A FAST diagram is a logic chart that organizes

the functions of a project and arranges them in a cause-and-effect
relationship. By constructing a FAST diagram you can ensure that
the functions have been properly defined and that nothing has been
overlooked. FAST diagrams are simple in concept, but creating a FAST
diagram is often difficult and frustrating, and it forces people to think
about their project in a detailed and precise manner. Constructing a
FAST diagram will create a focal point for the entire project, because
eventually, all important information on the project will be precisely
defined and displayed in the FAST diagram.
FAST diagrams are especially useful for projects where there are
widely different opinions, fuzzy understandings, or cloggy definitions
among team members, which are very common in the analysis of orga-
nizations, operations, and services. The construction of a FAST diagram
tends to pull together the thinking of a group to create a dynamic,
enthusiastic team.
Determining the basic function is the first step in constructing a FAST
diagram. The basic function is the function that cannot be eliminated
unless the product can be eliminated. There may be more than one, but
an effort should be made to determine the one most basic function.
We will use Example 9.5 as a starting point to discuss FAST diagrams.
Example 9.5 Portion of FAST Diagram for Youth Assistance

Program As you saw in Example 9.4, identify need, assist client,
and eliminate deviancy are considered to be basic functions for the
youth assistance program, because if any of these functions are not
performed, the whole youth assistance program will not perform as
intended. We can easily see that the functions are not working in isola-
tion; they are related to each other and together perform the overall
mission of the system. We can ask why do we need these three basic
functions (identify need, assist client, and eliminate deviancy)?
If we think really hard, we may come to the conclusion that because
this is a youth assistance program, we want to change the life of these
troubled youth so they can become better kids. Then we may come to
another function, modify behavior; that is, the three basic functions
are needed because we want to modify behavior. Figure 9.2 illustrates
this relationship.
At the top of Figure 9.2, the Why-How arrow indicates the relation-
ships among these four functions:
How do you modify behavior? You have to identify needs, assist

clients, and eliminate deviancy.
Why we need to perform these three functions? Because we want to
modify behavior.
316 Chapter Nine
Why How
Eliminate
Deviancy
Modify Identify
Behavior Needs
Figure 9.2 Relationships between basic functions and

Assist higher-order function
Clients
Similarly, we can expand Figure 9.2 by adding more functions, as

Again, the function on the left gives the reason why the function on
the right should be performed, and the function on the right tells how
the function on the left can be accomplished. For example, the function
on the right of eliminate deviancy is plan activities, and the function
on the right side of plan activities is determine needs. We plan activi-
ties in order to eliminate deviancy, and we determine needs in order to
plan activities. How do we eliminate deviancy? By planning activities!
How do we plan activities? By determining needs!
9.2.3.2 Understanding the FAST Diagram Now lets look at how to es-
tablish a FAST diagram for a value-engineering project. The general
format of a FAST diagram is illustrated in Figure 9.4.
The following terms are used in FAST diagrams:
Scope of the project Scope of the project is defined by two verti-
cally dotted lines. The portion between the two lines is the scope of
the project.
How Why
Eliminate Plan Determine

Deviancy Activities Needs
Modify Identify Define Plan

Behavior Needs Problems Treatment
Figure 9.3 Part of FAST

diagram for youth assistance
Assist Improve program
Client Process
Minor primary path

How Why
Independent
Dependent
Function
Function
(Supporting)
Objectives
or
Specifications
High Lowest
Order Basic Dependent Order
Function Function Function Function
(Output) (Input)
When Major primary path Activity
Scope of the project
Figure 9.4 FAST diagram format
Highest-order function(s) This is the highest level of function.

It is the overall objective of the product or service under study. The
highest-order function is the output of the basic function(s). The high-
est-order function should be positioned just outside of the left scope
line and to the left of the basic function(s). In general, any function
to the left of another function on the primary path is a higher-order
function.
Lowest-order function(s) The functions on the rightmost position
and outside of the right scope line are the lowest order functions. They
represent the initial inputs to all other functions. In general, any func-
tion to the right of another function on the critical path is a lower
order function. The higher- or lower-order functions do not carry any
meanings of relative importance.
Basic functions Basic functions are the functions located to the
immediate right of the left scope line. The basic functions represent
the purpose of the system under study. In value engineering, the basic
functions cannot change. Secondary functions can be changed, com-
bined, or eliminated.
Concept All the functions that are located to the right of the
basic function(s) are used to describe the means to accomplish basic
function(s). The concepts could be either existing conditions or pro-
posed approaches to accomplish basic functions(s).
318 Chapter Nine
Objectives or specifications Objectives or specifications are par-

ticular parameters or requirements that must be achieved in order
to satisfy the highest-order functions. Objectives or specifications are
not functions. In value engineering, objectives and specifications are
studied because they will influence the method selection to achieve
the basic functions and satisfy the users requirements.
Primary path functions Any function on the How or Why logic
is a primary path function. If the function is along the Why direction
that enters the basic function(s), then it is a major primary path.
Otherwise, it will be an independent (supporting) function, and it is
on a minor critical path. Supporting functions are usually second-
ary functions. Supporting functions are needed to help the system
to achieve the objectives or specifications of the basic functions; they
could also be needed because of a particular technical approach.
Dependent functions From the first function to the right of the
basic function(s), each successive function is dependent on the func-
tion that is on its immediate left, or higher-order function.
Independent functions Independent functions do not depend on
another function or method. Independent functions are located above
the critical path functions.
Activity The activity is the method selected to perform a function,
or a group of functions.
Primary path functions Any function on the How or Why logic
is a primary path function. If the function along the Why direction
enters the basic function(s), it is a major primary path; otherwise, it
will be identified as an independent (supporting) function and be a
minor critical path. Supporting functions are usually secondary. They
exist to achieve the performance levels specified in the objectives or
specifications of the basic functions, or because a particular approach
was chosen to implement the basic function(s).
Independent functions (above the critical path) and activities (below
the critical path) are the result of satisfying the When question.
Dependent Functions Starting with the first function to the right
of the basic function, each successive function is dependent on the
one to its immediate left, or higher-order function, for its existence.
That dependency becomes more evident when the How question and
direction is followed.
Independent (or Supporting) Function(s) Independent (or sup-
porting) functions do not depend on another function or method selected
to perform that function. Independent functions are located above the
critical path function(s) and are considered secondary with respect to
the scope, nature, and level of the problem, and its critical path.
Activity The method selected to perform a function (or a group of

functions) is an activity.
9.2.3.3 Symbols and Graphs Used in FAST Diagram Construction

Figure 9.5 shows the directions in a FAST diagram. The How and Why
directions are always along the primary path, whether it is a major or
minor primary path. The When direction indicates an independent or
supporting function (up) or an activity (down).
We have already discussed the How and Why directions in Example
9.5the lower-order function in the How direction (immediate right)
always explains how a particular function can be accomplished; the
higher-order function in the Why direction (immediate left) always tells
the reason why a particular function should be performed. All the func-
tions or activities along the When direction with a particular function
will happen at the same time. We can detect these functions or activities
by asking the question When the function occurs, what else happens?
The independent functions and supporting functions are listed above
the particular function; the activities will be listed under the particular
function.
AND and OR Symbols Along the Primary Path In the primary path
of a FAST diagram, it is possible that several functions have to be per-
formed simultaneously as the preconditions for lower-order function(s).
Sometimes, these functions are related by a logical AND, and some-
times they are related by a logical OR. Figures 9.69.9 illustrate such
cases.
In both Figure 9.6 and Figure 9.7, the fork is read as AND. In
Figure 9.6, how do you build swim club? By construct pool AND
construct club house. Construct pool and construct club house are
Independent
Function
When
Function Why Function

Function
(Higher Order) (Lower Order)
How
When
Activity
Figure 9.5 Directions in FAST diagram

320 Chapter Nine
Construct
Pool
Build Contract
Swim Club Builder
Construct
Club House
Figure 9.6 Two equally important functions in an AND relation
equally important. In Figure 9.7, how do you determine compliance

deviations? By analyze design AND review proposals. However,
analyze design is more important than review proposals.
In both Figure 9.8 and Figure 9.9, the multiple exit lines represent
an OR. In Figure 9.8, how do you convert bookings (to delivery)? By
extend bookings OR forecast orders, not both. Extend bookings and
forecast orders are equally important. In Figure 9.9, how do you identify
discrepancies? By monitor performance OR evaluate design. However,
evaluate design is less important than monitor performance.
AND Symbols Along the When Direction In a FAST diagram, the

When direction is vertical. When several functions are located along the
same vertical line, it means that these functions will be performed at
the same time. In addition, when these functions are connected by lines,
it means that there is an AND relationship among them. Figure 9.10
illustrates such an example.
In Figure 9.10, when you influence the customer, you inform cus-
tomer AND apply skills. If it is necessary to rank the AND functions,
those closest to the primary path should be the most important.
9.2.3.4 Completing a FAST Diagram Now that you know the symbols
and notation used in FAST diagrams, we can look at the step-by-step
procedure for completing a FAST diagram.
Step 1 List all functions by using the function list as illustrated
in Table 9.6. Be sure to identify each function with a verb and noun.
Identify basic functions and secondary functions.
Determine Analyze Determine

Compliance Design Criteria
Deviations
Review
Proposals
Figure 9.7 Two unequally important functions in an AND relation

Extend
Bookings
Convert
Bookings
(To delivery) Figure 9.8 Two equally important
Forecast functions in an OR relation
Orders
Step 2 Prepare a 1" 2" card for each function. Take a close look at
all functions and try to identify the relationships among them.
We can use the following logical questions for this purpose:
How? How is this function accomplished?

Why? Why is this function performed?
When? When is this function performed?
Select the function that you think is the basic function and apply the
logic questions to the right and left of the basic function. Ask How is
this function performed? to determine the function to the right. Ask
Why is this function performed? to determine the function to the left.
Repeat this process until the lowest-order functions are included. The
path of functions thus created is called a primary path. You may end up
with multiple primary paths.
Step 3 When a primary path has been selected and positioned on
the chart, position all secondary functions that did not fit into the pri-
mary path by applying the When question, and add them above or below
the primary path depending on whether they are supporting functions,
independent functions, or actions. If the secondary functions are actu-
ally objectives or specifications, put them into the upper-left corner of
the FAST diagram.
Example 9.5 illustrates this step-by-step process.
Example 9.5 Cigarette Lighter FAST Diagram In this example we

will create a FAST diagram for a typical cigarette lighter, as shown in
Figure 9.11. Assume that we have compiled the function list in Table 9.10.
To create the FAST diagram, we first pick up the basic function pro-
duce flame, and ask why and how, as illustrated by Figure 9.12.
Monitor
Identify Performance
Discrepancies
Evaluate Figure 9.9 Two unequally important

Design functions in an OR relation
322 Chapter Nine
Apply
Skill
Inform
Customer
Reach Influence Demonstrate

Agreement Customer Capabilities
Figure 9.10 AND relationship on the When direction
The basic function of the cigarette lighter is to produce flame. By

answering the question why produce flame?, we get the higher-order
function ignite cigarette. By answering the question how to produce
flame? we get the lower-order function ignite fuel.
TABLE 9.10 Function List for Cigarette Lighter

1. Produce Flame
2. Protect Flame
3. Manage Flame
4. Ignite Fuel
5. Release Fuel
6. Produce Spark
7. Control Flow
8. Restrict Exit
9. Energize Particles
10. Strike Flint
11. Generate Heat
12. Contain Fuel
13. Open Valve
14. Depress Lever
15. Enclose Fuel
16. Rub Material
17. Rotate Wheel
18. Apply Force
19. Activate Thumb
20. Accommodate Hand
21. Stimulate Muscle
Figure 9.11 A cigarette lighter
We then can ask the further question, how to ignite fuel? By answer-
ing this question, we will find that we need two lower-order functions to
be performed, produce spark AND release fuel. These two functions
are of equal importance, so we add these two functions in the FAST
diagram as illustrated in Figure 9.13.
Now we can continue to ask why and how questions to find lower-
order functions for release fuel and produce spark, and continue this
process. We will end up with the diagram in Figure 9.14. The functions
shown in this figure form the primary path of the FAST Diagram of the
cigarette lighter.
There are still many functions in the function list that cannot be fit
into the primary path. By asking when, we can fit the rest of the func-
tions into the FAST diagram, as illustrated in Figure 9.15.
9.2.3.5 Cost-Function Relationship When the FAST diagram is fin-

ished, it is possible to complete the cost-function worksheet. The cost-
function worksheet lists all functions versus all parts of a product or
actions of a system, procedure, or administrative activity. The objective
is to convert product cost to function cost.
The cost of each piece of hardware or service activity is redistributed
to the function performed. This proportional redistribution of cost to
function requires information, experience, and judgment, and all team
members must contribute their expertise.
After the cost of each part or action has been redistributed to the
functions performed, the cost columns are totaled to obtain the function
cost. This cost is then placed on the FAST diagram. The FAST diagram
then becomes a very valuable tool. It tells what is happening, why, how,
when, and what it costs to perform the function. It is now possible to
evaluate the functions to determine if they are worth what is being paid
for them. In other words, a value must be set on each function.
Why Produce How?

Flame
Figure 9.12 Start of FAST diagram

Ignite Produce Ignite construction for cigarette lighter
Cigarette Flame Fuel
324 Chapter Nine
Release How?
Fuel
Ignite Produce Ignite
Produce How?
Spark
Figure 9.13 Partial FAST diagram for cigarette lighter
Determining the value of each function is a subjective process.

However, it is a key element in the value process. Comparing the func-
tion cost to function value provides an immediate indication of the ben-
efit being obtained for expended funds. The ratio of value cost to function
cost is the performance index. The sum of all values is the value of the
system or the lowest cost to reliably provide the basic function. It should
be compared to the preliminary goal set earlier.
It may be that the new goal is considerably higher than the original.
If this is the case, an evaluation of the diagram will indicate what must
be done to achieve the original goal. It may indicate that an entirely new
concept is required or it may be that it will be acceptable to settle for less.
It is often the case that the original goal and the new value are close. An
analysis of the function costs will again indicate necessary action.
This analysis clearly defines the task for product improvement. It
breaks the problem down to functions that must be improved, revised,
or eliminated to achieve the goal. The FAST diagram clearly identifies
functions and their relationship to each other. Cost-visibility analysis
can identify high cost areas.
We now are ready to identify the relationship between cost and func-
tion. Specifically, we are ready to identify the cost for each function. Also,
after clearly defining each function, we are able to identify the best cost
for each function. The difference between the current cost and the best
cost is the profit improvement target, which will provide us with an
estimate of profit improvement potential.
Example 9.6 Cost-Function Worksheet for a Pencil Table 9.11 pro-

vides an example of a cost-function worksheet for the pencil discussed
previously in Example 9.3.
Open Depress
Valve Level
Release
Fuel Contain Enclose Apply
Ignite Produce Ignite Fuel Fuel Force
Produce Energize Strike Rub
Spark Particles Flint Material
Figure 9.14 Primary path of FAST diagram for cigarette lighter

How Why
Manage Control Restrict

Flame Flow Exit Open Depress
Valve Level
Protect Release
Flame Fuel
Contain Enclose Apply Activate Stimulate
Fuel Fuel Force Thumb Muscle
Ignite Produce Ignite
Produce Energize Strike Rub Accommodate

Spark Particles First Material Hand
When
Rotate
Scope Wheel
Figure 9.15 FAST diagram for cigarette lighter
We need to determine the cost of each function by distributing the

cost of each part to its related function. For example, the cost for the
pencil body is 0.94 cent, and 50 percent of pencil body cost is used
to perform the function transmit force, 40 percent of its cost is used to
perform the function support lead, and 10 percent of its cost is used
to perform the function display information. This breakdown of cost is
subjective; hopefully, it is not too biased because it is based on the consensus
of the team. By adding all the cost portions from all relevant parts for a
function, we can get the cost for performing that function. For example, in
Table 9.11, the cost of the function transmit force consists of 25 percent
of metal band cost, which is 0.06 cent, and 50 percent of pencil body cost,
which is 0.47 cent, therefore the cost of transmit force is 0.53 cent.
We also need to determine the best cost for each function. By defini-
tion, the best cost is the lowest cost to adequately and reliably provide
the function. The best way to determine the best cost of a function is
by comparison to another function that we know is a best deal. For
example, if a function is tell time, we need to know the time precision
requirement; for example, the required precision might be +/- 30 seconds
after a month of use. Next, we can find a watch that can provide time
with this precision reliably, but that doesnt provide any other functions,
such as decoration, brand name recognition, and so on. In this way, a
cheap, no-brand, plain, 99-cent electronic watch might be adequate. In
this case, the best cost for tell time is 99 cents.
To make sure we determine the best value, we can ask the following
questions:
1. Can we do without it? (If yes, the best cost is zero.)

2. Does it need all its features? (If no, get rid of all unnecessary features
and then figure out the best cost.)
326
TABLE 9.11 Cost-Function Worksheet for Pencil
Functions
Accommodate
Information
Appearance
Transmit
Improve
Support
Remove
Display
Protect
Secure
Eraser
Marks
Marks
Make
Force
Wood
Lead
Grip
Cost (Cents)
Cost
Cost
Cost
Cost
Cost
Cost
Cost
Cost
Cost
Pencil
%
Components
Eraser .43 100 0.43
Metal Band .25 50 0.13 25 .06 25 .06
Lead 1.2 100 1.2
Body .94 50 .47 10 .09 40 .38
Paint 0.10 50 .05 50 .05
Total Cost 2.92 16 .43 5 .13 4 .11 40 1.2 17 .53 3 .09 13 .38 2 .05
Best Cost .34 .10 .10 0.8 .30 .09 .28 .04
Profit .09 .03 .01 0.4 0.03 0 0.1 0.01
Improvement
Potential
Function Number Cost
3 $4.69
Plan
Treatment
$4.07
Figure 9.16 A fully-marked function
block in a FAST diagram
Function Name Best Cost
3. Is anyone buying it for less?

4. Is there something better that can do the job?
5. Can it be made by a less costly method?
6. Can a standard item be used?
7. Can another dependable supplier provide it for less?
8. Would you pay the price if you were spending your own money?
Best cost is not always lower than the current cost. The best cost is
the lowest cost to adequately and reliably provide this function, and it
is possible that this function is not adequately and reliably provided in
the current system. In this case, we might have to increase the cost for
this function. This is also why Question 4 Is there something better
that can do the job? is asked.
The cost and the best cost for functions are also often marked in the
FAST diagram. A fully-marked function block in a FAST diagram has
the format shown in Figure 9.16. Figure 9.17 shows a portion of a FAST
diagram for the youth assistance program with fully-marked function
blocks.
After the FAST diagram is fully developed and the cost-function
worksheet fully filled, it is time for the creative phase of the value-
engineering job plan.
1. 0.0 2. 0.0 3. $ 4.69 4. $3.85 5. $ 6.88

Direct Prescribe Plan Diagnose Collect
Behavior Action Treatment Problem Information
0.0 0.0 $ 4.07 $2.42 $9.42
Figure 9.17 A portion of a FAST diagram with fully-marked function blocks for youth
assistance program
328 Chapter Nine
9.3 Creative Phase

The difference between the cost and the best cost of each function is
the profit-improvement potential, and functions that have high profit-
improvement potentials are the perfect candidates for cost saving. The
creative phase of the value-engineering project uses team members
creativity to develop alternative solutions to perform the functions that
have high profit-improvement potential.
The creative phase is where free development of ideas is fostered.
These ideas will form the basis for concepts that will lead to recom-
mendations for improvement.
9.3.1 Brainstorming
Brainstorming is extremely helpful in the creative phase of value-
engineering projects. It helps to loosen the mental barriers to creativity in
order to create a great volume of ideas. In the beginning of brainstorming,
the quantity of ideas is importantmany obvious wrong ideas will be
generated, but the more ideas that are generated, the better chance that
some really brilliant but not obvious ideas will be among these ideas. The
ideas will be screened and evaluated in the next phase.
In the brainstorming process, an atmosphere is generated that per-
mits each person to lower his or her mental barriers. This may feel
uncomfortable, but it is a necessary stretch that pushes you to think
more freely and more creatively.
The following ground rules for brainstorming must be followed to
create the proper environment for developing ideas:
1. No criticism is allowed during the session.

2. A peer group is desired. Never have high-level management and
their assistants attend.
3. Quantity of ideas is desired. The more ideas, the more likelihood of
at least one outstanding item.
4. A group size of six to ten participants is best.
5. No publicity on the session after its completion.
6. Combine ideas.
7. Wild ideas wanted. Usually the first 90 percent of all ideas will be
those that have come up before.
8. Record all ideas; recording on paper is best.
The first step in brainstorming is to select questions for discussion.

These questions are often selected based on the functions that have
high profit-improvement potentials. The question is often in the form:
TABLE 9.12 Idea Generation Form
What else can perform the function?

