Voice of The Customer - Capture and Analysis
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
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.
DOI: 10.1036/0071465448
Professional
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
vii
viii Contents
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
References 399
Index 403
This page intentionally left blank
Chapter
Copyright 2008 by The McGraw-Hill Companies. Click here for terms of use.
2 Chapter One
Most
Microsoft operation companies Starbucks
systems
iPod Nano
Intel CPUs
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
This book will show you how to capture and analyze the voice of the
customer in all these four scenarios.
8 Chapter One
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.
Copyright 2008 by The McGraw-Hill Companies. Click here for terms of use.
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:
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
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:
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
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.
The Product Development Process 13
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.
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 5: Production
Process operation, control, and adjustment
Supplier/parts management
Stage 7: Disposal
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.
(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.
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.
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
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.
Controllable Factors
x1 x2 ..... xn
...
Input Output y
Process Response
...
z1 z2 ..... zp Figure 2.8 A Process Model of DOE
Uncontrollable factors
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?
The Product Development Process 27
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
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
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
The Product Development Process 37
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
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:
The Product Development Process 39
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.
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
1 2
2
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
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
1 m m
I = log 2 = log 2 P = log 2 Pi = log 2 Pi (2-7)
P i =1 i =1
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
The Product Development Process 45
Information
Modified
Information Knowledge and/or New
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:
Internal Customers
Y=f(x)
FRs DPs
External CAs PVs
.
Customers .
.
. . .
.
. . .
. Y=f(x) .
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.
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.
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.
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,
The Product Development Process 55
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
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.
Explicit Knowledge
At Next Level
Tacit Knowledge
At Next Level
Explicit
Knowledge
Figure 2.20 Knowledge Creation
Tacit
Spiral
Knowledge
58 Chapter Two
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
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).
Transportation
Staging Setup Inspection Staging
Time
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:
Start
End
3 min 60 min
confirm wait for
PC
One-piece ow
Work cells (cellular manufacturing)
Pull-based production
Raw Product
Material
Batches of Semi
Finished products
Figure 2.24 Flow Patterns of a Batch Flow Shop
70 Chapter Two
Raw Product
Material
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.
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:
Here are some of the most important lean concepts in the product
development process:
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.
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.
Reinvention
Mismatch of subsystems
Information loss and re-creation
Miscommunication
The Product Development Process 79
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.
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.
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
Problem statement
Current situation
Prior research or work
Root cause analysis
Experimental methods
Data analysis
Recommendations
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,
The Product Development Process 83
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.
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.
Product Design
Process Design
Time
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)
Low High
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.
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
Copyright 2008 by The McGraw-Hill Companies. Click here for terms of use.
90 Chapter Three
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.
Functions
Kaufman V=
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:
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).
Customer Value and the Voice of the Customer 93
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)
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.
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.
Use value Properties that make something work; this can also be
called functional value.
96 Chapter Three
Self-
Esteem Actualization
Belonging
Safety
Relative Importance
Survival
Psychological Development
Figure 3.1 Value and Psychological Development
Customer Value and the Voice of the Customer 97
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
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
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
Customer Value and the Voice of the Customer 105
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:
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.
Machine Up Time
Print Speed
Image Quality
Ease of Use
Repair Time
Quality of Service
Purchase Price
Toner
Paper
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
Customer Value and the Voice of the Customer 111
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:
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
Customer Value and the Voice of the Customer 113
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
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?
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
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:
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.
Alternative industries
Strategic groups
Buyer groups
Complementary product and service offerings
Functional-emotional orientation
Time
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.
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
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
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.
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.
Matrix A Y=f(x)
Matrix B
Y=f(x)
Physical Mapping Process Mapping
Customer Functional Design Process
Attributes Requirements Parameters Variables
1. Product development
2. Market and product planning
Customer Value and the Voice of the Customer 127
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
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
Customer Value and the Voice of the Customer 131
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
133
Copyright 2008 by The McGraw-Hill Companies. Click here for terms of use.
134 Chapter Four
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.
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.
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.
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.
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:
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.
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:
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.
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?
______________________________________________________________
Customer Survey Design, Administration, and Analysis 147
The respondents will be able to answer the questions, but they will get
disoriented after answering 10 or more of these misplaced questions.
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.
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.
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:
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.
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.
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(1 p)
p = Z = 3%
2
n
Customer Survey Design, Administration, and Analysis 155
Therefore,
2
Z p(1 p)
n= 2 (4-4)
p
Z2 [ p(1 p)]N
n= 2 (4-5)
Z2 [ p(1 p)] + ( N 1) 2p
2
156 Chapter Four
This sample size is smaller than the one calculated in Example 4.4.
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
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
Z2 2
n= 2 1.962 60002
2 + Z2 2 ( N 1) = 1.962 60002 = 135
2 10002 +
4999
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.
4.6.1.4 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.
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:
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.
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.
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.
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.
Response rate
Cost
Lead time
Data quality
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
169
Copyright 2008 by The McGraw-Hill Companies. Click here for terms of use.
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
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.
Contextual observations
Participant observations
Video recording
Open-ended interviews
Focus groups
Content analysis (text, audio, and video)
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.