Function: Enhance Appearance
1 Use laminate instead of paint.
2 Use stainless steel parts.
3 Use plastic material.
4 Use curves instead of sharp edges.
5 Delete complicated features.
6 Paint parts individually before assembly.
7 Use chrome plating.
8 Use multicolor paint.
9 Use gold material.
What else can perform this function? The questions are then presented
to the group and the group will toss out ideas regarding the question.
Any idea is acceptable and no discussion is allowed. For each question,
it is desirable that at least one new idea should be generated that has
never been thought of before.
Table 9.12 shows an example of ideas generated in a brainstorming
session. The function under discussion is enhance appearance for a
decoration.
9.4 Evaluation Phase

Evaluating the ideas developed during the creative phase is a critical
step in the value-engineering job plan. The ideas generated will include
practical suggestions as well as wild ideas. Each and every idea must
be evaluated without prejudice to determine if it can be used or what
characteristics of the idea may be useful.
Proper evaluation of the ideas is a critical step. Remember, if an idea
is discarded without thorough evaluation, the key to a successful solu-
tion may be lost. The time to create ideas is in the creative phase. If an
idea is discarded, there may not be another opportunity to develop it
again.
During the screening process, it must be kept in mind that the objec-
tive is not to discard ideas but to look for the good in them. All too fre-
quently, a new idea will create a negative reaction. For example, Thats
a great idea but let me tell you what is wrong with it. We should say,
Thats a great idea. What can we do to make it work? There never
seems to be any problem thinking of reasons why something will not
work. However, developing ways to make an idea work takes ingenuity.
330 Chapter Nine
The state of mind during the screening process should be How can we
make it work or what is there about this idea we can use? Evaluation
processes can range from the simple to the complex. The method selected
depends to some degree on the quantity and quality of ideas generated.
The number of ideas can run from less than a hundred to over a thou-
sand depending on the scope of the project. The first screening of the list
should be to eliminate the ones that obviously are of no use to the project.
However, each idea must be reviewed with a positive attitude. Look for
the good rather than the bad and dont be too critical.
9.4.1 Relatively Simple Evaluations

The following process is suggested for initial screening.
Step 1 Each item on the idea generation form will be read out loud
to the team. Each team member will vote whether to keep the idea for
future evaluation or drop it. This is an impulse decision. Each person
will decide by his initial reaction as to whether to keep the idea or not.
However, if one person on the team wants to keep the idea, it must be
kept on the list without question. During this initial screening, there
should be no discussion of the idea; only a yes or no vote is acceptable.
The result will be elimination of the obviously impractical ideas for
this project.
Step 2 Each of the remaining items on the list will be read out and
the group will discuss each idea. Table 9.13 shows a template that can
be used for this step. The intent is to determine what there is about
each idea that may be useful and decide whether to keep it on the list
or to drop it. At this stage, many ideas will combine with other ideas to
form basic groups or categories, such as materials, methods, organiza-
tion, etc. The discussion may also result in new ideas that can be added
to the list.
TABLE 9.13 Idea Screening Worksheet 1
Ideas Implementation Cost Development Cost Total Cost

1.
2.
3.
4.
5.
6.
7.
8.
9.
Step 3 After the initial screening process has been completed, it

will be necessary to resort to systems designed to aid in identifying the
best choice and an alternative, or to rank and weigh alternatives. It is
always important to have a second choice to fall back on just in case
the first choice cannot be implemented for reasons that may not become
apparent until detailed development is underway.
When the initial list of ideas has been reduced to a choice of only a
few alternatives, the simple system illustrated in Table 9.14 may be
used. This sheet identifies the advantages and disadvantages of each
alternative concept. In most cases, an idea listing more advantages
than disadvantages will be the first choice. However, there may be an
overpowering disadvantage that creates a serious roadblock. Can it be
eliminated? If it can, the choice may be clear. If it cannot, the second
alternative may be the best choice.
9.4.2 More Complex Evaluations

There may also be situations where the choice of alternatives will
require more complex systems to aid in the evaluation process. Two
systems that are favored because of their convenience, simplicity,
and effectiveness are Pareto voting and paired comparisons. They
may be used separately or in sequence depending on the situation.
Each of these systems is described in detail in the following sections.
They have been found applicable in a large number of cases and are
extremely useful.
TABLE 9.14 Idea Screening Worksheet 2
Idea 1 Idea 2
Advantage Disadvantage Advantage Disadvantage
Idea 3 Idea 4
Advantage Disadvantage Advantage Disadvantage
332 Chapter Nine
There are also cases involving high risk or a substantial amount of

money where even more detailed analysis is required. These may be
situations where risk is critical and alternatives and trade-offs are nec-
essary. In these cases, a matrix analysis may be necessary.
Experience has shown that this evaluation process is a difficult task.
The impulse to quickly screen the list to zero in on the best ideas must
be controlled. The mass of data must be handled systematically to obtain
maximum benefit from the creative phase. Careful screening is essential
to isolating the best concept to carry over into the planning phase where
the idea will be developed into a practical recommendation for action.
9.4.3 Selection and Screening Techniques

A difficult problem that frequently confronts decision makers is the
need to organize a large amount of data so that one or several of the
most important items may be identified. It may be necessary to deter-
mine which of several alternatives appears to be the best, or it may be
necessary to select a number of items so that they can be ranked and
weighted by order of importance or some other criteria. Most people are
not able to handle this task quickly and effectively.
However, experience has shown that a combination of two simple
methods, Pareto voting and paired comparisons, will satisfy a majority
of requirements. (A literature search identified 13 other methods for
evaluating data to aid in decision making; these are listed in Tables 10.3
and 10.4 in Chapter 10).
9.4.3.1 Pareto Voting Pareto voting is based on Paretos Law of

Maldistribution. Alfredo Pareto (18461923), a political economist, ob-
served a common tendency of wealth and power to be unequally distrib-
uted, and this observation has been refined to the degree that it can be said
that there is an 80/20 percent relationship between similar elements.
For example, 20 percent of the parts in an assembly contain 80 per-
cent of the cost. This is most useful information in cost estimating;
however, the relationship holds for many diverse examples, such as the
following:
Twenty percent of the states use 80 percent of the fuel oil.

Twenty percent of the activities create 80 percent of the budgeted
expense.
Twenty percent of the items sold generate 80 percent of the profit.
In value engineering it is frequently necessary to select the best ideas,

the highest-value functions, the highest-potential projects, or any of a
number of other requirements. It has been found that the application of
Pareto voting can help to simplify the list and will in most cases ensure
that the most important items have been selected. It also produces
results quickly and can be incorporated into the value-engineering pro-
cess to allow continuous operations without undue disruptions.
Pareto voting is conducted by requesting each team member to select
what they believe are the items or elements that have the greatest effect
on the system. This list of items is limited to 20 percent of the total
number of items. For example, each team member would be allowed to
select 6 items out of a list of 30. The vote is on an individual basis to
obtain as much objectivity as possible.
The resultant lists are then compared and arranged into a new con-
solidated list in descending order based on the number of votes each
item received. Usually, several items will have been selected by two
or more team members. The top 10 to 15 items are then ranked and
weighted in a second step by using paired comparisons.
Example 9.7 Pareto Voting This example is based on the idea gener-
ation form that was used in Table 9.12. A team of six people conducted a
Pareto vote on the nine ideas. Each member can only vote for two ideas,
so a total of 12 votes will be received. The number of votes for each idea
will be tallied, and the results are summarized in Table 9.15.
9.4.3.2 Paired Comparisons Paired comparison, or numerical

evaluation, as it is sometimes called, compares a list of items to rank
and weights them in order of importance or some other criteria. Ranking
is the assignment of a preferred order of importance to a list of items.
Weighting is the determination of the relative degree of difference
between items.
TABLE 9.15 Pareto Voting
What else will perform the function?

Rank Function: Enhance Appearance Vote Received
1 Use laminate instead of paint. 5
2 Paint parts individually before assembly. 4
3 Use curves instead of sharp edges. 2
4 Use plastic material. 1
5 Use stainless steel parts. 0
5 Delete complicated features. 0
5 Use chrome plating. 0
5 Use multicolor paint. 0
5 Use gold material. 0
334 Chapter Nine
TABLE 9.16 Pencil Improvement ideas
Key Letter Idea

A Eliminate paint.
B Reduce the length of lead.
C Remove eraser.
D Stain wood in lieu of paint.
E Make body out of paper.
In paired comparisons each item is compared to every other item on

the list in turn, using a simple matrix. It is most convenient for up to
15 items.
A comparative decision is made between any two items on a two-
or three-level basis. In a two-level comparison, 2 = major difference;
1=minor difference. In a three-level comparison, 3 = significant differ-
ence, 2 = moderate difference, 1 = minimal difference.
Example 9.8 illustrates how paired comparisons work.
Example 9.8 Paired Comparison for Pencil Improvement This example

is based on the pencil improvement project presented in Examples 9.3 and
9.6. After some discussion about how to improve the pencil, several ideas
about cost reduction for the pencil are proposed as shown in Table 9.16.
The next step will be evaluating idea A with respect to B, A versus C, and
so on, for all possible pairs. Is A or B the better idea based on cost, benefit,
customer satisfaction, etc? When comparing A and B, a B-2 result indicates
that the team thinks that B is moderately better than A. Similarly, when
comparing A and C, an A-1 result indicates A is minimally better than C.
Table 9.17 summarizes the comparisons for all possible pairs.
After comparing all pairs, all the boxes in Table 9.17 will be filled in. By
adding the values for each idea (for example, A = 1, B = 2+3+2+1= 8), we
get the results shown in Table 9.18. Clearly, B and E are the top choices.
TABLE 9.17 Paired Comparison of Pencil Ideas
B C D E
A B-2 A-1 D-2 E-2
B B-3 B-2 B-1
C D-1 E-3
D E-2
3 Significant
2 Moderate
1 Minimal
TABLE 9.18 Final Evaluation Results for Paired Comparison
Key Letter Idea Value

A Eliminate paint. 1
B Reduce the length of lead. 8
C Remove eraser. 0
D Stain wood in lieu of paint. 3
E Make body out of paper. 7
The whole evaluation phase may go through several screening steps.

Table 9.19 is a convenient template for recording the whole evaluation
phase.
9.5 Planning Phase

After the evaluation phase, we have a final list of ideas that can be
recommended to management for implementation. Now is the time to
develop the best ideas in detail so that recommendations can be made
convincingly. At this stage, we need to determine costs more accurately
and discuss proposed solutions with relevant people. We need to get
the latest material, labor, process and cost data. We must develop a
cooperative atmosphere with everyone able to contribute to a successful
solution, refine the cost of each solution, and determine the best and
alternate recommendation for the performance of basic functions.
For a successful project completion, we also need to determine poten-
tial roadblocks, where they may come up, and how they may be elimi-
nated. Table 9.20 can be used as a template.
TABLE 9.19 Idea Screening Result
What else will perform the function ? First Second Final

Function: Enhance Appearance Screening Screening Screening
1 Use laminate instead of paint.

2 Use stainless steel parts.
3 Use plastic material.
4 Use curves instead of sharp edges.
5 Delete complicated features.
6 Paint parts individually before assembly.
7 Use chrome plating.
8 Use multicolor paint.
9 Use gold material.
336 Chapter Nine
TABLE 9.20 Identify Roadblocks
Best Idea: Reduce the Length of Lead

Roadblock Where/Why Action Required
Differ from Design/out of 1. Show that people seldom use the full
traditional design specification pencil length
practice Marketing/bad 2. Show that good style and low price is
customer image more important to customers
Alternative Idea: Use Body Out of Paper
Roadblock Where/Why Action Required
Effect on strength Design/no previous Show strength/durability test results
and durability experience with this
unknown design
Perceived as a risk Marketing/no idea if Show this new design can make pencil
idea customer will buy in body self-peeling to expose lead, no
need for pencil sharpeners
In the planning phase, it is also very important to discuss how this

project can be sold and implemented. Table 9.21 is a planning form
that can be used to list the names of everyone who will be involved in
accepting and implementing the proposal. We need to figure out possible
problem areas and decide how they can be eliminated.
9.6 Reporting Phase

The object of the value-engineering study is to develop a successful rec-
ommendation for improvement in products, systems, organizations, etc.,
and therefore, in turn, for profits. It must be presented to management
so that it will be accepted and implemented if it is to be worth anything
to you and your company.
Your best recommendation must be prepared for presentation. Before
and after costs and potential savings must be shown and clearly defined.
TABLE 9.21 Action Plan for Selling Value-Engineering Ideas
Department Supervisors Action Required Problem How to Solve Problems

1.
2.
3.
4.
5.
6.
7.
Sketches should show the basic charges in whatever detail is necessary

to prove results. You may need to provide simple models in some cases.
You should list all advantages and disadvantages and show how the
disadvantages were considered in your decision. If the procedure has
been followed, all necessary data should be available in your notes and
records.
The importance of the reporting phase should not be overlooked. If
the recommendations are not presented properly and effectively, a good
idea or an excellent recommendation may not be taken seriously. It is
necessary to present the recommendation in a manner that will clearly
demonstrate its advantages from the standpoint of the organization
required to implement it.
The worksheets used during this chapter have been developed to
provide all of the information necessary to prepare an effective recom-
mendation. It is complete and concise. The next step is to arrange the
material so that it will sell your idea. One of the most important con-
siderations here is to provide complete information.
Failure to provide complete information has been proven to be a major
cause for rejection of a proposal. Persons who are required to review or
approve proposals of one type or another will verify that it is rare for
complete information to be provided.
Keep these points in mind when preparing the final recommendation
and report:
1. Cover all of the facts. Do not skip an important consideration think-

ing that it can be considered later. Do not plan surprises.
2. Justify the recommendation on both technical and economic grounds.
Show the risk and the rewards involved, and show the cost to verify
the idea as well as total lifetime program costs, such as design and
developmental expense, capital investments necessary for buildings,
tools, etc.
3. Indicate the effect on corporate profit, competitive position. or other
important factors.
4. Discuss the proposal with people who will be affected by the idea.
9.7 Implementation Phase

The implementation phase of a value-engineering project deals with
changing the product or service designs, based on the findings of the
value-engineering analysis, and implementing the design changes.
The objective of a value-engineering study is the successful incorpo-
ration of recommendations into the product or operations. However,
the success of a project often depends on the beginning of the study.
338 Chapter Nine
Each project must be thoroughly analyzed at the outset to determine

its potential for benefit and the probability of implementation. This is
as important as the knowledge and skill required to apply the system
to attain successful results.
An excellent idea is worthless unless it can be properly implemented,
and it must be implemented in the manner intended. There are many
cases on record where the idea could not be implemented because of the
high cost of making the change, and other cases where the recommenda-
tions were not properly understood and the implementation resulted in
increased cost. This often results in disillusionment or the feeling that
value engineering doesnt work for those sorts of problems. However,
in most cases this is the result of inefficient preliminary analysis and
preparation.
Selection of projects is a part of the entire value-engineering imple-
mentation process. Often management will assume that any project
will prove profitable. This is not always the case. The project must be
practical in relationship to its effect on the organization. It is not reason-
able to expend effort and funds on a value study without first having
done the necessary work to ensure that the project is practical, that it
can be implemented, and that the necessary funds and manpower will
be available.
To begin with, we will look at the overall organization and the imple-
mentation of value-engineering operations. Then we will look at some
of the details that make for success.
9.7.1 Setting a Goal

What do we want to get from the implementation of value engineering
work? What will be the objective? This is the first question to answer.
The goal should be very specific. Whatever the goal is, it should be
defined in specific terms, such as: increase productivity by a specific per-
cent; reduce product cost by a specific number of dollars per unit, etc.
The goal should be known to everyone. It can be product-oriented or
directed towards manufacturing or administrative operations. It need
not be company-wide. However, the scope can be broadened at any time.
Once the goal has been determined, the means to achieve the objective
can be developed.
9.7.2 Develop An Implementation Plan

The first task in the implementation is to create an implementation leader
or coordinator. The coordinator should work with a trained value-engi-
neering expert. The coordinator and the value-engineering expert should
develop a training program for the implementation team members.
After the training, a carefully designed implementation plan should

be developed jointly by the coordinator, the value-engineering expert,
and team members.
From what we have noted here, it is obvious that the problem is com-
plex from the standpoint of options. However, successful operations do
not have to be extensive. Starting small and developing successfully is
preferred to a lot of noise and a big crash because of poor planning.
One of the most important factors in value engineering is attitude
on the part of management and the people on task teams. A positive,
cooperative, supportive attitude is required.
In most cases whenever a new idea is presented to a group the initial
reaction could be negative. The first remarks are, Its interesting but let
me tell you whats wrong with it. The best approach to this reaction is
to listen carefully. They may have some ideas you overlooked. After all
negative reaction has run out, be prepared to ask some specific positive
questions of the group that will develop positive responses. For example,
I understand your difficulty in producing this part in the plant. What
do you think we would have to do to make this practical? Do you see
any changes we might make to satisfy our methods? This will usually
bring about a positive result.
Never argue. In many cases it is beneficial to solicit negative ideas,
but be prepared to develop positive questions. Our attitude is that we
must begin to ask, Whats good about this idea? How will it help us
to do a better job?
Changing peoples attitudes is difficult and may never happen, but
understanding the reasons behind the negative reaction should make
it possible to persuade most people that they can benefit from success.
Remember, there is a risk of failure in new ideas. New ideas require
change and they may not work. People want proof that a new idea will
work before they will support it. However, you may be able to show that
the benefits are greater than the risk.
9.8 Automobile Dealership Construction

(Park 1999)
A large realty company built and maintained many automobile
dealership facilities all over the country. One major problem faced by
this company was the long time between dealership project authorization
and dealer occupancy. History showed that this long construction cycle
caused tremendous lost sales, so the company wanted to use the value-
engineering technique to shorten this cycle.
At the start of this project, it was found that the average duration
from site selection to leasing was 502 days, or about 1.5 years. A review
of the project process flow chart identified the activities that filled the
340 Chapter Nine
502 days: selecting and obtaining the options on the land, topographical
surveys, soil borings, facility layouts, bid estimates and analysis, budget
reviews, design, construction, and many, many others.
In this value-engineering project, a FAST diagram was developed,
as illustrated in Figure 9.17. Out of the 30 functions, three (resolve
restrictions, obtain data, and construct facility) took 85 percent of the
time. This evaluation was obtained by using time, instead of cost, as a
measure in the FAST diagram.
As a result of project recommendations, the project process procedure
was revised to make it possible to conduct several long-term activi-
ties in parallel with other activities. For example, approval for early
site work was obtained from property owners before ownership was
transferred so that topographic surveys and soil boring could be done
as soon as possible. Standard designs were developed for several parts
of the facility to reduce overall design and development time, and a
single-source contracting procedure was developed to reduce contractor
project interface.
The result of these recommendations was a potential average saving
of 262 days, or a 47 percent average saving in time per project. Based on
the average annual construction program, the yearly benefit in increased
rent would be over $1,250,000 per year. The additional increased vehicle
sales were not included in the benefit.
Select Review
How Vendor Bids
Scope Assure
Competence
Lease Construct Award Secure Prepare Develop

Dealership Facility Purchase Order Approvals Proposal Layout
Establish
Budget
Resolve
Restrictions
When Purchase
Property
Why Scope
Analyse Secure Obtain Assign Issue Obtain

Conditions Option Data Responsibility Authorization Approval
Estimate
Cost
Identify
Requirements
Negotiate
Contract
Figure 9.18 FAST diagram for automobile dealership construction

9.9 Engineering Department Organization

Analysis (Park 1999)
A leading automobile company spent about $200 million dollars per year
and employed 4,000 people, including engineers, designers, technicians,
technical specialists, financial analysts, and so on. An economic down-
turn forced the company to cut costs. Painful lessons had previously been
learned from across-the-board budget cuts, where an equal share of budget
was cut across all departments. The result of that kind of budget cut was
that some vital operations were seriously damaged, and other areas were
okay. This time, the situation was critical; the budget had already been cut
several times, and no one knew where to look next. A value-engineering
project was initiated in order to identify hidden, unnecessary cost.
In this project, with 72 hours of total effort by a team of six people,
a FAST diagram was developed that had 72 functions. The chart was
then thoroughly discussed to ensure that it covered all aspects of the
operation, and a glossary of the functions was made to ensure future
understanding.
Table 9.22 gives a partial list of the functions for this engineering
operation.
Here is a sample from the glossary of functions:
Create Design: To generate a new system, assembly, or component,
measured by time, which includes time to come up with design ideas, design
and layout time, engineers working time, programming time, and so on.
The FAST diagram provided some interesting information. Most
important, it showed that many functions were performed to satisfy
functions outside the scope of engineering responsibilities. Many of
these functions contributed to higher-order functions to support other
company operations, such as the purchasing and legal department.
The next step was to determine how much each function costs and
how funds were distributed among all the functions. To do that, depart-
mental managers were asked to distribute their departmental cost by
function. Cost-function worksheets were filled out. One portion of a
cost-function worksheet is illustrated in Table 9.23.
TABLE 9.22 Partial List of Functions in Engineering Operation
Verb Noun Verb Noun

Create Design Prepare Plan
Transmit Information Negotiate Alternatives
Evaluate Information Evaluate Capabilities
Confirm Design Allocate Resources
Model Concept Appropriate Funds
342
TABLE 9.23 Cost/Function Worksheet
Cost (Hours) Functions

Item No Activity Hours Trans Info Create Design Auth Prog Conf Design Evalu Info Collect Data Make Model
1 Manager 1300 100 60 40 150 40
2 Secretary 1736 1000 60 60
3 Design Supervisor 1438 40 40 60 40 40
4 Engineering 1344 20 100 40 80 80 40
Supervisor
5 Development 2270 160 280
Supervisor
6 Technical Specialist 1078 200 40
7 Sr Design Engineers 2790 40 200 40 160 200 120
8 Sr Develop Engineers 2790 320 1290 560 120
9 Design Engr 11109 280 3360 140 420 560 560
10 Develop Engr 22560 1200 4200 3000 750
11 Design leader 4909 190 480 180 180
12 Technician 22204 8320 1040 780
13 Modeler 17580 4800
14 Clerk 3392 880 800 400
15 Mechanics 26528 4800 8600
16 Material $198,000
Total Material
Total Hours
A partial FAST diagram is illustrated in Figure 9.19.

As you can see in the FAST diagram, 43 percent of the available funds
went to confirm design, and only 14 percent went to create design.
This was considered to be a poor distribution of funds. Team members
thought that this lopsided distribution of funds was a major source
of problems, and across-the-board budget cuts would likely create big
problems in new product design.
It was recognized that confirm design was a required function.
However, changing the way that this function was performed would offer
opportunities for major improvement in productivity, could improve the
overall engineering operation, and could make more efficient use of
engineering funds. The immediate recommendation was to review all
areas involved in confirm design. By following the value-engineering
job plan, major changes were made in several areas that substantially
increased output, cost savings, and avoidance of major capital invest-
ment. Substantial cost reduction was achieved without affecting the
vital design functions.
43.1% total budget

Confirm
8.80% Design 13.6%
Produce Transmit Create Authorise Prepare Establish Utilize
Product Information Design Program Plan Need Knowledge
Confirm Determine
Information Direction
Support
Activity
Figure 9.19 FAST diagram for engineering operation

Chapter
Customer Value Creation Through

10
Creative Design (TRIZ)
In developing any product or service, creative design combined with

the right customer value position will usually bring huge success in
the marketplace. Creative design can make your product the first of
its kind in the marketplace, and it can make your product difficult for
competitors to copy. A company with a good reputation in creativity
can often build a strong brand image, and the brands power further
enhances the customer value of the companys products.
The Theory of Inventive Problem Solving (TRIZ) is an effective
methodology that can help companies and product development people
improve their creativity. The nature of TRIZ is to shortcut the creative
process, and to effectively reuse the knowledge base developed in similar
inventions in order to avoid reinvention. TRIZ is an indispensable tool
for any customer-value-centric company. In this chapter we will cover
the philosophical aspects of TRIZ in order to lay a foundation. Then we
will look at the four-step TRIZ problem definition and solving process,
together with the tools used in TRIZ.
10.1 Theory of Inventive Problem Solving (TRIZ)

TRIZ (Teoriya Resheniya Izobreatatelskikh Zadatch) was developed in
the Soviet Union starting in the late 1940s and in English is known as
the Theory of Inventive Problem Solving (TIPS). TRIZ has been devel-
oped based on more than 1500 person-years of research and study over
many of the worlds most successful solutions of problems from science
and engineering, and systematic analysis of successful patents from
around the world, as well as the study of the psychological aspects of
human creativity.
345
346 Chapter Ten
Dr. Genrich S. Altshuller, the creator of TRIZ, started his investigation of

invention and creativity in 1946. After initially reviewing 200,000 patent
abstracts, Altshuller selected 40,000 as representatives of inventive solu-
tions. He separated the patents into five levels according to their degree
of inventiveness, with Level 1 being the lowest and Level 5 being the
highest. He found that almost all inventions contain at least one contra-
diction, which means that an attempt to improve one feature of a system
detracts from another feature. He found that the level of inventiveness
often depends on how well the contradiction is resolved.
Level 1: Apparent or Conventional Solution Inventions at

Level 1 represent 32 percent of patent inventions and employ obvi-
ous solutions drawn from only a few clear options. These are not real
inventions but narrow extensions or improvements of existing sys-
tems, which are not substantially changed due to the application of
the invention. Usually a particular feature is enhanced or strength-
ened. Examples of Level 1 invention include increasing the thickness
of walls to allow for greater insulation in homes, or increasing the
distance between the front skis on a snowmobile for greater stability.
These solutions may represent good engineering, but contradictions
are not identified and resolved.
Level 2: Small Invention Inside Paradigm Inventions at Level 2
offer small improvements to an existing system by reducing a contra-
diction inherent in the system but still requiring obvious compromises.
These solutions represent 45 percent of the inventions. A Level 2 solu-
tion is usually found through a few hundred trial-and-error attempts,
and it requires knowledge of only a single field of technology. The exist-
ing system is slightly changed, and includes new features that lead to
definite improvements. The new suspension system between the track
drive and the frame of a snowmobile is a Level 2 invention. The use of
an adjustable steering column to increase the range of body types that
can comfortably drive an automobile is another example at this level.
Level 3: Substantial Invention Inside Technology Inventions
at Level 3, which significantly improve the existing system, represent
18 percent of the patents. At this level, a contradiction is resolved within
an existing system, usually through introducing some new elements. The
Level 3 invention may involve hundreds of trials and errors. Examples
of Level 3 invention include automatic transmission of cars. These Level
3 inventions usually borrowed some ideas from other industries, which
are not known in the industry where the invention is applied.
Level 4: Invention Outside Technology Inventions at Level 4
are found in science, not in technology. Such breakthroughs represent
about 4 percent of inventions. Tens of thousands of random trials
Customer Value Creation Through Creative Design (TRIZ) 347
are usually required for these solutions. Level 4 inventions usually

lie outside the technologys normal paradigm and involve using a
completely different principle for the primary function. In Level 4
solutions, the contradiction is eliminated because its existence is
impossible within the new system. That is, Level 4 breakthroughs
use physical effects and phenomena that had previously been little
known within the area. A simple example involves using materials
with thermal memory (shape-memory metals) for a key ring. Instead
of taking a key on or off a steel ring by forcing the ring open, the ring
is placed in hot water. The metal memory causes it to open for easy
replacement of the key. At room temperature, the ring closes.
Level 5: Discovery Inventions at Level 5 exist outside the con-
fines of contemporary scientific knowledge. Such pioneering works
represent less than 1 percent of inventions. These discoveries require
lifetimes of dedication for they involve the investigation of tens of thou-
sands of ideas. This type of solution occurs when a new phenomenon is
discovered and applied to the invention problem. Level 5 inventions,
such as lasers and transistors, create new systems and industries.
Once a Level 5 discovery becomes known, subsequent applications
or inventions occur at one of the four lower levels. For example, the
laser, technological wonder of the 1960s, is now used routinely as a
lecturers pointer and a land surveyors measuring instrument.
Other major findings of TRIZ are also based on extensive studies of
inventions:
A very small number of inventive principles and strategies summarize
most innovations.
Outstanding innovations often feature complete resolution of contra-
dictions, not merely compromising on contradictions.
Outstanding innovations often transform wasteful or harmful ele-
ments in the system to useful resources.
Technological innovation trends are highly predictable.
10.1.1 What Is TRIZ?

TRIZ is a combination of methods, tools, and a way of thinking (Mann
2002). The ultimate goal of TRIZ is to achieve absolute excellence in
design and innovation. In order to achieve absolute excellence, TRIZ
has five key philosophical elements:
Ideality This is the ultimate criterion for system excellence; ideality

is the maximization of the benefits provided by the system, and mini-
mization of the harmful effects and costs associated with the system.
348 Chapter Ten
Functionality This is the fundamental building block of system

analysis; it builds models about how a system works and how it cre-
ates benefit, harm, and costs.
Resource Maximum utilization of resources is one of the keys to
achieving maximum ideality.
Contradictions Contradictions are a common inhibitor for increas-
ing functionality, and removing contradictions usually greatly
increases the functionality and raises the system to a totally new
performance level.
Evolution The evolutionary trend of technological development is
highly predictable, and it can be used to guide further development.
Based on these five key philosophical elements, TRIZ researchers

developed a process of problem-solving. This is a four-step process, con-
sisting of problem definition, problem classification and tool selection,
solution generation, and evaluation.
10.1.1.1 Problem Definition This is a very important step in TRIZ. If you

define the right problem and do it accurately, this represents 90 percent of
the solution. The problem-definition step includes the following tasks:
Function analysis This includes function modeling and analysis of

the system. This is the most important task in the definition step. TRIZ
has very well-developed tools for function modeling and analysis.
Technological evolution analysis This task looks into the relative
technological maturity of all subsystems and parts. If a subsystem or
part is technically too mature, it may reach its limit in performance
and thus become a bottleneck for the whole system.
Ideal final result The ideal final result is the virtual limit of the
system in TRIZ. It may never be achieved but it provides us with an
ultimate dream and will help us to think out of box.
10.1.1.2 Problem Classification and Tool Selection TRIZ has a large

array of tools for inventive problem solving; however, we must select
the right tool for the right problem. In TRIZ, we must first classify the
problem type and then select the tools accordingly.
10.1.1.3 Solution Generation In this step, we apply TRIZ tools to gen-

erate solutions for the problem. Because TRIZ has a rich array of tools,
it is possible to generate many solutions.
10.1.1.4 Evaluation In any engineering project, we need to evaluate the

soundness of the new solution. TRIZ has its own evaluation approach.
However, other non-TRIZ methods might also be used at this stage,

such as axiomatic design and design vulnerability analysis.
10.2 TRIZ Fundamentals

Functionality, use of resources, ideality, contradictions, and evolution
are the pillars of TRIZ. These elements make TRIZ distinctly different
from other innovation and problem-solving strategies. In this section,
we will look at all five elements.
10.2.1 Function Modeling and Analysis

Function modeling and functional analysis originated in value engineer-
ing (Miles 1961). A function is defined as the natural or characteristic
action performed by a product or service.
However, products or services often provide many functions. For exam-
ple, an automobile provides customers with the ability to get from point
A to point B, but it also provides a comfortable riding environment, air
conditioning, music, and so on. There are several types of functions:
Main basic function Among all the functions, the most important
function is called the main basic functionit is the primary purpose
or the most important action performed by the product or service. The
main basic function must always exist, although methods or designs
for achieving the function may vary. For example, the ability to get
from A to B is the main basic function of an automobile.
Besides the main basic function, there are other useful functions as
well; these are the secondary useful functions. There are several kinds
of secondary useful functions:
Secondary basic functions These are not main basic functions,
but customers definitely need them. For example, providing a com-
fortable riding environment is a must-have for automobiles.
Nonbasic but beneficial functions These are functions that pro-
vide customers with esteem value, comfort, and so on. For example,
the paint finish on an automobile provides both basic and nonbasic
functions; it protects the automobile from corrosion and rust, as well
as creating a sleek look for the car.
Besides secondary useful functions, there are two other types of
functions:
Supporting functions These are functions that support the main
basic function or other useful functions. Supporting functions result from
a design approach chosen to achieve the main basic function or other
350 Chapter Ten
useful functions. If the design approach used to achieve the main basic
function and other useful functions is changed, supporting functions may
also change. There are at least two kinds of supporting functions:
Assisting functions These are functions that assist other useful
functions. For example, the engine suspension system provides the
function of locking the position of the engine in the automobile so
the engine can provide power without falling off the car.
Correcting functions These are functions that correct nega-
tive effects of another useful function. For example, the main basic
function of the engine is to provide power for the automobile, but
internal combustion engines also create heat, which is a negative
effectthe water pumps function is to circulate water through the
engine to correct this negative effect. If we change the design and
use electricity as the power source of the automobile, the function of
a water pump will no longer be needed.
Harmful functions These are unwanted, negative functions caused
by the method used to achieve useful functions. For example, an inter-
nal combustion engine not only provides power, but it also generates
noise, heat, and pollution, which are harmful functions.
In summary, the main basic function and secondary useful functions

provide benefits for the customer. Supporting functions are useful, or
at least they are not harmful, but they do not provide benefits directly
to the customer and incur costs. Harmful functions are not useful and
provide no benefits at all.
10.2.1.1 Functional Statement A function can be usually fully described

by a statement consisting of three parts: a subject, a verb, and an object.
For example, an automobiles main basic function can be described as
follows:
Car moves people
(Subject) (Verb) (Object)
A toothbrushs main basic function can be described as
Toothbrush brushes teeth
(Subject) (Verb) (Object)
10.2.1.2 Functional Analysis Diagrams The functional analysis diagram

is a graphical tool used for describing and analyzing functions. The fol-
lowing graph is a typical template for a functional analysis diagram.
Action
and/or
Subject Object
Field
Normal useful action
Insufficient useful action
Excessive useful action Figure 10.1 Arrows for various actions

in functional analysis diagrams
Harmful action
The subject is the source of the action, and the object is the action
receiver. Action is the verb in a functional statement, and it is represented
by an arrow. In technical systems, the action is often accomplished by
applying some kind of field, such as a mechanical, electrical, or chemical
field. For example, the function brush teeth can be described by the
following functional analysis diagram:
Brush
Toothbrush Teeth
Mech.
In this diagram, Mech stands for mechanical field. Clearly, brush-

ing teeth is an application of one kind of mechanical field, force.
In functional analysis diagrams, there are four types of actions, and they
are represented by four types of arrows as illustrated in Figure 10.1.
Example 10.1 Brushing Teeth If we use a toothbrush correctly, and

clean our teeth properly, we call this brush action a normal useful
action. We can illustrate that with the following functional analysis
diagram:
Brush
Toothbrush Teeth
Mech.
However, if we use the toothbrush too gently and do not brush long
enough, or we use a worn-out toothbrush, our teeth will not get enough
cleaning. This is shown in the following functional analysis diagram:
Brush
Toothbrush Teeth
Mech.
Clearly, this is a case of insufficient useful action.

If we use a very strong toothbrush, and brush our teeth with great
force and big strokes, our gums and teeth will be hurt. We can use the
following functional analysis diagram to describe this situation:
Brush
Toothbrush Teeth
Wear
Tear Gums
352 Chapter Ten
SS4 Height adjustment

(Angle-adjusted mirror)
SS3 Focusing device
(Rack and pinion)
SS2 Magnifying lens
SS1 Projection lamp

SS5 Cooling fan
SS6 Power supply regulator
Figure 10.2 Overhead projector
That is, the toothbrush delivers excessive brushing action to the teeth,
which will deliver a harmful action, tear the gums, and make them bleed;
and the teeth also may deliver a harmful action, wear the toothbrush.
Example 10.2 Functional Modeling and Analysis Figure 10.2 shows

a schematic view of an overhead projector.
The following graph is the functional modeling and analysis diagram
for the whole system:
M Focusing adjuster
Hand M
M
Angular adjuster
M
Heat
Focused
E Light Light image
Electrical Projection Image
E Heat Lens Heat Film Mirror Screen
power lamp
E Cold air
Cold air
Cold air
Fan
In this functional analysis graph, E stands for electrical field, and

M stands for mechanical field.
There are many chains of action in this graph; that is, an object can
be another objects subject. This produces a sequence of subject-action-
object-action chains. Each chain describes a complete function. We can
identify the following functions:
From Electrical power to screen is the function to project image

in the film to screen. This is the main basic function.
From Hand to focusing adjustor to mirror is the function to focus

the image. That is a secondary basic function.
From Hand to angular adjuster to mirror is the function to proj-
ect image to right position in screen. That is also a secondary basic
function.
From Electrical power to film is a harmful function chain, which,
without correction, will damage the film and device.
From Electricity to fan to projection lamp, lens, and film is a correcting
function to compensate for the negative effect of the harmful function.
10.2.2 Use of Resources

Using resources effectively is very important in TRIZ. We also need to
make use of resources in creative ways.
The primary mission for any product or process is to deliver functions.
Because substances and fields are basic building blocks of functions,
they are important resources from the TRIZ point of view. However,
substances and fields are not sufficient to build and deliver functions;
space and time are also important and necessary resources. From the
TRIZ point of view, information and knowledge about how to use avail-
able resources are also important resources.
We can divide resources into the following categories:
Substance resources
Raw materials and products
Waste
By-product(s)
System elements
Substance from surrounding environments
Inexpensive substance
Harmful substance from the system
Altered substance from system
Field resources
Energy in the system
Energy from the environment
Energy/field that can be built upon existing energy platforms
Energy/field that can be derived from system waste
Space resources
Empty space
354 Chapter Ten
Space at interfaces of different systems

Space created by vertical arrangement
Space created by nesting arrangement
Space created by rearrangement of existing system elements
Time resources
Pre-work period
Time slot created by efficient scheduling
Time slot created by parallel operation
Post-work period
Information/knowledge resources
Knowledge about all available substances (material properties,
transformations, etc.)
Knowledge about all available fields (field properties, utilizations,
etc.)
Past knowledge
Other peoples knowledge
Knowledge on operation
Functional resources
Unutilized or underutilized existing system main functions
Unutilized or underutilized existing system secondary functions
Unutilized or underutilized existing system harmful functions
In TRIZ, it is more important to look into cheap, ready-to-use, abun-

dant resources than expensive, hard-to-use, and scarce resources, as
demonstrated in Example 10.3.
Example 10.3 Cultivating Fish in Farmland The southeast part of

China is densely populated, so land is a scarce resource. Much land is
used to plant rice. Agriculture experts suggest that farmland can be
used to cultivate fish while it is used to grow rice, because water is a
free and ready resource in rice paddies, and the waste produced by fish
can be used as fertilizer for the rice.
10.2.3 Ideality
Ideality is a measure of excellence. In TRIZ, ideality is defined by the
following equation
Benefits
Ideality = (10-1)
Costs + Harm
Where
Benefits is the sum of the values of the systems useful functions.

Here the supporting functions are not considered useful functions,
because they will not bring benefits to customers directly. Supporting
functions are considered part of the cost to make the system work.
Costs is the sum of the expenses for the system.
Harm is the sum of harms created by harmful functions.
In equation (10-1), a higher ratio indicates a higher ideality. When a

new system is able to achieve a higher ratio than the old system, it is
considered a real improvement.
In TRIZ, there is a law of increasing ideality, which states that the
evolution of all technical systems proceeds in the direction of increasing
ideality. The ideality of the system will increase in the following cases:
Benefits are increased.

Costs are reduced.
Harms are reduced.
Benefits are increased faster than costs and harms.
From the TRIZ point of view, technical systems or products are not
goals in themselves. The real value of the product/system is in its useful
functions. Therefore, the better system is the one that consumes fewer
resources in both initial construction and maintenance.
The ideal final result (IFR) is when the ratio becomes infinite. The
IFR system requires no material, consumes no energy and space, needs
no maintenance, and will not break.
10.2.4 Contradictions
From the TRIZ standpoint, a challenging problem can be expressed as
either a technical contradiction or physical contradiction.
10.2.4.1 Technical Contradictions A technical contradiction is a situ-

ation where efforts to improve technical attributes of a system lead to
deterioration of other technical attributes. For example, as a container
becomes stronger, it becomes heavier. Faster automobile acceleration
reduces fuel efficiency.
A technical contradiction can be resolved either by finding a trade-off
between the contradictory demands, or by overcoming the contradiction.
Trade-off or compromise solutions do not eliminate the technical contra-
dictions, but rather soften them, thus retaining the harmful (undesired)
action or shortcoming in the system.
356 Chapter Ten
Altshuller analyzed thousands of inventions and formulated typi-

cal technical contradictions, such as productivity versus accuracy, reli-
ability versus complexity, shape versus speed, etc. It was discovered
that despite the immense diversity of technological systems and even
greater diversity of inventive problems, there are only about 1250 typi-
cal system contradictions. These contradictions can be expressed as a
table of contradiction of 39 design parameters (see Table 10.1).
From the TRIZ standpoint, overcoming a technical contradiction is
very important both because attributes in the contradiction can be
improved drastically, and the performance of the system will be raised
to a whole new level. TRIZ has many tools for eliminating technical
contradictions.
TABLE 10.1 Thirty-Nine Contradictory Design Parameters
1 Weight of moving object 21 Power
2 Weight of nonmoving object 22 Waste of energy
3 Length of moving object 23 Waste of substance
4 Length of nonmoving object 24 Loss of information
5 Area of moving object 25 Waste of time
6 Area of nonmoving object 26 Amount of substance
7 Volume of moving object 27 Reliability
8 Volume of nonmoving object 28 Accuracy of measurement
9 Speed 29 Accuracy of manufacturing
10 Force 30 Harmful factors acting on object
11 Tension, pressure 31 Harmful side effects
12 Shape 32 Manufacturability
13 Stability of object 33 Convenience of use
14 Strength 34 Repairability
15 Durability of moving object 35 Adaptability
16 Durability of nonmoving object 36 Complexity of device
17 Temperature 37 Complexity of control
18 Brightness 38 Level of automation
19 Energy spent by moving object 39 Productivity
20 Energy spent by nonmoving object

10.2.4.2 Physical Contradictions A physical contradiction means that

a subject or an object has to be in mutually exclusive physical states.
A physical contradiction typically has this pattern: To perform func-
tion F1, the element must have property P, but to perform function F2, it
must have property -P, or the opposite of P. For example, an automobile
has to be light in weight (P) to have high fuel economy (F1), but it also
has to be heavy in weight (-P) in order to be stable for driving (F2).
Example 10.4 Problem: Some buildings are supported by piles. The

pile should have a sharp tip to facilitate the driving process. However,
sharp piles have reduced support capability. For better support capac-
ity, the piles should have blunt ends. However, it is more difficult to
drive a blunt-tipped pile.
Contradiction: A pile should be sharp to facilitate the driving process
and it should be blunt to provide better support of the foundation.
TRIZ Solution: The situation clearly calls for a solution providing
separation of contradictory properties in time. The pile is sharp during
the driving process, and then its base is expanded, which could be real-
ized by a small explosive charge.
During driving After driving
Explosives
Conventional design philosophy is based on compromises (trade-offs).

Contrary to this approach, TRIZ offers several methods of overcoming
physical contradictions completely; these methods will be discussed
thoroughly later in this chapter.
10.2.5 Evolution
TRIZ researchers have found that the trends of evolution of many techni-
cal systems are similar and predictable. They found that many techni-
cal systems will evolve through five stages: pregnancy, infancy, growth,
maturity, and decline. If we plot a time line on the horizontal axis (X-axis),
and plot performance index, level of inventiveness, number of inventions
(relating to the system), and profitability of inventions on the vertical
358 Chapter Ten
axis (Y-axis), we will get the four curves shown in Figure 10.3. Because
the shape of the first curve (performance versus evolutionary stages) has
a S shape, it is also called an S-curve.
10.2.5.1 Pregnancy For a technical system, the pregnancy stage is the

time between an ideas inception and its birth. A new technological sys-
tem emerges only after the following two conditions are satisfied:
There is a need for the function of the system.

There are means (technology) to deliver this function.
The development of an airplane can be used as an example. The

need for the function of airplane, that is, to fly, existed long ago in
many peoples dreams and desires. However, the technical knowledge
of aerodynamics and mechanics was not sufficient for the development
of human flight until the 1800s.
The technologies for the airplane were available since the develop-
ment of glider flight in 1848 and the gasoline engine in 1859. It was the
Wright brothers who successfully integrated both technologies in their
aircraft in 1903and a new technology got off the ground.
10.2.5.2 Infancy The birth of a new technical system is the starting

point of the infancy stage. This is the first stage of an S-curve. The
new system appears as a result of a high-level invention. Typically,
y Decline Decline
urit Maturity
Level of Inventions
Mat
Performance Index
Growth
h
Infancy
t
ow
Gr
Infancy
Time Time
(a) (b)
Profitability of Inventions
Decline Decline
Number of Inventions
Maturity Maturity
Growth Growth
Infancy Infancy
Time Time
(c) (d)
Figure 10.3 Curves of technical system evolution

the system is primitive, inefficient, unreliable, and has many unsolved

problems. It does, however, provide some new functions, or the means
to provide the function.
In the infancy stage, the performance level is low and its improvement
is slow (Figure 10.3a), due to the lack of human and financial resources.
Most people may not be convinced of the usefulness of the system, but a
small number of enthusiasts who believe in the systems future continue
to work towards its success. The level of inventiveness is usually high,
because the initial concept is often very inventive and patentable. It is
usually Level 3, 4, or even 5 (Figure 10.3b). But the number of inven-
tions in this system is usually low (Figure 10.3c), because the system is
fairly new. The profit is usually negative (Figure 10.3d), because at this
stage of the technology the customers are usually few and the expense
is high.
10.2.5.3 Growth (Rapid Development) The growth stage begins when

society recognizes the value of the new system. By this time, many prob-
lems have been overcome; efficiency and performance have improved in
the system, and people and organizations invest money in developing
the new product or process. This accelerates the systems development,
improving the results and in turn, attracting greater investment. Thus,
a positive feedback loop is established, which serves to further acceler-
ate the systems evolution.
In this stage, the improvement of performance is fast (Figure 10.3a)
because of the rapid increase in investment and the removal of many
technical bottlenecks. The level of inventiveness is lower, because
most inventions in this stage are incremental improvements. They
are mostly Level 1 or Level 2 (Figure 10.3b). But the number of inven-
tions is usually high (Figure 103c). The profit is usually growing fast
(Figure 10.3d).
10.2.5.4 Maturity In the maturity stage, system development slows as

the initial concept upon which the system was based nears exhaustion
of its potential. Large amounts of money and labor may be expended,
but the results are usually very marginal. At this stage, standards are
established. Improvements occur through system optimization and
trade-offs. The performance of the system still grows but at a slower
pace (Figure 10.3a). The level of invention is usually low (Figure 10.3b),
but the number of inventions in the form of industrial standards is quite
high (Figure 10.3c). The profitability is usually dropping because of the
saturation of the market and increased competition (Figure 10.3d).
10.2.5.5Decline At this stage, the limits of the technology have been

reached and no fundamental improvement is available. The system
360 Chapter Ten
New System
Infancy Growth Maturity

Performance
Effort (Time)
Figure 10.4 S-curve for two generations of a system
may no longer be needed, because the function provided may no longer

be needed.
It is really important to start the next generation of technical systems
long before the decline stage in order to keep the company from failing.
Figure 10.4 illustrates the S-curves of two successive generations of a
technical system.
10.3 The TRIZ Problem-Solving Process

TRIZ has a four-step problem-solving process. The four steps are prob-
lem definition, problem classification and problem tool selection, prob-
lem solution generation, and problem concept evaluation. We will look
at each step in detail.
10.3.1 Problem Definition

Problem definition is a very important step. The quality of the solution
is highly dependent on problem definition.
Defining the problem starts with several questions:
1. What is the problem?

2. What is the scope of the project?
3. What subsystem, system, and components are involved?
4. Do we have a current solution for the problem? Why is the current
solution not good?
These are common questions that are asked in any engineering project.
By answering these questions, we can define the scope of the project and
focus on the right problem area.
Besides answering these common questions, several TRIZ methods

are also very helpful in the problem definition stage:
Functional modeling and functional analysis
Ideality calculations
S-curve analysis
Contradiction analysis
10.3.1.1 Functional Modeling and Functional Analysis After identifying

the project scope, it is very helpful to establish a functional model of the
subsystem involved in the project. Functional modeling and analysis
enables us to see the problem more clearly and precisely.
We will take another look at the toothbrush example to illustrate how
functional analysis can help the problem definition.
Example 10.5 Toothbrush Problem Revisited Assume that we are a

toothbrush manufacturer and that the current regular toothbrush does
not perform satisfactorily, that is, teeth cannot be adequately cleaned.
We can first draw the following functional diagram:
Brush
Toothbrush Teeth
Mech.
By analyzing the functional diagram, we may come up with the fol-

lowing possibilities:
The current lack of performance may be caused by inadequate action,

that is, the actual functional diagram is the following:
Brush
Toothbrush Teeth
Mech.
If that is the case, it belongs to the problem of inadequate functional

performance; we can use the TRIZ standard solution technique to
resolve this problem.
We may find that the current functional model is too limiting, because
the function statement toothbrush brush teeth limits our solution to
using only a brush and to use mechanical action only. We can develop
the following alternative functional model:
Tooth cleaning Remove

Dirt in teeth
device
The subject toothbrush is replaced by a more general tooth cleaning

device. The object teeth is changed to dirt in teeth, which is more precise.
362 Chapter Ten
The action brush is changed to a more general term remove. Under this
alternative functional modeling, many possible subjects and actions are
open for selection. For example, we can use hydraulic action to clean teeth,
or chemical action to clean teeth. We can even consider pretreatment of
teeth to keep them dirt free, and so on. Clearly this alternative functional
modeling opens a gate for problem solving and innovation.
10.3.1.2 Ideality and Ideal Final Result After functional modeling and
functional analysis, we can evaluate the ideality of the current system:
Ideality = Benefits
Costs + Harm
The ideal final result (the ultimate optimal solution) for a system is
one where
Benefits , and Costs + Harm 0

By comparing the ideality of the current system with the ideal final
result, we can identify where the system improvements should go and
what aspects of the system should be improved. This will definitely
help with the problem definition and identify what problem should be
solved.
10.3.1.3 S-Curve Analysis It is very beneficial to evaluate the evolu-

tionary stage of the current technical system involved in any TRIZ proj-
ect. For example, if the current subsystem is at the growth stage, we
should focus our attention on gradual improvement. If our subsystem is
near maturity, we will know that it is time to develop the next genera-
tion of this subsystem.
10.3.1.4 Contradiction Analysis By using the method described in

Figure 10.4, we can identify whether there are any physical or techni-
cal contradictions in our current system. TRIZ has many methods to
resolve contradictions.
10.3.2 Problem Classification and Tool Selection

After the problem has been defined, we should be able to classify the
problem into one of the following categories. For each category, there are
many TRIZ methods available to resolve the problem:
Physical contradiction
Method: Separation principles
Technical contradiction
Method: Inventive principles
Imperfect functional structures

This problem occurs when
There are inadequate useful functions or a lack of needed useful
functions.
There are excessive harmful functions.
Method: Functional improvement methods

Excessive complexity
This problem occurs when the system is too complex and costly and
some of its functions can be eliminated or combined.
Method: Trimming
System improvement
This problem occurs when the current system is doing its job but
enhancement is needed to beat the competition.
Method: Evolution of technological systems
Develop useful functions
This problem occurs when we can identify what useful functions are
needed to improve the system but we do not know how to create these
functions.
Method: Physical, chemical, and geometric effects database
10.3.3 Problem-Solution Generation

After the problem is classified, there are usually many TRIZ methods
available for solving the problem, so many alternative solutions can be
found. These solutions will be evaluated in the next step.
10.3.4 Problem Concept Evaluation

There are many concept evaluation methods that can be used to evalu-
ate and select the best solution. These methods are often not TRIZ-
related. Frequently used concept evaluation methods include Pugh
concept selection, and value engineering.
10.4 Technical Contradiction Elimination

and Inventive Principles
Genrich Altshuller analyzed more than 40,000 patents and identi-
fied about 1250 typical technical contradictions. These contradictions
are further expressed in a matrix of 39 39 engineering parameters
(contradictions). To resolve these contradictions, Altshuller compiled
364 Chapter Ten
40 principles. Each of 40 principles contains subprinciples, totaling

86 subprinciples.
It should be noted that the 40 principles are formulated in a general
way. If, for example, the contradiction table recommends principle 30,
flexible shell and thin films, it means that the solution of the problem
relates somehow to changing the degree of flexibility or adaptability
of a technical system being modified. The contradiction table and the
40 principles do not offer direct solutions to problems; they only suggest
the most promising directions for searching for a solution. The problem
solver has to interpret these suggestions and find their own way to a
particular situation.
Usually people solve problems by analogical thinking. We try to relate
the problem confronting us to some familiar standard class of problems
(analogs) for which a solution exists. If we draw upon the right analog,
we arrive at a useful solution. Our knowledge of analogous problems is
the result of educational, professional, and life experiences.
What if we encounter a problem for which we have no analog? This
obvious question reveals the shortcomings of the usual approach to
invention problems. The contradiction table and 40 principles offer us
clues to solving the problems with which we are not familiar.
When using the contradiction table and 40 principles, the following
simple procedure will be helpful.
1. Decide on the attribute to be improved, and use one of the 39

parameters in the contradiction table to standardize or model this
attribute.
2. Answer the following questions:
a) How can this attribute be improved using conventional means?
b) Which attribute would suffer if conventional means were used?
3. Select an attribute in the contradiction table corresponding to step 2b.
4. Use the contradiction table, identify the principles in the intersection
of the row (attributes improved) and column (attribute deteriorated)
for overcoming the technical contradiction.
Here are the 40 principles for reference.

Principle 1: Segmentation
Divide an object into independent parts.
Make an object easy to disassemble.

Increase the degree of fragmentation (or segmentation) of an object.
Principle 2: Taking out
Separate an interfering part (or property) from an object, or single

out the only necessary part (or property) of an object.
Principle 3: Local quality
Change an objects structure from uniform to nonuniform; change
an external environment (or external influence) from uniform to

nonuniform.
Make each part of an object function in conditions most suitable for
its operation.
Make each part of an object fulfill a different and useful function.
Principle 4: Asymmetry
Change the shape of an object from symmetrical to asymmetrical.
If an object is asymmetrical, increase its degree of asymmetry.

Principle 5: Merging
Bring closer together (or merge) identical or similar objects; assemble
identical or similar parts to perform parallel operations.

Make operations contiguous or parallel, and bring them together in
time.
Principle 6: Universality
Make a part or object perform multiple functions, to eliminate the
need for other parts.
Principle 7: Nested doll
Place one object inside another, place each object, in turn, inside the
other.
Make one part pass through a cavity in the other.
Principle 8: Anti-weight
To compensate for the weight of an object, merge it with other objects
that provide lift.
To compensate for the weight of an object, make it interact with the
environment (for example, use aerodynamic, hydrodynamic, buoyancy,
and other forces).
Principle 9: Preliminary anti-action
If it will be necessary to do an action with both harmful and useful
effects, this action should be replaced later with anti-actions to control

harmful effects.
Create beforehand stresses in an object that will oppose known unde-
sirable working stresses later on.
366 Chapter Ten
Principle 10: Preliminary action

Perform, before it is needed, the required change of an object (either
fully or partially).
Prearrange objects so that they can come into action from the most
convenient place without losing time for their delivery.
Principle 11: Beforehand cushioning
Prepare emergency means beforehand to compensate for the rela-
tively low reliability of an object.

Principle 12: Equipotentiality
In a potential field, limit position changes (for example, change oper-
ating conditions to eliminate the need to raise or lower objects in a
gravity field).
Principle 13: The other way around
Invert the action(s) used to solve the problem (for example, instead of
cooling an object, heat it).

Make movable parts (or the external environment) fixed, and fixed
parts movable.
Turn the object (or process) upside down.
Principle 14: Spheroidality
Instead of using rectilinear parts, surfaces, or forms, use curvilinear
ones, move from flat surfaces to spherical ones, from parts shaped as
a cube (parallelepiped) to ball-shaped structures.
Use rollers, balls, spirals, domes.
Go from linear to rotary motion, use centrifugal forces.
Principle 15: Dynamics
Allow (or design) the characteristics of an object, external environ-
ment, or process to change them to optimal, or to find an optimal
operating condition.
Divide an object into parts capable of movement relative to each other.
If an object (or process) is rigid or inflexible, make it movable or
adaptive.
Principle 16: Partial or excessive actions
If 100 percent of an effect is hard to achieve using a given solution
method, the problem may be considerably easier to solve by using

slightly less or slightly more of the same method.
Principle 17: Another dimension
Move an object in two- or three-dimensional space.

Use a multistory arrangement of objects instead of a single-story
arrangement.
Tilt or reorient the object; lay it on its side.
Use another side of a given area.
Principle 18: Mechanical vibration
Cause an object to oscillate or vibrate.
Increase its frequency (even up to the ultrasonic).

Use an objects resonance frequency.
Use piezoelectric vibrators instead of mechanical ones.
Use combined ultrasonic and electromagnetic field oscillations.
Principle 19: Periodic action
Instead of continuous action, use periodic or pulsating actions.
If an action is already periodic, change the periodic magnitude or
frequency.
Use pauses between impulses to perform a different action.
Principle 20: Continuity of useful action
Carry on work continuously; make all parts of an object work at full
load, all the time.

Eliminate all idle or intermittent actions or work.
Principle 21: Skipping
Conduct a process, or certain stages (e.g. destructive, harmful, or haz-
ardous operations) at high speed.
Principle 22: Blessing in disguise
Use harmful factors (particularly, harmful effects of the environment
or surroundings) to achieve a positive effect.

Eliminate the primary harmful action by adding it to another harmful
action to resolve the problem.
Amplify a harmful factor to such a degree that it is no longer harmful.
Principle 23: Feedback
Introduce feedback (referring back, cross-checking) to improve a pro-
cess or action.
If feedback is already used, change its magnitude or influence.
Principle 24: Intermediary
368 Chapter Ten
Use an intermediate carrier article or intermediary process.

Merge one object temporarily with another (which can be easily
removed).
Principle 25: Self-service
Make an object serve itself by performing auxiliary helpful functions.
Use waste resources, energy, or substances.

Principle 26: Copying
Instead of an unavailable, expensive, fragile object, use simpler and
inexpensive copies.
Replace an object or process with their optical copies.
If visible optical copies are already used, move to infrared or ultravio-
let copies.
Principle 27: Cheap short-living
Replace an expensive object with a multitude of inexpensive objects,
compromising certain qualities (such as service life, for instance).
Principle 28: Mechanics substitution
Replace a mechanical means with a sensory (optical, acoustic, taste,
or smell) means.
Use electric, magnetic, and electromagnetic fields to interact with the
object.
Change from static to movable fields, from unstructured fields to those
having structure.
Use fields in conjunction with field-activated (for example, ferromag-
netic) particles.
Principle 29: Pneumatics and hydraulics
Use gas and liquid parts of an object instead of solid parts (e.g. inflat-
able, filled with liquids, air cushion, hydrostatic, hydro-reactive).

Principle 30: Flexible shells and thin films
Use flexible shells and thin films instead of three-dimensional
structures.
Isolate the object from the external environment using flexible shells
and thin films.
Principle 31: Porous materials
Make an object porous or add porous elements (inserts, coatings, etc.).
If an object is already porous, use the pores to introduce a useful
substance or function.
Principle 32: Color changes

Change the color of an object or its external environment.
Change the transparency of an object or its external environment.

Principle 33: Homogeneity
Make objects interact with a given object of the same material (or a
material with identical properties).

Principle 34: Discarding and recovering
Make portions of an object that have fulfilled their function go away
(discard by dissolving, evaporating, etc.) or modify these directly
during operation.
Conversely, restore consumable parts of an object directly during
operation.
Principle 35: Parameter changes
Change an objects physical state (e.g. to a gas, liquid, or solid).
Change the concentration or consistency.
Change the degree of flexibility.
Change the temperature.
Principle 36: Phase transitions
Use phenomena occurring during phase transitions (e.g. volume
changes, loss or absorption of heat, etc.).

Principle 37: Thermal expansion
Use thermal expansion (or contraction) of materials.
If thermal expansion is being used, use multiple materials with dif-

ferent coefficients of thermal expansion.
Principle 38: Strong oxidants
Replace common air with oxygen-enriched air.
Replace enriched air with pure oxygen.
Expose air or oxygen to ionizing radiation.
Use ozonized oxygen.
Replace ozonized (or ionized) oxygen with ozone.
Principle 39: Inert atmosphere
Replace a normal environment with an inert one.
Add neutral parts, or inert additives to an object.
Principle 40: Composite materials
370 Chapter Ten
Change from uniform to composite (multiple) materials.

Example 10.6 Using 40 Principles and Contradiction Matrix to
Improve Wrench Design When we use a conventional wrench to undo
an overtightened or corroded nut (as shown in the following image),
one of the problems is that the corners of the nut take a concentrated
load and may wear out quickly. You can reduce the clearance between
wrench and nut, but it will be difficult to fit the wrench onto the nut. Is
there anything we can do to solve this problem?
It seems obvious that we want to reduce the clearance between the

wrench and nut to improve reliability, but this leads to the deterioration
of operationsthe wrench is harder to use. From the TRIZ standpoint, a
technical contradiction is present when a useful action simultaneously
causes a harmful action.
A problem associated with a technical contradiction can be resolved
either by finding a trade-off between the contradictory demands, or by
overcoming the contradiction. Trade-off or compromise solutions do not
eliminate the technical contradictions, but rather soften them, thus
retaining harmful (undesired) actions or shortcomings in the system.
An engineering problem becomes an inventive one when the technical
contradiction cannot be overcome by conventional means and trade-off
solutions are not acceptable. The 40 principles and the contradiction
matrix are important tools for overcoming contradictions.
1. Build contradiction model.

Look into the problems and find a pair of contradictions. The contradic-
tions should be described using two of the 39 parameters for technical
contradictions. In this problem, the contradictions are:
Things we want to improve: Reliability (parameter 27)
Things are getting worse: Ease of operation (parameter 33)
2. Check contradiction matrix.

Locate the parameter to be improved in the row, and the parameter to be
deteriorated in the column in the contradiction matrix for inventive prin-
TABLE 10.2 Partial Contradiction Matrix
What is deteriorated ?
31. Harmful effect caused by the object

29. Manufacturing precision
30. Harmful action at object

28. Measurement accuracy
26. Quantity of substance
32. Ease of manufacture
33. Ease of operation

What should be
34. Ease of repair

25. Waste of time
improved?
35. Adaptation
27. Reliability
25. Waste of time 35 10 24 24 35 35 35 4 32 35

38 30 34 26 18 22 28 28 1 28
18 4 28 28 34 18 34 10 10
16 32 18 39 4 34
26. Quantity of 35 18 3 33 35 3 29 35 2 15
substance 38 3 2 30 33 35 1 29 32 3
18 28 28 29 40 35 10 10 29
16 40 31 39 27 25 25
27. Reliability 10 21 32 11 27 35 27 1 13
30 28 3 32 35 2 17 11 35
4 40 11 1 2 40 40 8
3 23 40 26 24
28. Measurement 24 2 5 28 3 6 1 1 13
accuracy 34 6 11 24 33 35 13 32 35
28 32 1 22 39 25 17 13 2
32 23 26 10 18 34 11
29. Manufacturing 32 32 11 26 4 1 25
precision 26 30 32 28 17 32 10
28 1 10 34 35
18 36 26 23
ciples. The matrix offers the following principles: 27, 17, and 40 (see the
partial matrix that follows; the full matrix is shown in Tables 10.3ac).
3. Interpret principles.
Read each principle and construct analogies between the concepts
of principle and your situation, then create solutions to your problem.
In this case, Principle 17 (another dimension) indicates that the wrench
372
TABLE 10.3a Contradiction Table of Inventive Principles 113 (continued)
13. Objects composition stability

What is deteriorated ?
7. Volume of movable object

3. Length of movable object
1. Weight of movable object
5. Area of movable object
8. Volume of fixed object

4. Length of fixed object
2. Weight of fixed object
6. Area of fixed object
11. Stress, pressure
12. Shape
10. Force
9. Speed
What should be
improved?
1. Weight of movable 15 8 29 17 29 2 2 8 8 10 10 36 10 14 1 35
object 29 34 38 34 40 28 15 38 18 37 37 40 35 40 19 39
2. Weight of fixed object 10 1 35 30 5 35 8 10 13 29 13 10 26 39
29 35 13 2 14 2 19 35 10 18 29 14 1 40
3. Length of movable 8 15 15 7 17 13 4 17 10 1 8 1 8 1 8
object 29 34 17 4 4 35 8 4 35 10 29 15 34
4. Length of fixed object 35 28 17 7 35 8 28 10 1 14 13 14 39 37
40 29 10 40 2 14 35 15 7 35
5. Area of movable object 2 17 14 15 7 14 29 30 19 30 10 15 5 34 11 2
29 4 18 4 17 4 4 34 35 2 36 28 29 4 13 39
6. Area of fixed object 30 2 26 7 1 18 10 15 2 38
14 18 9 39 35 36 36 37
7. Volume of movable 2 26 1 7 1 7 29 4 15 35 6 35 1 15 28 10
object 29 40 35 4 4 17 38 34 36 37 36 37 29 4 1 39
8. Volume of fixed object 35 10 19 14 35 8 2 18 24 35 7 2 34 28
19 14 2 14 37 35 35 40
9. Speed 2 28 13 14 29 30 7 29 13 28 6 18 35 15 28 33
13 38 8 34 34 15 19 38 40 18 34 1 18
10. Force 8 1 18 13 17 19 28 10 19 10 1 18 15 9 2 36 13 28 18 21 10 35 35 10
37 18 1 28 9 36 15 36 37 12 37 18 37 15 12 11 40 34 21
11. Stress, pressure 10 36 13 29 35 10 35 1 10 15 10 15 6 35 35 24 6 35 36 35 35 4 35 33
37 40 10 18 36 14 16 36 28 36 37 10 36 21 15 10 2 40
12. Shape 8 10 15 10 29 34 13 14 5 34 14 4 7 2 35 15 35 10 34 15 33 1
29 40 26 3 5 4 10 7 4 10 15 22 35 34 18 37 40 10 14 18 4
13. Objects composition 21 35 26 39 13 15 37 2 11 39 28 10 34 28 33 15 10 35 2 35 22 1
stability 2 39 1 40 1 28 13 19 39 35 40 28 18 21 16 40 18 4
14. Strength 1 8 40 26 1 15 15 14 3 34 9 40 10 15 9 14 8 13 10 18 10 3 10 30 13 17
40 15 27 1 8 35 28 26 40 29 28 14 7 17 15 26 14 3 14 18 40 35 40 35
15. Duration of moving 19 5 2 19 3 17 10 2 3 35 19 2 19 3 14 26 13 3
objects operation 34 31 9 19 19 30 5 16 27 28 25 35
16. Duration of fixed 6 27 1 40 35 34 39 3
objects operation 19 16 35 38 35 23
17. Temperature 36 22 22 35 15 19 15 19 3 35 35 38 34 39 35 2 28 35 10 35 39 14 22 1 35
6 38 32 9 9 39 18 40 18 6 4 36 30 3 21 19 2 19 32 32
18. Illumination 19 1 2 35 19 32 19 32 2 13 10 13 26 19 32 30 32 3
32 32 16 26 10 19 6 27
19. Energy expense of 12 18 12 28 15 19 35 13 8 15 16 26 23 14 12 2 19 13
movable object 28 31 25 18 35 21 2 25 29 17 24
20. Energy expense of 19 9 36 37 27 4
fixed object 6 27 29 18
373
374
TABLE 10.3b Contradiction Table of Inventive Principles 1425 (continued)
15. Duration of moving
24. Loss of information

deteriorated ?
19. Energy expense of
20. Energy expense of

objects operation
objects operation
23. Loss of substance

16. Duration of fixed
26. Quantity of sub-

22. Waste of energy
movable object
25. Waste of time

What is
17. Temperature
18. Illumination
fixed object
14. Strength
21. Power
stance
What should be improved?
1. Weight of movable object 28 27 5 34 6 29 19 1 25 12 13 36 6 2 5 35 10 24 10 35 3 26

18 40 31 35 4 38 32 34 31 18 31 34 19 3 31 35 20 28 18 31
2. Weight of fixed object 28 2 2 27 28 19 35 19 18 19 15 19 18 19 5 8 10 15 10 20 19 6
10 27 19 6 32 22 35 28 1 18 22 28 15 13 30 35 35 26 18 26
3. Length of movable object 8 35 19 10 15 32 8 35 1 35 7 2 4 29 1 24 15 2 29 35
29 34 19 24 35 39 23 10 29
4. Length of fixed object 15 14 1 40 3 35 3 25 12 8 6 28 10 28 24 26 30 29
28 26 35 38 18 24 35 14
5. Area of movable object 3 15 6 3 2 15 15 32 19 32 19 10 15 17 10 35 30 26 26 4 29 30
40 14 16 19 13 32 18 30 26 2 39 6 13
6. Area of fixed object 40 2 10 35 39 17 32 17 7 10 14 30 16 10 35 2 18
19 30 38 30 18 39 4 18 40 4
7. Volume of movable 9 14 6 35 34 39 10 35 35 6 7 15 36 39 2 22 2 6 29
object 15 7 4 10 18 13 2 13 18 13 16 34 10 34 10 30 7
8. Volume of fixed object 9 14 35 34 35 30 6 10 39 35 16 35 3
17 15 38 6 4 35 34 32 18
9. Speed 8 3 3 19 28 30 10 13 8 15 19 35 14 20 10 13 13 26 10 19
26 14 35 5 36 2 19 35 38 38 2 19 35 28 38 29 38
10. Force 35 10 19 2 35 10 19 17 1 16 19 35 14 15 8 35 10 37 14 29
14 27 21 10 36 37 18 37 40 5 36 18 36
11. Stress, pressure 9 18 19 3 35 39 14 24 10 35 2 36 10 36 37 10 14
3 40 27 19 2 10 37 14 25 3 37 36 4 36
12. Shape 30 14 14 26 22 14 13 15 2 6 4 6 2 14 35 29 14 10 36 22
10 40 9 25 19 32 32 34 14 3 5 34 17
13. Objects composition 17 9 13 27 39 3 35 1 32 3 13 19 27 4 32 35 14 2 2 14 35 27 15 32
stability 15 10 35 35 23 32 27 15 29 18 27 31 39 6 30 40 35
14. Strength 27 3 30 10 35 19 19 35 35 10 26 35 35 28 29 3 29 10
26 40 10 35 28 31 40 28 10 27
15. Duration of moving 27 3 19 35 2 19 28 6 19 10 28 27 10 20 10 3 35
objects operation 10 39 4 35 35 18 35 38 3 18 28 18 10 40
16. Duration of fixed 19 18 16 27 16 10 28 20 3 35
objects operation 36 40 18 38 10 16 31
17. Temperature 10 30 19 3 19 18 32 30 19 15 2 14 21 17 21 36 35 28 3 17
22 40 39 36 40 21 16 3 17 17 25 35 38 29 31 21 18 30 39
18. Illumination 35 19 2 19 32 35 32 1 32 35 32 19 16 13 1 1 6 19 1 1 19
6 19 19 1 15 1 6 26 17
19. Energy expense of 5 19 28 35 19 24 2 15 6 19 12 22 35 24 35 38 34 23
movable object 9 35 6 18 3 14 19 37 18 15 24 18 5 19 18 16 18
20. Energy expense of fixed 35 19 2 28 27 3 35
object 35 32 18 31 31
375
376
TABLE 10.3c Contradiction Table of Inventive Principles 27-39 (continued)
caused by the object
36. Device complexity
38. Degree of automa-

deteriorated ?
30. Harmful action at
33. Ease of operation
37. Measurement or
test complexity
29. Manufacturing
31. Harmful effect
34. Ease of repair

28. Measurement
What is
manufacture
39. Productivity
35. Adaptation
27. Reliability
precision
accuracy
32. Ease of
object
tion
What should be improved?
1. Weight of movable object 3 11 28 27 28 35 22 21 22 35 27 28 35 3 2 27 29 5 26 30 28 29 26 35 35 3
1 27 35 26 26 18 18 27 31 39 1 36 2 24 28 11 15 8 36 34 26 32 18 19 24 37
2. Weight of fixed object 10 28 18 26 10 1 2 19 35 22 28 6 13 2 27 19 15 1 10 25 28 2 26 1 28
8 3 28 35 17 22 37 1 39 1 9 1 32 28 11 29 26 39 17 15 35 15 35
3. Length of movable object 10 14 28 10 28 1 15 17 15 1 29 15 29 1 28 14 15 1 19 35 1 17 24 14 4
29 40 32 4 29 37 17 24 17 35 4 10 1 16 26 24 26 24 26 16 28 29
4. Length of fixed object 15 29 32 2 32 1 18 15 17 2 25 3 1 35 1 26 26 30 14
28 28 3 10 27 7 26
5. Area of movable object 29 9 26 28 2 32 22 33 17 2 13 1 15 17 15 13 15 30 14 1 2 36 14 30 10 26
32 3 28 1 18 39 26 24 13 16 10 1 13 26 18 28 23 34 2
6. Area of fixed object 32 35 26 28 2 29 27 2 22 1 40 16 16 4 16 15 16 1 18 2 35 23 10 15
40 4 32 3 18 36 39 35 40 36 30 18 17 7
7. Volume of movable object 14 1 25 26 25 28 22 21 17 2 29 1 15 13 10 15 29 26 1 29 35 34 10 6 2
40 11 28 2 16 27 35 40 1 40 30 12 26 4 16 24 34
8. Volume of fixed object 2 35 35 10 34 39 30 18 35 1 1 31 2 17 35 37
16 25 19 27 35 4 26 10 2
9. Speed 11 35 28 32 10 28 1 28 2 24 35 13 32 28 34 2 15 10 10 28 3 34 10 18
27 28 1 24 32 25 35 23 35 21 8 1 13 12 28 27 26 4 34 27 16
10. Force 3 35 35 10 28 29 1 35 13 3 15 37 1 28 15 1 15 17 26 35 36 37 2 35 3 28
13 21 23 24 37 36 40 18 36 24 18 1 3 25 11 18 20 10 18 10 19 35 37
11. Stress, pressure 10 13 6 28 3 35 22 2 2 33 1 35 11 2 35 19 1 2 36 35 24 10 14
19 35 25 37 27 18 16 35 37 35 37
12. Shape 10 40 28 32 30 22 1 35 1 1 32 32 15 2 13 1 15 16 29 15 13 15 1 17 26
16 32 1 40 2 35 17 28 26 1 29 1 28 39 32 34 10
13. Objects composition 13 18 35 24 35 40 35 19 32 35 2 15 35 30 2 35 35 22 1 8 23 35
stability 18 30 27 39 30 10 16 34 2 22 26 39 23 35 40 3
14. Strength 11 3 3 27 3 27 18 35 15 35 11 3 32 40 27 15 3 2 13 27 3 15 29 35
16 37 1 22 2 10 32 28 2 11 3 32 28 15 40 10 14
15. Duration of moving 11 2 3 3 27 22 15 21 39 27 12 27 29 10 1 35 10 4 19 29 6 10 35 17
objects operation 13 16 40 33 28 16 22 1 4 27 13 29 35 39 35 14 19
16. Duration of fixed objects 34 27 10 26 17 1 22 35 10 1 1 2 25 14 1 20 10
operation 6 40 24 40 33 6 35 16 38
17. Temperature 19 35 32 19 24 22 33 22 35 26 27 26 27 4 10 2 18 2 17 3 27 26 2 15 28
3 10 24 35 2 2 24 16 27 16 35 31 19 16 35
18. Illumination 11 15 3 32 15 19 35 19 19 35 28 26 15 17 15 1 6 32 32 15 2 26 2 25
32 32 39 28 26 19 13 16 19 13 10 16
19. Energy expense of 19 21 3 1 1 35 2 35 28 26 19 35 1 15 15 17 2 29 35 38 32 2 12 28
movable object 11 27 32 6 27 6 30 17 28 13 16 27 28 35
20. Energy expense of fixed 10 36 10 2 19 22 1 4 19 35 1 6
object 23 22 37 18 16 25
377
378 Chapter Ten
problem may be resolved if you move an object in two or three-dimen-

sional space or use a different side of the given area. From Principle 27
(cheaper short-living) and Principle 40 (composition material), we may
replace an expensive object with a multitude of inexpensive objects
and change from uniform material to composite material.
4. Resolve the problem.
The working surface of a wrench can be redesigned in nonuniform shape
by applying Principle 17 (see the illustration below). Principles 27 and
40 can be used together: Attach soft metal or plastic pads on the wrenchs
working surfaces when tightening or undoing expensive nuts.
Table 10.3(ac) is the complete contradiction matrix, which can be

used to select the most appropriate inventive principles.
Chapter
Statistical Basics and Six

11
Sigma Metrics
Six Sigma is a data-driven management system with near-perfect-

performance objectives (Pande et al. 1998). In this context, data-driven
means that in Six Sigma, the real data collected in the process under study
is the only source for measuring the current performance, analyzing the
root causes of the problem, and deriving improvement strategies. Near-
perfect-performance objectives means that in Six Sigma, the process will
be continuously improved until it achieves very low levels of defects and
very high levels of performance. Clearly, it also needs real data from the
process to verify that the desired performance requirements are met.
The theoretical basis for modern data analysis is statistics, and there
are different methods in statistics that can be used to analyze the data.
Some of them are very simple, such as descriptive statistics, which can
provide intuitive display and analysis of the data. Some of them are
more sophisticated, such as probability distribution models and statisti-
cal inferences; these analyses are more powerful and they can provide
more insights and provide credible inferences and predictions about
processes. All popular Six Sigma performance metrics are based on the
theory of statistics, so familiarity with basic statistics is essential in
understanding Six Sigma metrics.
In this chapter, we will review several descriptive statistical methods
and several commonly-used probability distribution models, as well as
Six Sigma metrics.
11.1 Six Sigma and Data Analysis

Data analysis is a very important part of Six Sigma. In real business
and engineering processes, much of the data that we collect consists
379
380 Chapter Eleven
of random variables, that is, their value will vary with some degree of
uncertainty. Consider the case in Example 11.1.
Example 11.1 A Data Set from a Semiconductor Manufacturing

Process In a semiconductor manufacturing process, we have a step
where an oxide film is grown on a silicon wafer in a furnace. In this step,
a cassette of wafers is placed in a quartz boat and the boats are placed
in the furnace. A gas flow is created in the furnace and it is brought
up to temperature and held there for a specified period of time. In this
process, the most desirable oxide film thickness is 560 Angstroms, and
the specification of the oxide thickness is 560 100 Angstroms; that is,
an oxidized wafer is out of specification if its thickness is either lower
than 460 angstroms or higher than 660 Angstroms.
We collected the following film thickness data in the process:
547 563 578 571 572 575 584 549 546 584 593 567 548 606 607 539 554 533 535
522 521 547 550 610 592 587 587 572 612 566 563 569 609 558 555 577 579 552
558 595 583 599 602 598 616 580 575
How well does this process satisfy our quality requirement?

In Example 11.1, the film thickness clearly varies from wafer to wafer,
so it is a random variable. A random variable can be either discrete or
continuous. If the set of all possible values is finite or countably infinite,
the random variable is discrete; if the set of all possible values of the
random variable is an interval, then the random variable is continuous.
Clearly, the film thickness data is a continuous variable.
11.2 Descriptive Statistics

Descriptive statistics are a set of simple graphical and numerical meth-
ods that can quickly display some intuitive properties found in data.
The commonly-used graphical descriptive statistical methods include
dot plots, histograms, and box plots.
11.2.1 Dot Plot

The dot plot, as illustrated in Figure 11.1, is a simple yet effective dia-
gram in which each dot represents a data point from Example 11.1. Dot
plots can display distribution patterns and the spread of data points.
11.2.2 Histogram
A histogram is a diagram displaying a frequency distribution. In a his-
togram, the horizontal axis is partitioned into many small segments,
Statistical Basics and Six Sigma Metrics 381
Dot Plot of Film Thickness
532 546 560 574 588 602 616

Film thickness
Figure 11.1 Dot Plot of film thickness data
and the number of data points (or the percentage of points) that fall in
each segment is called the frequency. This determines the height of the
bar for that segment.
For example, the histogram for the film thickness data in Example
11.1 is displayed in Figure 11.2. The leftmost segment is the bracket
(515 to 525), and in the data set, there are two data points (521 and
522) in this range, so the height of the bar is 2. We can see that a large
portion of the data falls in the 545 to 585 segment.
Histogram of Film Thickness
9
8
7
6
Frequency
5
4
3
2
1
0
520 540 560 580 600 620
Film thickness
Figure 11.2 Histogram of film thickness data

382 Chapter Eleven
11.2.3 Box Plot

A box plot is another very useful way of displaying data. A box plot will
display the following aspects of the data:
Minimum value (lowermost point of the vertical line)

Maximum value (uppermost point of the vertical line)
Median value (horizontal center line)
Twenty-fifth percentile (lower bar of the box)
Seventy-fifth percentile (upper bar of the box)
Figure 11.3 shows the box plot of the data in Example 11.1. The cen-
terline of the box corresponds to 572, which is the median of the data.
The lower bar of the box corresponds to 552, which is the 25th percentile
of the data, and the upper bar corresponds to 592, which is the 75th
percentile of the data. The uppermost point corresponds to 616, which
is the maximum of the data, and the lowermost point corresponds to
521, which is the minimum of the data.
11.2.4 Numerical Descriptive Statistics

Numerical descriptive statistics are numbers calculated from a data set
in order to help us create a mental image of the distribution pattern of
the data. There are three types of numerical descriptive statistics:
1. The numerical measure that describes the central tendency of the

data, that is, where the center of the data set is. Frequently used
measures of central tendency include the mean and median.
Box Plot of Film Thickness

620
600
Film thickness
580
560
540
520
Figure 11.3 Box Plot of film thickness data

2. The numerical measures that describe the spread of data, also called
the measures of variation. The frequently used measures of varia-
tion include variance, standard deviation, range, maximum, and
minimum.
3. The numerical measures that describe the relative position of the
data, often called the measures of relative standing. The frequently
used measure here includes the percentile points.
11.2.4.1 Measures of Central Tendency The mean and median are the
most frequently used measures for central tendencies. I will define both
measures and give numerical examples in the following sections.
Mean The mean, y , is also called the arithmetic mean. For a set of
n measurements, y1 , y2 ,..., yn, y is the average of the measurements,
specifically:
1 n
y = yi (11.1)
n i=1
Example 11.2 The Calculation of Mean The mean for the data in
Example 11.1 is
1
y= (547 + 563 + .... + 575) = 572.02.
47
Clearly, the mean is simply a numerical average, which gives a good

sense of where the center is for a data set. It is the most commonly
used measure of central tendency. However, in some cases, it is not a
preferred measure. For example, assume that in a subdivision, there
are 20 families. Most of the families have an annual income around
$40,000, but there is one family with an annual income of $1,000,000. If
we use the arithmetic mean as the measure of central tendency, y will
be approximately $90,000, which is by no means a typical income in this
circumstance. The median will be a better measure in this case.
Median The median of a set of measurements, y1 , y2 ,..., yn , is the

middle number when the measurements are arranged in ascending (or
descending) order. Specifically, let y( i) denote the ith value of the data
set when y1 , y2 ,..., yn are arranged in ascending order; then the median
m is the following:
y[( n+1)/ 2] if n is odd
m = y( n / 2) + y( n / 2+1) if n is even. (11.2)

2
384 Chapter Eleven
Example 11.3 The Calculation of Median The median of the data

set in Example 11.1 can be calculated by first arranging the data in
ascending order:
y(1) , y( 2) ,..., y( n1) , y( n) = 521, 522, 533, , 610, 612, 616
In this data set, n = 47, which is an odd number, so n + 1/2 = 48/2 =

24, and y24 = 572
Therefore, m = 572.
11.2.4.2 Measures of Variation The most commonly used measures of

variation are the range, the variance, and the standard deviation.
Range The range is equal to the difference between the largest

(maximum) and the smallest (minimum) measurements in a data set:
Range = Maximum Minimum (11.3)
Example 11.4 The Calculation of Range For the data set in Exam-
ple 11.1, the maximum = y( n) = y( 47) = 616, and the minimum = y(1) = 521.
Therefore, Range = 616-521 = 95.
The range is very easy to compute, but it only gives the distance
between the two most extreme observations. It is not a good measure
of variation for the whole data set. Variance and standard deviation are
better measures in this respect.
Variance The variance of a sample of n measurements, y1 , y2 ,..., yn ,

is defined as:
1 n
n 1
s2 = ( yi y)2 . (11.4)
i =1
Example 11.5 The Calculation of Variance For the data set of

Example 11.1, the variance can be computed as:
1
s2 = [(547 572.02)2 + (563 572.02)2 + ... + (575 572.02)2 ]
47 1
= 601.72
Sample variance, s2, is obviously an average of the sum of the squared

deviation from the mean of all observations. Squared deviation makes
sense because whether an observation is smaller or larger than the
mean, the squared deviation will always be positive. The average of

squared deviation is a measure of variation for the whole data set. The
drawback to variance, however, is that the measurement unit of vari-
ance is the square of the measurement unit of the original data. So if the
original data is length in inches, the variance will be in square inches,
which cannot be compared with the original data.
Standard Deviation The standard deviation is the square root of

the variance. Specifically, the standard deviation of a sample of n mea-
surements, y1 , y2 ,..., yn , is defined as:
1 n
n 1
s = s2 = ( yi y)2 . (11.5)
i=1
One advantage of standard deviation over variance is that the stan-

dard deviation has the same measurement scale as that of the mean.
Also, the spread of data is usually measured by the mean and standard
deviation.
Example 11.6 The Calculation of Standard Deviation For the data

in Example 11.1, the standard deviation can be computed as:
s = s2 = 601.72 = 24.53.
11.2.4.3 Measure of Relative Standing The measure of relative stand-

ing provides a numerical value or score that describes a predefined
location relative to other observations in a data set. A very commonly
used measure of relative standing is the 100pth percentile, often simply
called percentile points.
The 100pth percentile of a data set is a value y located so that 100p
percent of the data is smaller than y, and 100(1-p) percent of data is
larger than y, where 0 p 1.
Example 11.7 The Calculation of Percentile The median is the 50th

percentile, because 50 percent of the data points are smaller than the
median, and 50 percent of the data points are larger. The 25th percen-
tile is often called lower quartile, QL, or Q1. Twenty-five percent of the
data points will be smaller than QL, and 75 percent of the data points
will be larger in a given data set. The 75th percentile is often called the
upper quartile, denoted by QU or Q3. Seventy-five percent of the data
points will be smaller than QU, and 25 percent of the data will be larger
than QU.
386 Chapter Eleven
MINITAB is a popular statistical software that can compute all types

of descriptive statistics conveniently. The following MINITAB output is
the descriptive statistics for the data set in Example 11.1.
Descriptive Statistics: Film Thickness
Variable N N* Mean SE Mean StDev Minimum Q1 Median Q3
Film Thickness 47 0 572.02 3.58 24.53 521.00 552.00 572.00 592.00
or
Variable Maximum
Film Thickness 616.00
11.3 Random Variables and Probability

Distributions
The data set collected in a process, such as the data set described in
Example 11.1, is called a sample of data, because it only reflects a snap-
shot of the process. For example, the data set in Example 11.1 is only a
small portion of production data. If we were able to collect all the film
thickness data for all wafers in the whole life cycle of the oxidation fur-
nace, we would have collected a whole population of data.
In real-world business decision making, the population is of more
interest for the decision makers. The decision makers are definitely
more interested in the overall quality level for the population. Random
variables and probability distributions are the mathematical tools used
to describe the behavior of populations.
11.3.1 Discrete and Continuous

Random Variables
A random variable can be defined as a variable that takes different
values according to some specific probability distribution. A random
variable can be a discrete random variable if it can take only a count-
able number of values. A simple example of a discrete random variable
is the values from rolling a fair six-faced die: it can only be 1, 2, 3, 4, 5,
6, with equal probabilities. A random variable can also be a continu-
ous random variable, if it can take all real values in a given interval.
For example, the height of a person you meet randomly on a street is a
continuous random variable.
The probability structure of a random variable, y, is described by
its probability distribution. If y is a discrete random variable, its
probability distribution is described by the probability function, often
denoted by p(y). If y is a continuous random variable, its probability
distribution is described by its probability density function, often
denoted by f(y).
The properties of the probability function p(y) and probability density
function f(y) are summarized as follows:
For a discrete random variable y

0 p( yi ) 1 yi where represents the possible values that y
can take
P(y = yi) = p( yi)
p( yi ) = 1.
All yi
For a continuous random variable y
f ( y) 0
b
P(a y b) = f ( y)dy
a
+
f ( y)dy = 1.

11.3.2 Expected Values, Variance,

and Standard Deviation
Random variable and probability distributions deal with the population,
and the populations mean, , is the most frequently used measure of
central tendency for populations. is also called the expected value of
the random variable y.
The expected value E(y) = is defined as follows:
= E( y) = yp( y) if y is discrete
All y
+
= E( y) = yf ( y)dy if y is continuous.

Population variance is often denoted by 2, and it is often simply

called the variance. The variance of a random variable y is defined as
follows:
2 = ( y )2 p( y) if y is discrete
All y
+
2 = ( y )2 f ( y)dy if y is continuous.

Variance is often denoted by Var(y). From the definition of expected

value, it is clear that:
2 = Var( y) = E[( y )2 ] .
388 Chapter Eleven
The population standard deviation is often called the standard devia-

tion, , which is simply the square root of variance 2:
= 2
11.3.3 Probability Distribution Models

In Six Sigma-related applications, there are several probability dis-
tribution models that are frequently used as the basis for data-driven
decision making. They are the normal distribution, exponential distri-
bution, binomial distribution, and Poisson distribution. We will look at
each of these in turn.
11.3.3.1 Normal Distribution The normal distribution (or Gaussian dis-

tribution) was first proposed by Gauss (17771855). Normal distribu-
tion is often used to model the probability distribution of continuous
variables that have the following properties:
There are many random factors that can affect the value of the random
variable.
Each of these random factors has a relatively small influence on the
random variable; there is no dominant factor.
Normal distribution is the most popular distribution in quality engi-
neering and Six Sigma. It is often used to model the following random
variables:
Quality characteristic of parts from suppliers

Students test scores or employee performance scores
The probability density function of the normal distribution is as

follows: 2
1 y
1
2 < y < +
f ( y) = e
2
For a normal distribution:
E(y) =
Var(y) = 2
A normal random variable y with E(y) = and Var(y) = 2 is denoted

by N(, 2).
The probability density function f(y) displays a bell-shaped curve
as illustrated in Figure 11.4. The distribution is centered at , and a
smaller will result in a tighter curve, and vice versa.
f(y)
Distribution
with smaller
Distribution
with larger
Figure 11.4 Normal probability density curve
An important special case of the normal distribution is the stan-

dard normal distribution. In the standard normal distribution, = 0
and 2 = 1. The standard normal random variable is often denoted by
z ~N(0,1). The standard normal distribution table is mainly used to
calculate probability for all kinds of normal distribution. The shape and
percentage distribution properties of the standard normal distribution
are illustrated in Figure 11.5.
Figure 11.5 shows that if y ~ N(,2), then P( y + ) = P(1
z 1) = 0.6826 = 68.27 percent, that is, 68.27 percent of observations
from a normal population will locate within one standard deviation
distance from the mean. Similarly, P( 2 y + 2 ) = P(2 z 2) =
0.9545 = 95.45 percent, that is, 95.45 percent of observations from a
normal population will locate within two standard deviations distance
from the mean. P( 3 y + 3 ) = P(3 z 3) = 0.9973 = 99.73
percent, that is, 99.73 percent of observations from a normal population
will locate within three standard deviations distance from the mean.
68.27%
95.45%
99.73%
Figure 11.5 Percentage distribution properties of normal random variable
390 Chapter Eleven
11.3.3.2 Exponential Distribution An exponential distribution is fea-

tured by the following probability density function:
y

e for 0 y <
f ( y) =

with mean and variance
E(y) = = Var(y) = 2 .
Exponential distributions are often used to model:
Lifetime of some electronic components

Time between customers entering a service facility
Time between consecutive machine failures or earthquakes
11.3.3.3 Binomial Distribution A binomial distribution is a discrete

probability distribution that characterizes a binomial random variable.
Binomial random variables can be used for situations where all of the
following apply:
There are n successive trials, and each trial will only have two distinct
outcomes, S (success) or F (failure).
The probability of success P(S) = p, P(F) = 1P(S) = 1p.
The result of each trial will not affect the results of any other trials.
If all these conditions are true, the number of successes, y, out of n

trials, will be a binomial random variable and its probability function
p(y) will be:
n! y = 0,1 ,, n.
p( y) = p y (1 p) n y
y !( n y)!
The mean and variance of the binomial random variable are:

E(y) = = np and Var(y) = 2 = np(1p).
The binomial distribution is often denoted by y ~B(n, p).

These are examples of binomial random variables:
The number of defective parts, y, in a lot of n parts in a sequential

quality inspection
The number of positive customer responses, y, in a survey involving
n customers
11.3.3.4 Poisson Distribution The Poisson probability distribution

provides a model for the probability of occurrence of rare events that
happen in a unit of time, area, volume, and so on. Actually, the Poisson
distribution is an extreme case of binomial distribution, where n is very
large, and p is very small. That is, the probability of a rare event occur-
rence, p = P(S), is very small, but the number of trials, n, is very large.
In a Poisson distribution, the parameter ( = np) is used, and the
probability function of Poisson distribution, p(y), is:
y e y = 0, 1, 2,
p( y) =
y!
The mean and variance of the Poisson distribution are:

E(y) = and Var(y) = 2 = .
11.3.4 Statistical Parameter Estimation

All probability distribution models depend on population parameters,
such as , 2 in normal distributions, and p in binomial distributions.
Without these parameters, no probability distribution model can be
used. In real-world applications, these population parameters are usu-
ally not available, but statistical estimates of these population param-
eters can be computed based on a sample of data from the population.
The commonly-used statistical estimate for in a normal distribution
is the sample mean y, where
1 n
n
y= yi .
i=1
For a sample of n observations, y1 , y2 ,..., yn, from y ~ N(, 2 ), the

commonly-used statistical estimate for 2 in normal distribution is the
sample variance s2, where
1 n
n 1
s2 = ( yi y)2 .
i=1
For a sample of n observations y1 , y2 ,..., yn from y ~ N(, 2), and

1 n
y = n yi .
i =1
The commonly-used statistical estimate for p in a binomial distribu-

tion B(n, p) is the sample ratio, p, where
y
p =
n
where y is the actual number of successes (S) in n trials.
392 Chapter Eleven
However, statistical estimates are only approximations of the true pop-

ulation parameters. When sample size is small, there will be substantial
discrepancies between population parameters and statistical estimates.
As the sample size gets larger, the discrepancies will get smaller.
11.4 Quality Measures and Six

Sigma Metrics
For any product or business process, there are always performance met-
rics to measure and improve. For example, in a loan approval process,
the cycle time (the time from loan application to loan decision) is a
performance metric. In Example 11.1, the oxide film thickness is a per-
formance metric and the ideal thickness is 560 Angstroms. Most process
performance metrics are random variables, like the loan cycle time and
the oxides film thickness.
There are many quality measures that are developed to measure
the performance of a process when the process involves randomness.
Quality measures compare the degree of randomness in the processs
performance and compare the degree of randomness with the processs
performance specification. The most commonly used process perfor-
mance quality measure is the process capability index. In Six Sigma
practice, many other process performance-related metrics have also
been developed, such as Sigma quality level, DPMO (defect per million
opportunities), and so on.
In this section, we will look first at the process capability index, and
then at other Six Sigma metrics.
11.4.1 Process Performance

Process performance is a measure of how well a process performs. It
is measured by comparing the actual process performance level to the
ideal process performance level. For the oxide film build process, its per-
formance may be measured by its thickness, and its ideal performance
level would be 560 Angstroms.
For most processes, performance level is not constant. We call this
variation the process variability. If the process performance can be
measured by a real number, then the process variability can usually
be modeled by normal distribution, and the degree of variation can be
measured by the standard deviation of that normal distribution.
If the process performance level is not a constant but a random vari-
able, we can use process mean and process standard deviation as key
performance measures. Mean performance can be calculated by averag-
ing a large number of performance measurements.
If processes follow the normal probability distribution, a high percent-

age of the process performance measurements will fall between 3 of
the process mean, where is the standard deviation. That is, approxi-
mately 0.27 percent of the measurements would naturally fall outside
the 3 limits, and the balance of them (approximately 99.73 percent)
would be within the 3 limits.
Since the process limits extend from 3 to +3 , the total spread
amounts to about 6 total variation. This total spread is often used to
measure the range of process variability, also called the process spread.
For any process performance measure, there are usually some per-
formance specification limits. For example, if the oxide film thickness
in a wafer is too high or too low, it will not function well. For example, if
deviation cannot be more than 100 Angstroms from the target value of
560 Angstroms, its specification limits would be 560 100 Angstroms.
We would say that its specification spread is (460, 660), where 460
Angstroms is the lower specification limit, or LSL, and 660 Angstroms
is the upper specification limit, or USL.
11.4.2 Process Capability Indices

Capability indices are simplified measures that quickly describe the
relationship between the variability of a process and the spread of the
specification limits.
11.4.2.1 The Capability Index Cp The equation for the simplest capabil-
ity index, Cp, is the ratio of the specification spread to the process spread,
the latter represented by six standard deviations or 6. Cp assumes that
the normal distribution is the correct model for the process.
USL LCL
Cp =
6
Cp can be translated directly to the percentage or proportion of non-

conforming product outside specifications, if the mean of the process
performance is at the center of the specification limits.
When Cp = 1.00, approximately 0.27 percent of the parts are outside
the specification limits (assuming that the process is centered on the
midpoint between the specification limits) because the specification
limits closely match the process USL and LSL. We say this is about 2700
parts per million (ppm) nonconforming.
When Cp = 1.33, approximately .0064 percent of the parts are outside
the specification limits (assuming the process is centered on the mid-
point between the specification limits). We say this is about 64 parts
per million (ppm) nonconforming. In this case, we would be looking at
normal curve areas beyond 1.33 3 = 4 from the center.
394 Chapter Eleven
When Cp = 1.67, approximately .000057 percent of the parts are out-

side the specification limits (assuming the process is centered on the
midpoint between the specification limits). We say this is about .6 parts
per million (ppm) nonconforming. In this case, we would be looking at
normal curve areas beyond 1.67 3 = 5 from the center of the normal
distribution.
11.4.2.1 The Capability Index Cpk The major weakness in Cp is that,

for many processes, the mean performance of the process is not equal
to the center of the specification limit. Also, many process means will
drift from time to time. When that happens, the probability calculation
for nonconformance will be wrong when we still use Cp. Therefore, one
must consider where the process mean is located relative to the speci-
fication limits. The index Cpk is created to do exactly this.
USL LSL
C pk = Min , = Min CPU , CPL .{ }
3 3
For example, suppose the process standard deviation is = .8 with

USL = 24, LSL = 18, and the process mean = 22.
LSL = 18 USL = 24
Process center
Center = 22
24 22 22 18
C pk = Min , = Min(0.83, 1.67) = 0.83 .
38 38
It is also clear that
CPU = 0.83, CPL = 1.67.
If the process mean were exactly centered between the specification

limits:
Cp = Cpk = 1.25.
Cpk is a more accurate process performance metric when process mean

is not at the nominal design point.
Example 11.8 Process Capability Calculation For the film thickness

data given in Example 11.1, LSL = 460, USL = 660, and we do not know
the exact value of and . However, we can calculate that y = 572.02
and s = 24.53. Because the sample size of this data set is fairly large
(n = 47), we can substitute and by using y and s. This gives the
following result:
C pk = Min { 660 572.02 572.02 460

3 24.53
,
3 24.53 }
= Min (1.19, 1.51) = 1.19
MINITAB can be used to conduct a comprehensive process capabil-

ity analysis. The following MINITAB output is the process capability
analysis for the film thickness data.
Process Capability of Film Thickness
LSL USL
Process Data Within
LSL 460.00000 Overall
Target
USL 660.00000 Potential (Within) Capability
Sample Mean 572.02128 Cp 1.35
Sample?N 47 CPL 1.51
StDev(Within) 24.66804 CPU 1.19
StDev(Overall) 24.66804 Cpk 1.19
CCpk 1.35
Overall Capability
Pp 1.35
PPL 1.51
PPU 1.19
Ppk 1.19
Cpm
480 510 540 570 600 630 660

Observed Performance Exp. Within Performance Exp. Overall Performance
PPM?<?LSL 0.00 PPM?<?LSL 2.80 PPM?<?LSL 2.80
PPM?>?USL 0.00 PPM?>?USL 180.89 PPM?>?USL 180.89
PPM?Total 0.00 PPM?Total 183.68 PPM?Total 183.68
11.4.3 Sigma Quality Level

(Without Mean Shift)
In 1988, the Motorola Corporation was the winner of the Malcolm
Baldridge National Quality Award. Motorola bases much of its quality
effort on its Six Sigma Program. The goal of this program was to reduce
the variation in every process to such an extent that a spread of 12 (6
on each side of the mean) fits within the process specification limits.
Figure 11.6 gives a graphical illustration of this Six Sigma quality.
If the actual variation measured by standard deviation is , 6 quality
means that the total spread of the specification is six times the standard
deviation on each side of the mean.
USL LSL 12
For a Six Sigma quality level, C p = = = 2.
6 6
396 Chapter Eleven
LSL USL
6 3 +3 +6
Figure 11.6 Six Sigma quality
By using normal probability distribution, it can be computed that
P(y) will be in specification if P(y) =

P ( LSL y USL) = P ( 6 z 6) = 0.999999998 = 99.9999998 percent.
Clearly, P(y) will be out of specification if P(y) = 1 0.9999999998 =

0.000000002, or 0.002 defects per million parts (ppm).
Similarly, if the spread of the specification is five times on each side
of the mean, it is called Five Sigma quality.
The following table summarizes the relationship between Cp, Sigma
quality level (without mean shift), percentage in specification, and
defective ppm.
Sigma Quality Level

(without mean shift) Cp Percentage in Specification Defective PPM
1 0.33 68.27 317300
2 0.67 95.45 45500
3 1.0 99.73 2700
4 1.33 99.9937 63
5 1.67 99.999943 0.57
6 2.00 99.9999998 0.002
11.4.4 Sigma Quality Level

(With Mean Shift)
In most actual processes, the process mean, y, is not a constant. The
process mean will shift from time to time. For example, in manufactur-
ing processes, with a change of raw material or operator, the process
may suddenly change its mean level. In service processes, with a change
of server or shift, the process mean may also change.
Motorola allocates 1.5 on either side of the process mean to take into
account this shifting of the mean. For a Six Sigma quality level, with
the maximum possible mean shift of 1.5, the minimum distance from
the process mean to one of the specification limits could be as small as
4.5. The following illustration shows the relationship between mean
shift and Six Sigma quality level.
4.5 1.5 1.5 4.5

Sigma Sigma Sigma Sigma
LSL USL
Process mean
Thus, even if the process mean strays as much as 1.5 from the pro-
cess center, a full 4.5 remains. This ensures a worst-case scenario of
3.4 parts per million (ppm) nonconforming on each side of the distribu-
tion. With the inclusion of a 1.5 mean shift, for the same sigma quality
level, the defective ppm will be much larger than when not considering
a mean shift.
The following table summarizes the relationship between Cp, Sigma
quality level (with mean shift), percentage in specification, and defec-
tive ppm.
Sigma Quality Level Percentage in

(with mean shift) Cp Specification Defective PPM
1 0.33 30.23 697700
2 0.67 69.13 308700
3 1.0 93.32 66810
4 1.33 99.3790 6210
5 1.67 99.97670 233
6 2.00 99.999660 3.4
References
Aaker, D. A. (1996) Building Strong Brands, New York: The Free Press.
Abrams, B. (2000) Observational Research Handbook: Understanding How Consumers
Live with Your Product, First Edition, New York: McGraw-Hill.
Agar, M. (1996) The Professional Stranger: An Informal Introduction to Ethnography,
Second Edition, Burlington, MA: Academic Press.
Allen, J., Robinson, C., and Stewart, D. eds., (2001) Lean Manufacturing: A Plant Floor
Guide. Dearborn, MI: Society of Manufacturing Engineers.
Altshuler, G. S. (1988) Creativity as Exact Science, New York: Gordon & Breach.
Altshuler, G. S. (1990) On the theory of solving inventive problems, Design Methods and
Theories 24 (2): 121622.
Anderson, E., Rornell, C., and Lehmann, D. (1994) Customer Satisfaction, Market Share
and Profitability: Findings from Sweden, Journal of Marketing, July, pp. 5366.
Arnold, D. (1992) The Handbook of Brand Management, Reading, MA: Addison-Wesley.
Ballard, B. (1987) Academic Adjustment: The Other Side Of The Export Dollar, Higher
Education Research and Development, Volume 6, Issue 2, pp. 109119.
Batra, R., Lehmann, D., and Singh, D. (1993) The Brand Personality Component of
Brand Goodwill: Some Antecedents and Consequences, Brand Equity and Advertising:
Advertisings Role in Building Strong Brands, Aaker D., and Biel, A. (eds.), Hillsdale, NJ:
Lawrence Erbaum Associates Publishers, pp. 8396.
Berry M. and Linoff G.S. (2000), Master Data Mining, New York: Wiley.
Biel A.L. (1993) Converting Brand Image into Equity, Brand Equity and Advertising:
Advertisings Role in Building Strong Brands, Aaker D., and Biel, A. (eds.), Hillsdale, NJ:
Lawrence Erbaum Associates Publishers, pp. 6782.
Clark, K. and Fujimori, T. (1991). Product Development Performance: Strategy, Organization
and Management in the World Auto Industry, Boston, MA: Harvard Business School
Press.
Clausing, D. P. (1994) Total Quality Development: A Step by Step Guide to World-Class
Concurrent Engineering, New York: ASME Press.
Cohen, L. (1988) Quality function deployment and application perspective from Digital
Equipment Corporation, National Productivity Review 7 (3): 197208.
Cohen, L. (1995) Quality Function Deployment: How to Make QFD Work for You. Reading,
MA: Addison-Wesley.
Cooper, R.G. (1990) Stage-gate systems: A new tool for managing new products, Business
Horizons, Vol 33, issue 3, pp. 4454.
Davenport, T.H., De Long, D.W., and Beers, M.C. (1998) Successful Knowledge
Management Projects, Management of Technology and Innovation, Vol. 39, No. 2,
pp. 4357.
Davis, S. M. (2000) Brand Asset Management. San Francisco, Jossey-Bass CA, Inc.
de Brentani, U. (1993) The new product process in financial services: Strategy for success,
International Journal of Bank Marketing 11 (3): 1522.
399
400 References
Deming, E. (1982) Out of the Crisis, Cambridge, MA: Massachusetts Institute of Technology,
Center for Advanced Engineering Study.
Dixon, J.R. (1966) Design Engineering: Inventiveness Analysis and Decision Making,
McGraw-Hill, New York.
Edelstein, H.A. (1999) Introduction to Data Mining and Knowledge Discovery, Third
Edition, Potomac, MD: Two Crows Corporation.
Farquhar, P.H. (1989) Managing Brand Equity, Marketing Research, Vol. 3, pp. 2433.
Fox, J.E., Mockovak, W., Fisher, S.K, and Rho, C. (2003) Usability Issues Associated with
Converting Establishment Survey to Web-Based Data Collection, Proceedings of the
Federal Committee of Statistical Methodology, Washington, D.C.
Fredriksson, B. (1994) Holistic systems engineering in product development, The Saab-
Scania Griffin, November.
Gale, B. (1994) Managing Customer Value, New York: The Free Press.
George, M. (2003) Lean Six Sigma for Service, New York: McGraw-Hill.
Goetsch, D. L., and Davis, S. B. (2000) Quality Management: Introduction to Total Quality
Management for Production, Processing, and Services, Third Edition, Upper Saddle River,
NJ: Prentice Hall.
Gold, R.L. (1958) Roles in Sociological Field Observations, Social Forces, Vol. 36, No. 3
pp. 217223.
Hauser, J. R., and Clausing, D. (1988) The House of Quality, Harvard Business Review,
66 (3): pp. 6373.
Hubka, V. (1980) Principles of Engineering Design, (translated by W.E. Eder), London:
Butterworth Scientific.
Huthwaite, B. (2004) The Lean Design Solution, Mackinac Island, MI: Institute for Lean
Design.
Jacobson, R. and Aaker, D. (1987) The strategic role of product quality, Journal of
Marketing 51: pp. 3144.
Kaufman, J. J. (1989) Value Engineering for the Practitioner, Second Edition, Raleigh, NC:
North Carolina State Univ. Press.
Kawakita, J. (1977) A Scientific Exploration of Intellect (chi no tankengaku), Tokyo:
Kodansha.
Kawakita, J. (1991) The Original kj Method, Tokyo: Kawakita Research Institute.
Keller, K. L. (1993) Conceptualizing, measuring, and managing customer-based brand
equity, Journal of Marketing 57: pp. 122.
Kennedy, M. (2003) Product Development for the Lean Enterprise, Richmond, VA: The
Oaklea Press.
Kim, W. and Mauborgne, R. (2006) Blue Ocean Strategy: How to Create Uncontested Market
Space and Make Competition Irrelevant. Boston, MA: Harvard Business School Press.
Kochan, N. (Ed.) (1997) The Worlds Greatest Brands, New York: University Press.
Levine, M. (2003) A Branded World, Hoboken, NJ: John Wiley and Sons, Inc.
Liker, J. K. (2004) The Toyota Way, New York: McGraw-Hill.
Loftus, E. and Wells, G. (1984) Eyewitness Testimony: Psychological Perspectives,
New York: Cambridge Univ. Press.
MacElroy, B. (2000) Variables influencing dropout rates in web-based surveys, Quirks
Marketing Research Review, July/August. Retrieved October 9, 2002, from http://www
.modalis.com/english/news/ newsitem000808.html.
Mann, D. (2002) Hands-on Systematic Innovation, Ieper, Belgium: CREAX Press.
References 401
Mann, D. (2004) Hands-on Systematic Innovation for Business and Management, Clevedon,
UK: IFR Press.
Mann, D. and Domb, E. (1999) 40 Inventive (business) principles with examples, TRIZ
Journal, September.
McCord, K. R., Eppinger, S. D. (1993) Managing the integration problem in concurrent
engineering, Working paper #3594-93-MSA, Cambridge MA: Sloan School of Management,
Massachusetts Institute of Technology.
Miles, L.D. (1961) Techniques for Value Analysis and Value Engineering, New York,
McGraw-Hill.
Miles, B. L. (1989) Design for assembly: A key element within design for manufacture,
Proc. IMechE, Part D, Journal of Automobile Engineering 203: pp. 2938.
Monden, Y. (1993) The Toyota Management System: Linking the Seven Key Functional
Areas, Portland, OR: Productivity Press.
Morgan, J. and Liker, J. (2006) The Toyota Product Development System. New York:
Productivity Press.
Murphy, J. (1990) Brand Valuation, Establish a True and Fair Value, London: Hutchinson
Business Books.
Nielsen, J. (2000) Designing Web Usability: The Practice of Simplicity, Indianapolis, IN,
New Riders Publishing.
Nonaka, I. and Takeuchi, H. (1995) The Knowledge-Creating Company: How Japanese
Companies Create the Dynamics of Innovation, New York: Oxford Univ. Press.
Ohno, T. (1990) Toyota Production System: Beyond Large-Scale Production, Portland, OR:
Productivity Press.
Park, R. J. (1992) Value Engineering, Birmingham, MI: R. J. Park & Associates, Inc.
Park, R. J. (1999) Value Engineering, a Plan for Invention, Boca Raton, FL: Saint Lucie
Press.
Penney, R. K. (1970) Principles of Engineering Design, Postgraduate, No. 46, pp. 344349.
Phal, G. and Beitz (1988) Engineering Design: A Systematic Approach, New York:
Springer-Verlag.
Phaal, R., et al. (2001) Technology Roadmapping: Linking technology resources to business
objectives, Cambridge, UK: Univ. of Cambridge, Centre for Technology Management.
Pimmler, T. U. and Eppinger, S. D. (1994) Integration Analysis of Product Decomposition,
Design Theory and Methodology, Vol 68, pp. 343351.
Polanyi, M. (1967) The Tacit Dimension, New York: Anchor Books.
Pugh, S. (1991) Total Design: Integrated Methods for Successful Product Engineering,
Reading, MA: Addison-Wesley.
Pugh, S. (1996) Creating Innovative Products Using Total Design, Reading, MA: Addison-
Wesley.
Ramaswamy, R. (1996) Design and Management of Service Processes, Reading, MA:
Addison-Wesley.
Rantanen, K. (1988) Altshulers methodology in solving inventive problems, Proceedings
of ICED 88, Budapest, pp. 2325 Aug.
Rea, L. M. and Parker, R. A. (1992) Designing and Conducting Survey Research: A
Comprehensive Guide, San Francisco, CA: Jossey-Bass Inc.
Reeve, D. (1975) Value engineering analysis of the Oakland County Youth Assistance
Program, Masters thesis, Oakland Univ.
Reinertsen, D. (1997) Managing the Design Factory, New York: The Freedom Press.
402 References
Reis, A. (1981) Positioning: The Battle for Your Mind, New York: McGraw-Hill.
Rother, M. and Shook, J. (2003) Learning to See: Value-Stream Mapping to Create Value
and Eliminate Muda, Brookline, MA: The Lean Enterprise Institute, Inc.
Rust, L. (1993) Observations: Parents and Children Shopping Together Journal of
Advertising Research, July/August, pp. 6570.
Sherden, W. A. (1994) Market Ownership, New York: American Management Association.
Shigeru, M. (1988) Management for Quality Improvement: The Seven New QC Tools,
Cambridge, MA: Productivity Press.
Shingo, S. (1989) The Toyota Production System from an Industrial Engineering Viewpoint,
Portland, OR: Productivity Press.
Smith, P. G. and Reinertsen, D. G. (1998) Developing Products in Half the Time: New
Rules, New Tools, New York: Van Nostrand Reinhold (International Thomson Publishing
Company).
Spradley (1980) Participant Observation. New York: Harcourt, Brace, Jovanovich.
Squires, S. and Byrne, B. (eds.) (2002) Creating Breakthrough Ideas: The Collaboration of
Anthropologists and Designers in the Product Development Industry, New York: Bergin
& Garvey.
Suh, N. P. (1984) Development of the science base for the manufacturing field through the
axiomatic approach, Robotics & Computer Integrated Manufacturing 1 (34).
Suh, N. P. (1990) The Principles of Design, First Edition, New York: Oxford Univ. Press.
Suh, N. P. (2001) Axiomatic Design: Advances and Applications, First Edition, New York:
Oxford Univ. Press.
Tauber, F. M. (1988) Brand Leverage: Strategy for Growth in a Cost Controlled World,
Journal of Advertising Research, Vol 28, pp. 2630.
Tsourikov, V. M. (1993) Inventive machine: Second generation, Artificial Intelligence &
Society 7: pp. 6277.
Ullman, D. G. (1992) The Mechanical Design Process, New York: McGraw-Hill.
Ulrich K. T. and Eppinger S. D. (2000) Product Design and Development, New York:
McGraw-Hill.
Ulwick, A. (2005) What Customers Want: Using Outcome-Driven Innovation to Create
Breakthrough Products and Services, New York: McGraw-Hill.
Watkins, T. (1986) Economics of the Brand, London: McGraw-Hill.
Wellner, A. S. (2002) The Test Drive-ethnographic research and marketing research,
American Demographics,Vol.Oct 1, 2002 http://findarticles.com/p/articles/mi_m4021/is_
2002_Oct_1/ai_92087420.
Wicker, A. W. (1969) Attitudes versus actions: The relationship of verbal and overt behav-
ioral responses to attitude objects, Journal of Social Issues 25 (4): pp. 4178.
Womack, J. P., Jones, D. T., and Roos, D. (1990) The Machine that Changed the World, New
York: Rawson Associates.
Womack, J. P. and Jones, D. T. (2003) Lean Thinking, New York: The Free Press.
Yang, K. and El-Haik, B. (2003) Design for Six Sigma: A Roadmap for Product Development,
New York: McGraw-Hill.
Index
A audio documentation, 178

A3 reports, 8183 audio recordings, 186
Aakers brand identity model, automobile dealership construction,
264272 339340
across-the-board budget cuts, 343 axiomatic design, 3543
actual costs, 303
administrative costs, 304 B
advertisement styles, 270 basic functions, 299, 309
affinity diagrams batch flow process, 69
data analysis, 216217 beforehand cushioning principle, 366
data collection, 215 beginners mind, 177
defining, 233 beliefs, 206
header cards for groups, 216 Beliefs and Values layer, 273
overview, 214215 benchmarking, 225
writing reports, 216221 Benefits layer, 273
allowances, 304 best cost, 299, 327
alternative industries, 117118 Bethway, Charles, 314
Altshuller, Genrich S., 20, 346 binomial distribution, 390
American Society for Quality binomial random variables, 390
(ASQ), 94 blessing in disguise principle, 367
analogs, 364 blue ocean strategy
analysis alternative industries, 117118
of competitor, 286287 appeals to customers, 122123
of customer, 284285 chain of buyers, 119121
of self, 285286 complementary product/service
analytical prototypes, 3132 offerings, 121122
another dimension principle, 366 defining, 89
anti-weight principle, 365 long term needs, 123
arithmetic mean, 383 overview, 111112
ASQ (American Society for strategic groups within industries,
Quality), 94 118119
assist client function, 312, 315 U.S. wine industry, 112116
assisting functions, 350 borders, 291
asymmetry principle, 365 box plots, 382
attitudes, 206 brainstorming, 299, 328329
attribute data, 213, 221 brand associations, 277
403
404 Index
brand attributes, 268 C

brand awareness, 280, 293 CAD (computer-aided design), 30
brand contracts, 289290 campaign guidelines, Marlboro
brand development cigarettes, 295
buying behavior, 262264 capability indexes, 393395
evaluation of, 293295 celebrity endorsements, 270
implementation of, 293 central tendencies, 383384
process CFD (computational fluid
controllable factors, 278279 dynamics), 30
paradigms, 281283 chain of buyers, 119121
purpose of, 287 cheap short-living principle, 368
strategy checklist survey questions, 141142
analysis, 283 classifications of problems, 362363
benefits/points of differentiation, close-ended questions, 141144
267, 289 cluster sampling, 152
brand positions of bookstores, color changes principle, 368
290291 commercial costs, 304
channel strategy, 292293 communication
communications, 291292 marketing strategies, 291292
competitor analysis, 286287 product development, 280
contract with market, 289 with survey participants,
customer analysis, 284285 162163
positioning, 287290 company image, 270
pricing, 293 comparative techniques, 308
product-brand coordination, competitive activity, 281
290291 competitive assessments, 239
public relations, 293 competitive brand positioning, 287
self-analysis, 285286 competitive customer value analysis,
target market segment, 107109
288289 competitors, 286287
vision development, 283284 complementary products/services,
brand dominance, 12 121122
brand equity, 275277 complexity, excessive, 363
brand extension, 267 composite materials principle, 369
brand identity, 278279, 287 computational fluid dynamics
brand leadership model, 281 (CFD), 30
brand names, 260261 computer simulation models, 2931
brand persona, 274 computer-aided design (CAD), 30
brand positioning, principles, 290 computers, usability research,
brand recall, 276 201202
brand symbols, 270 concept development
brand-as-organization attributes, 268 design phases, 19
brand-customer relationship, 287 lead users, 20
brands, defining, 259 overview, 1825
buffer inventory, 70 concept evaluation methods, 363
burden, 303 continuity of useful action
Burnett, Leo, 295 principle, 367
business strategy, 267 continuous factors, 27
Index 405
continuous random variables, importance of, 248

386387 product quality measures, 214
contracts, brand, 289 trees, 220
contradiction matrix, 370377 WHATs, 235
contradictions CTQs. See critical-to-quality
analyzing, 362 characteristics (CTQs)
defining, 21 CTSs. See critical-to-satisfaction
design parameters, 356 requirements (CTSs)
technical, 355357 cultural issues
controllable factors, 278 cultural meanings, 204207
conversations, 174 cultural tools, 204, 207211
copying principle, 368 customer behaviors, 204205
core operation, 11 overview, 178
correcting functions, 350 research, 208211
correlation matrix, 238 rules, 207
cost uncontrollable factor, 281
components, 11 Customer Competitive
defined, 303 Assessment, 239
figures, 301 Customer Desirability Index,
Internet survey, 166167 236, 237
product, 914 customer domains, 3637
and value, 94 customer value
cost information, sources of, 304305 analysis, 97109
cost-function relationship, 323327 competitive analysis, 107109
cost-function worksheets, 300302, market-perceived price profile,
341 102104
cost-visibility market-perceived quality profile,
analysis, 303, 324 98102
calculations, 305 overview, 9798
terminologies, 302304 value map, 104107
worksheets, 305 blue ocean strategy
countries of origin, 267, 270 alternative industries, 117118
create design function, 341 appeals to customers, 122123
creative phase, 300, 332 chain of buyers, 119121
creativity, 312314 complementary product/service
credibility principle, 290 offerings, 121122
Critical-To characteristics, 229 long term needs, 123
critical-to-quality characteristics overview, 111112
(CTQs) strategic groups within
components, 223 industries, 118119
for customer satisfaction, 223227 U.S. wine industry, 112116
for product development, 18 defining, 25
product quality measures, 214 deployment, 109111
quantitative measures, 226 dynamics of, 9597
trees, 224 elements of, 9197
critical-to-satisfaction requirements equation
(CTSs) benefits, 34
HOWs, 235, 238, 246248 liabilities, 4
406 Index
customer value (Cont.) telephone, 135

importance of, 25 types of questions, 285
innovation, 56 uncontrollable factor, 281
overview, 8991
quality function deployment D
(QFD), 110 data analysis, 140, 379380
red sea analogy, 111 data collection methods, 185189
survey forms, 101 data mining, 136
versatility, 9597 data processing
voice of customer (VOC) analyzing data, 214
collecting data, 129132 critical-to-quality characteristics,
data accuracy, 126 223227
data collectors, 128129 overview, 213
identifying target customers, quantitative data analysis,
127128 221223
mining, 67 types of data, 213214
organizing data, 132 data recording methods, 174175
overview, 126127 Daviss brand image model, 272275
plan for, 127 decline stage, 359360
customer-centric innovation, 12, 59 decoupled designs, 3839
customers. See also customer value define problems function, 312
analysis of, 284285 defined research, 176177
attributes of descriptive statistics, 380386
defining, 124 design
Kano model, 239 axiomatic, 3543
WHATs, 234235, 254255 experiment, 2529
buying decisions, 262 as information production factory,
competitors, 127 4345
cultures, 203211 as mapping process, 3537
current, 127 matrices, 3839
external, 48 practices, 77
input from, 224 principles, 35
internal, 48 product, 2533
lead, 128 Suh, 35
needs of, 219221, 285 Design for Six Sigma (DFSS), 76
potential, 128 Design of Experiments (DOE), 2529
surveys of design parameters (DPs), 3637
administering, 147148, 162166 DFSS (Design for Six Sigma), 76
goals, 137138 diagrams, functional analysis,
implementation of, 140 350353
in-person interviews, 134135 diaries, informant, 174
Internet, 157167 differentiation, 289
mail-out, 134 discarding/recovering principle, 369
overview, 133 discovery research
questionnaire development, defining, 175176
141147 projects, 179
sampling, 148157 discovery stages, 16
stages of, 136140 discrete data, 221
Index 407
discrete factors, 27 evaluation phase, 330335

discrete random variables, 386387 evaluation research
distribution channels projects, 179
in brand development, 280 purpose of, 177
strategies, 292293 evolution
strategy, 292 of technical systems, 357360
diversions, 205 of technological development, 348
DOE (Design of Experiments), 2529 excessive action principle, 366
domains, 3637 excessive complexity, 363
dot plots, 380 expected values, 387388
downstream stages, 14 experimental designs, 2728
DPs (design parameters), 3637 explicit knowledge, 51, 58
dynamics of value, 9597 exponential distribution, 390
dynamics principle, 366 external customers, 48
E F
EDA (electrical design Fallon, Carlos, 91
automation), 30 FAST (Functional Analysis System
electrical design automation Technique), 301, 314316,
(EDA), 30 319323, 340
eliminate deviancy function, 312, FEA (finite element analysis), 30
315, 316 Features and Attributes layer, 273
e-mail, 160163 feedback principle, 367
emotions feelings, 207
appeal to customer, 122123 field notes, 174175, 186, 353
benefit to customers, 271 fieldwork guides, 190191
cultural meanings, 207 finite element analysis (FEA), 30
engineering department organization first-hand information, 177
analysis, 341343 fish cultivation, 354355
equipment, usability research, 199 fit principle, 290
equipotentiality principle, 366 fixed burden, 303
Erlicher, Harry, 298 fixed costs, 303
ethnographic interviews, 194 flexible shells/thin films
ethnographic methods principle, 368
advantages of, 177178 focus groups, 139
data recording, 174175 focused prototypes, 3132
frequently used methods, 173174 forcing technique, 308
observations in product usage formal interviews, 174
processes, 198203 freight cost, 304
overview, 171173 Fujimori, 84
in product development, 175177 function analysis, 301, 313, 348
project execution, 191198 function determination, 301, 309
project planning function lists, 300
data collection methods, 185189 functional analysis diagrams,
objectives, 178180 350352
overview, 178 Functional Analysis System
recruiting informants, 180185 Technique (FAST), 301, 314316,
team/ground rules, 189191 319323, 340
408 Index
functional appeals, 122123 defining, 23, 233

functional benefits, 271 eye opener opportunities, 240
functional domains, 3637 of QFD, 230
functional modeling/analysis, significant points, 240
361362 HOWs, (critical-to-satisfaction
functional requirements, 36, 243 requirements), 245248
functional resources, 354 human factors, 79, 199
functional structures, 363 human relations, 301
functionality, 348 hydraulics, 368
functions
defining, 308, 314 I
development of, 363 idea generation, 1417
evaluating, 299 ideal final results, 348
identifying, 298299 ideal product development process,
fuzzy front end stages 6064
defining, 14 ideality, 347, 354355, 362
elements of, 16 identifying functions, 298299
idea generation/handling, 17 imagination, 299
imperfect functional structures, 363
G implementation of surveys, 140
Gale, Bradley, 89, 9697 implementation phase, 337339
games, 205 importance ratings, 237, 255
gates, 35 incremental costs, 303
Gaussian distribution, 388 independence axioms, 3742
General Electric Company, 298 independent functions, 318
generic products, 10 industries
George, Michael, 68 alternative, 117118
goal-oriented behaviors, 205 strategic groups within, 118119
goals, survey, 137138 wine, 112116
Go-Gurt, 175 inert atmosphere principle, 369
ground rules, 189191 infancy stage, 358359
grouped survey questions, 146147 influencers, 119
growth stage, technical system informal interviews, 174
evolution, 359 informant diaries, 174
informants
H defining, 172173
hard attributes, 277 recruiting, 180185
harmful functions, 350 information bases, survey, 138139
header cards, 216 information development, 301
higher-order functions, 314 information mining, 5052
highest-potential projects, 332 customer attributes, 124
highest-value functions, 332 explicit knowledge, 51, 58
histograms, 380381 tacit knowledge, 51, 58
homogeneity principle, 369 information phase, 300301
House of Quality (HOQ) information resources, 354
benchmarking, 241 information system design, 241251
blank/weak columns/rows, 240 information transformation process
components, 234 best-practice database, 55
conflicts, 240 flow down, 54
Index 409
information transformation/flow inventories, 70

adaption of technology, 86 inventory survey questions, 141142
human factors, 79
leveled product development J
process flow, 85 jargon, cultural meanings, 206
optimizing, 7886 job plan, value-engineering, 300
queueing theory, 7981 job shop process, 68
reduction of product development juice beverage shoppers,
system complexity, 8486 ethnographic research, 197198
stages, 8384
standardization, 8586 K
tools for standardization and Kano model of quality, 239240
organizational learning, 86 Kawakita, Jiro, 214
visible knowledge, 8183 KJ method. See affinity diagrams
visual communications, 86 knowledge
information-collecting phase, 301 explicit, 51, 58
initial screening process, 330 tacit, 51, 58
innovation, 56, 20, 59 visible, 8182
in-person interviews, 134135 knowledge conversion, 5859
in-person surveys, 148 knowledge resources, 354
intangible products, 9 Kobe Shipyards, 231, 236
intermediary principle, 367 Koltler, Philip, 259
internal customers, 48
L
Internet surveys, 157167
labor cost, 305
administering, 162166
labor, defined, 303
comparison with paper-based
language, 206
surveys, 166167
language and cultural meanings, 206
drawing people to, 158161
large-scale surveys, 167
instructions, 165166
law of increasing ideality, 355
overview, 157
lead time, Internet survey, 167
interpretation and cultural
lead users, 20
meanings, 206
lean product development, 6487
interval level survey
creating lean product, 8687
measurement, 143
front-loading product development
interval scale, 143, 156157
process, 7678
interviewers, survey, 140
information transformation and
interviews, 174, 185186, 191196
flow, 7886
inventive principles, 363378
adaption of technology, 86
contradiction table of, 372377
human factors, 79
improving wrench design using,
information flow across stages, 84
369371, 378
leveled product development
overview, 363364
process flow, 85
partial contradiction matrix, 371
queueing theory, 7981
principles, 364369
reduction of product development
inventive problem solving, theory
system complexity, 8486
of, 2022
stage overlapping, 8384
inventiveness, levels of (TRIZ),
standardization, 8586
346347
410 Index
lean product development (Cont.) marketing brands, 293

tools for standardization and marketing expenditures, 293
organizational learning, 86 marketing strategy, communications
visible knowledge, 8183 and, 291292
visual communications, 86 market-perceived price profile,
mining VOC to capture value, 102104
7475 market-perceived quality profile,
one-piece flow, 6870 98102
operation principles, 6465 Marlboro cigarette brand
overview, 64 development, 294295
principles for product development, Maslow, Abraham, 9697
7174 materials, defined, 303
pull-based production, 7071 matrix
technical competence, 7576 contradiction, 370377
value-stream mapping, 6668 planning, 237238
waste elimination in, 6566 relationship, 235236
level of measurements, close-ended maturity stage, technical system
questions, 142144 evolution, 359
leveled product development process McDonalds brand identity, 271272
flow, 85 mean, 383, 395397
levels, 27 measurement scales, 222
lifestyles of informants, 181 mechanical vibration principle, 367
Likert scale survey measurement, mechanics substitution principle, 368
143144 mechanism-based simulation
line flow process, 70 models, 30
lipstick study, 179180 median, 383384
local quality principle, 365 merging principle, 365
longitudinal study, 186 Miles, L. D., 298, 308, 313
lower-order function, 319 Mitsubishi Heavy Industry, 231
lower/upper triangle matrix, 38 modify behavior, 315
low-order functions, 314 modular design, 24, 77
loyalty, brand, 276 modularity, 2425
Monte Carlo simulation model, 3031
M Morris, Philip, 295
mail services, brand personas for, muda, 66
274275 multiple-choice survey questions, 141
mail-out surveys, 134, 147148 multipurpose questions, 145
main basic functions, 349 mystery customers, 226
manufacturing processes
batch flow process, 69 N
as compared to product negative ideas, 339
development process, 72 nested doll principle, 365
job shop process, 68 nominal level survey measurement,
line flow process, 70 142143
mapping process, design as, 3537 nominal scale, 142143
mapping, value-stream, 6668 nonparticipant observation, 173174
maps, customer value, 104107 nonprobability sampling,
margin of error, 154155 150, 152153
Index 411
normal distribution, 388389 pencils

normal probability density curve, cost-function worksheet for,
387388 324327
Novo Nordisk, 120121 function determination, 310
numerical descriptive statistics, paired comparisons for, 334335
382386 perceived quality, 95
percentage distribution properties of
O normal random variables, 389
Oakland County Youth Assistance percentile points, 385
Program, 310312 performance, 94, 266, 392393
objective information, 177 periodic action principle, 367
OBrien, Chris, 91 phase analysis, 255258
one-piece flow, 6870 phase transitions principle, 369
open-ended questions, 141, 144 photographs, 175, 186187
operation principles, 6465 physical contradictions, 357, 362
operations items, 255 physical domains, 3637
opinions, cultural meanings, 206 physical prototypes, 3133
opportunity identification, plan activities, 316
1417 planning matrix, 237238
ordinal level survey planning phase, 332
measurements, 143 planning phase, value-engineering
ordinal scale, 143 job plan, 335336
organizational attributes, 268 planning walls, 8283
organizational learning, 85 play, and cultural behaviors, 205
organizational patterns, for survey pneumatics/hydraulics principle, 368
questions, 146 pocket-sized cameras, ethnographic
organization-related attributes, 268 interview, 194196
orthogonal array experiment, 29 Poisson distribution, 391
other way around principle, 366 population, 139, 149150, 155157
overhead, 303 porous materials principle, 368
overhead projectors, 351353 positioning, 287290
oxidants, 369 pregnancy stage, technical system
evolution, 358
P preliminary action, 365366
packages, 269 preliminary action principle, 365
paired comparisons, 331, 332, preliminary anti-action
333335 principle, 365
paper-based surveys, 166167 price, 94, 279280
parameter changes principle, 369 price profile, market-perceived,
parameters, statistical, 391392 104107
Pareto, Alfredo, 332 pricing strategy, 292, 293
Pareto voting, 331, 332333 primary path, 321
Paretos Law of Maldistribution, 332 principles. See inventive principles
partial actions, 366 probability density functions,
partial contradiction matrix, 371 386387
partial/excessive actions probability distributions, 386392
principle, 366 discrete and continuous random
participant observation, 173, 185 variables, 386387
412 Index
probability distributions (Cont.) product cost, 914

expected values, variance and product design, 2533
standard deviation, 387388 computer simulation models,
models of, 388391 2931
statistical parameter estimation, design of experiments, 2529
391392 prototypes, 3133
probability functions, 386387 product scope, 266
probability sampling, 150 product usage processes, 198203
problem solving, 2022, 298. See also product-brand coordination, 290291
Theory of Inventive Problem profit, defined, 304
Solving (TRIZ or TIPS) profit improvement target, 324
process capability, 392395 profitability, 2
process domains, 3637 profit-improvement potential,
process efficiency, 68 324, 328
process lead time, 67 project control rooms, 5859
process performance, 392393 project process flow chart, 339340
process spread, 393 promotion, 280
process variability, 392 proportional variables, 153155
process variables (PVs), 37 prototypes, 3133
product development, 987. See also psychographics of informants, 181
lean product development public relations, 293
concept development, 1825 Pugh concept selection, 24
Pugh concept selection, 24 pull-based production, 7071
quality function deployment, purchasers, 119
2223 purchasing, 304
set-based design and modularity, push-based production, 70
2425 PVs (process variables), 37
theory of inventive problem
solving, 2022 Q
user innovation, 20 QFD. See Quality Function
value engineering, 22 Deployment (QFD)
customer and business quality, 9495, 266
requirement study, 1718 quality characteristics, 255
flowchart, 1314 Quality Function Deployment (QFD),
front-loading process of, 7678 8, 2223, 110, 229258
information/knowledge creation affinity diagram, 233
axiomatic design, 3543 analysis of, 240241
defining information and benefits, requirements, and
knowledge, 46 practicalities, 231232
design as information production effect on project resources,
factory, 4345 chart, 230
information mining, 4653 history of, 231
information transformation, House of Quality, 230, 233
5356 information system design,
overview, 3335 241251
opportunity identification and idea HOWs importance calculation,
generation, 1417 248249
overview, 9 overview, 241
Index 413
Phase I diagnostics, 249251 probability distribution models,

ranking customer input, 388391
241243 statistical parameter estimation,
ranking functional requirements, 391392
243248 ranges, 27, 384
Kano model of quality, 239240 ranking, 333
methodology, 232239 rapid development stage, technical
competitive assessments, 239 systems, 359
critical-to-satisfaction rating survey questions, 142
requirements, 235, 238 ratings, importance, 237
customer attributes, 234235 recalls, 293
importance ratings, 237 recommendations, 337
overview, 232234 recovering, 369
planning matrix, 237238 recreation, 205
relationship matrix, 235236 red sea markets, 111
targets and limits, 238 Reeve, David, 298, 304, 310
Yaesu Book Center, 251258 relationship matrix, 235236, 246248
customer attributes (WHATs), relationships and cultural
254255 meanings, 207
importance ratings for customer relative standing, 385386
attributes, 255 reporting phase, value-engineering
operations items, 255 job plan, 336337
overview, 251254 reports
phase analysis, 255258 affinity diagrams, 216221
quality characteristics survey, 140
(HOWs), 255 requirement studies, customer and
quality measures, 392397 business, 1718
overview, 392 research methods, ethnographic,
process capability indices, 393395 175177
process performance and process research project plan, 187189
capability, 392393 response rates, Internet survey, 166
Sigma, 395397 response variables, selection of, 27
quality, perceived, 104107, 277 restaurant service, customer needs
quantitative CTQs for restaurants, for, 219221
225227 rituals, cultural, 204
quantitative data analysis, robust design, 29
221223 robust technology development, 59
quantitative measures, 226 roles, cultural, 204205
questionnaires. See survey
instruments S
queueing theory, 7981 sales point, 237238
sampling, 148161, 186
R determining size of, 153157
random sampling, 151 methods of, 150153
random variables, 380, 386392 overview, 139, 148149
discrete and continuous, 386387 population and sampling frame,
expected values, variance and 149150
standard deviation, 387388 snowball, 153
Index 415
subjective measures, ethnographic partial contradiction matrix, 371

research, 200 principles, 364369
Suh, 3536, 42 overview, 355356
suppliers, 304 technical dominance, 12
supply-and-demand equation, 10 Technical Importance Ratings,
supporting functions, 318, 349350 236, 237
survey instruments, 139147. See technical systems, evolution of,
also customers, surveys of 357360
close-ended questions, 141144 technological evolution analysis, 348
formatting for Internet surveys, technology
163165 adaption of, 86
open-ended questions, 144 development of, 59
order of questions, 145147 telephone surveys, 135, 148
overview, 139141 Theory of Inventive Problem Solving
questionnaire length, 147 (TRIZ or TIPS), 20, 34, 345378
wording of questions, 144145 contradictions, 21, 355357
sustainability principle, 290 evolution of technical systems,
sustainable prices, 2 357360
symbols function modeling and analysis,
and cultural meanings, 206 349353
in FAST diagram construction, ideal final solution, 60
319320 ideality, 354355
system constraints, 225 overview, 5657, 345349
systematic sampling, 151152 problem-solving process,
360363
T technical contradiction elimination
tacit knowledge, 51, 58 and inventive principles,
Taguchi method, 29 363378
taking out principle, 364 contradiction table of inventive
tangible products, 9 principles, 372377
target market segment, 288289 improving wrench design using,
target orientations, 238 369371, 378
target populations, Internet overview, 363364
survey, 158 partial contradiction
targets, 238 matrix, 371
team rules, ethnographic projects, principles, 364369
189191 use of resources, 353354
technical competence, 7576 thermal expansion principle, 369
Technical Competitive Assessment, thin films, 368
239 time resources, 354
technical contradictions TIPS. See Theory of Inventive
elimination of and inventive Problem Solving
principles, 363378 toaster customer value analysis,
contradiction table of inventive 105107
principles, 372377 topographic surveys, 340
improving wrench design using, total cost, defined, 304
369371, 378 TRIZ. See Theory of Inventive
overview, 363364 Problem Solving
416 Index
U value proposition, 270271

uncontrollable factors, 278, 280281 value-engineering study, 306, 337
uncoupled design matrices, 3839 values, 178, 206
uniqueness principle, 290 Values layer, 273
unit of analysis, 149 values principle, 290
universality, 365 value-stream mapping, 6668
universality principle, 365 variable burden, defined, 303
upstream stages, 14 variable data, 221
usability research, product usage variables. See random variables
processes, 199203 variance, 384388
variation, measures of, 384385
V verbs in functional analysis
value engineering, 22, 297343 diagrams, 351
automobile dealership construction versatility of value, 9597
(Park 1999), 339340 vibration, mechanical, 367
collecting information and creating videos, 175, 178, 187
design alternatives, 298300 visible knowledge, 8183
creative phase, 328329 visual communications, 86
developing alternatives, 299 von Hippel, Eric, 20
engineering department
organization analysis (Park W
1999), 341343 waste elimination, 6566
evaluating, planning, reporting and water faucet design, 3941
implementing, 300 web posted survey instruments, 162
evaluation phase, 329335 website-based solicitation, Internet
complex evaluations, 331332 surveys, 161
selection and screening wine industry, 112116
techniques, 332335 work functions, 308
simple evaluations, 330331 work-in-process inventories, 70
implementation phase, 337339 workplaces, usability research, 203
information phase, 300327 wrench design using inventive
function analysis and evaluation, principles, 369371, 378
314327
function determination, 306314 Y
information development, Yaesu Book Center, 251258
301306 customer attributes, 254255
overview, 300301 operations items, 255
job plan, 300 phase analysis, 255258
overview, 298300 quality characteristics, 255
planning phase, 335336 Yokoya, Yuji, 171
reporting phase, 336337 Youth Assistance Program, 304305
414 Index
SAVE (Society of American Value snowball sampling, 153

Engineers), 297, 314 Society of American Value Engineers
screening, 181185, 329330 (SAVE), 297, 314
S-curve analysis, 362 solicitation, e-mail, 160161
segmentation principle, 364 spheroidality principle, 366
self analysis, 285286 sponsorship, 270
self-expressive benefits, 271 squared deviation, 384385
self-service principle, 368 Squires, 175, 176
sell functions, 309 stage-gate process, 14
semiconductor manufacturing standard deviation, 385, 387388
process, 380 standardization, 8586
service industry, 304 Starbucks brand contract, 289290
services, complementary, 121122 statistical parameter estimation,
set-based design, 2425, 7778 391392
Sherden, William, 89 statistically designed experiments.
sidewalk surveys, 152 See Design of Experiments
signs and cultural meanings, 206 (DOE)
simulation models, 2931 statistics. See descriptive statistics
Six Sigma, 379397 strategic analysis of brand
and data analysis, 379380 development, 283287
descriptive statistics, 380386 competitor analysis, 286287
box plots, 382 customer analysis, 284285
dot plots, 380 self-analysis, 285286
histograms, 380381 vision development, 283284
numerical descriptive statistics, strategic groups within industries,
382386 118119
overview, 379 strategy development, brands,
quality measures and, 392397 287293
overview, 392 benefits and points of
process capability indices, differentiation, 289
393395 brand positions of bookstores, 290
process performance and process channel strategy, 292293
capability, 392393 communications and marketing
Sigma quality level, 395397 strategy, 291292
random variables and probability contract with market, 289
distributions, 386392 positioning, 287290
discrete and continuous random pricing strategy, 293
variable, 386387 principles of effective brand
expected values, variance and positioning, 290
standard deviation, 387388 product-brand coordination,
probability distribution models, 290291
388391 Starbucks implicit brand
statistical parameter estimation, contract, 289290
391392 target market segment,
sketches, 337 288289
skipping principle, 367 stratified random sampling, 152
slang, 206 strong brands, 260261
small-scale surveys, 167 strong oxidants principle, 369

Voice of The Customer - Capture and Analysis

Uploaded by

Copyright:

Available Formats

Voice of The Customer - Capture and Analysis

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Voice of The Customer - Capture and Analysis

Uploaded by

Copyright:

Available Formats

Voice of the Customer: Capture and Analysis

Dr. Kai Yang

New York Chicago San Francisco Lisbon London Madrid

THE WORK IS PROVIDED AS IS. McGRAW-HILL AND ITS LICENSORS MAKE NO

Want to learn more?

Chapter 1. Value, Innovation, and the Voice of the Customer 1

Chapter 2. The Product Development Process 9

Chapter 3. Customer Value and the Voice of the Customer 89

Chapter 4. Customer Survey Design, Administration, and Analysis 133

4.6.2 Administering a Survey on the Internet 162

Chapter 5. Proactive Customer Information Gathering

Chapter 6. VOC Data Processing 213

Chapter 7. Quality Function Deployment (QFD) 229

7.8 QFD Case Study 2: Yaesu Book Center 251

Chapter 8. Customer Value Creation by Brand Development 259

Chapter 9. Value Engineering 297

10.2.3 Ideality 354

Chapter 11. Statistical Basics and Six Sigma Metrics 379

Value, Innovation, and the

In todays global economy, several business functions have become global

Value level Value level

Figure 1.1 Value and innovation

customer-centric innovation. In both aspects, thorough understanding

1.1 Defining Customer Value

Service and convenience benefits This category includes avail-

Economic liabilities This category includes all monetary expenses

General Motors have a joint venture in California that produces an

1.2 Innovation Roadmap

We define innovation as the successful implementation of creative ideas

1.3 Voice of the Customer:

factor, whether you are dealing with a technology push-or-market pull

1.4 Overview of This Book

The Product Development

2.1 Defining Product Cost and Development

Value = Benets Liabilities

Figure 2.2 Business Operation of Product Producing Companies

Cost of product development

Figure 2.3 The Economic Model of Product Development and

Product Performance Product performance depends on how much

Robustness In addition, the robustness (including quality, reliabil-

On a longer time scale, product development lead time, efficiency,

2.1.1 Product Development Process

1. Determine that a new product is required to serve some need

2.2 The Product Development

2.2.1 Opportunity Identification and

Stage 1: Customer and business requirements study

Stage 2: Concept Development

Stage 3: Product/Service Design/Prototyping

Stage 4: Manufacturing Process Preparation/Product Launch

Stage 6: Product/Service Consumption

Figure 2.4 A typical product/service life cycle

identified by marketing people or upper management, and then the

is technology analysis. Chapter 6 of this book will discuss voice-of-

2.2.2 Customer and Business Requirement

2.2.3 Concept Development: Stage 2

2.2.3.1 User Innovation Sometimes the best product concept ideas

2.2.3.2 Theory of Inventive Problem Solving TRIZ (Teoriya Resheniya

Though there are millions of patents and inventions, all innovations

and learned. By learning and using these principles and strategies,