Contextual observations
Directed and semistructured interviews
Ratings, rankings, and comparisons of the relevant products
Participant observation
Scenario analysis
Surveys
Focus groups
Usability tests
Product simulation
User-experience interviews
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.
and them screening them. The initial pool of candidates can come from
the following sources:
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)
Proactive Customer Information GatheringEthnographic Methods 185
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:
The selection of data collection methods should fit the objectives of the
ethnographic research project and should be practical in the field.
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
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.
Proactive Customer Information GatheringEthnographic Methods 189
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.
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.
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
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.
important, the brand? The quality? The price? The next two examples
illustrate these objectives.
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?
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.
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.
Proactive Customer Information GatheringEthnographic Methods 203
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.
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.
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.
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):
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
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).
213
Copyright 2008 by The McGraw-Hill Companies. Click here for terms of use.
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
6.2.2.1 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:
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.
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
VOC Data Processing 217
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
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
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)
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
TABLE 6.2 Two Levels of Affinity Groups for Restaurant Service (Ramaswamy 1996)
TABLE 6.4 Statistical Measures for Data with Different Measurement Scales
Characteristic
Weight of saw
(Y)
Pure weight in kg
Measure (Weight of electrical
wire not included)
Target 2 kg
Delivery Speed
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:
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.
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
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
229
Copyright 2008 by The McGraw-Hill Companies. Click here for terms of use.
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
HOWs
CTSs
HOWs
Functional
Attributes
Customer
WHATs
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
HOWS
Operation
CTSs
Variables
Attributes
Customer
WHATS
WHATS
House of House of
CTSs
quality Quality
#1 #2
CTS Correlation
CTSs
(Hows)
Direction of Improvement
Customer Desirability
Attributes
Customer
(Whats)
Relationship
Planning Matrix
Matrix
Importance Rating
Competitive
Benchmarks
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):
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.
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
Importance
Extremely Important 5.0
Very Important 4.0
Somewhat Important 3.0
A little Important 2.0
Not Important 1.0
Trade Offs
Synergy 1.0
Compromise 1.0
Excitement
Quality
Performance
Satisfaction
Customer
Quality
Wow!
Degree of CTS
Achievement
..
of...
re Basic
Mo Unspoken Wants
ve Quality
Gi
Customers
HOWs
Data Workers
Weight
Site Manager
Partners
WHATs
Intensity of Use 5
CRITERIA
1 49. 147
3 15. 47
ABSOLUTE IMPORTANCE
RELATIVE IMPORTANCE
RANK
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
5
y2 = pi 2 xi = 9 5 + 3 1 + 9 5 + 3 3 + 9 5 = 147
i=1
5
y3 = pi3 xi = 3 5 + 1 5 + 9 3 + 9 5 = 47
i=1
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.
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).
Project Objectives:
Design a global commercial process with high performance.
MANAGEMENT REPORTS
IMPORTANCE
Fuel Consumption
Distance Traveled
HOWs
Run Distribution
Hour Statistics
Market Visits
Expenditures
Truck Usage
Payments
Failures
WHATs
FUNCTIONAL REQUIREMENTS
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
Quality Function Deployment (QFD) 245
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%.
7.7.1.1 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
Direction of Improvement
Available Products
Professional Staff
Flexible Processes
Knowledgeable Staff
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
Quality Function Deployment (QFD) 247
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.
Consultative
Answer Questions
Courteous
Adequate Follow-Up
Direction of Improvement
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
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
Competitor 2: Dealers
Cost-Effective Products
Competitor 4: Competitor B
Competitor 3: Competitor A
Importance of Product Attributes
Chapter Seven
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
Competitor 2: Dealers
Trade off
Competitor 3: Competitor A
Competitor 4: Competitor B
1.0
1.0
Min = 1.5
9
8
7
6
5
4
3
2
1
Quality Function Deployment (QFD) 251
% of Employees Trained
Systems Uptime
Target Values
Maximize 1.0
Target 0.0
Minimize 1.0
1
3
Direction of Improvement 1
% Svc Req Res by Promise Date 2 5004.0 15.5 100% of Service Requests Resolved by Promise Date 2
Average Speed of Answer 8 1278.0 4.0 80% of Calls Answered in < 24 Seconds 8
18.6 483.3
17.1 443.7
7.8 202.9
89.6
89.6
53.2
3.4
2.0
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%
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:
Product Trade-offs
Competitor 3: Competitor A
Competitor 4: Competitor B
Direction of Improvement
Competitor 2: Dealers
Adequate Follow-Up
Improvement
Answer Questions
Target Values
Maximize 1.0
Consultative
Our Product
Courteous
Target 0.0
Minimize 1.0
10
1
8
Direction of Improvement 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
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
3.1
1.5
Trade Offs
Our Product 3
3
Synergy 1.0
2.5
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
10
1
6
7
Figure 7.12 shows a partial tree diagram that organizes the customer
attributes in this case.
Customer
Satisfaction
Pleasantness Fast
Clean books
First Pleasant
Fast
Level Atmosphere
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
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
Quality Function Deployment (QFD) 257
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
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
0.1
Operation Items
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
Delivery rate
Procurement time
Cleanliness
Lighting
Damage rate
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
259
Copyright 2008 by The McGraw-Hill Companies. Click here for terms of use.
260 Chapter Eight
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
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
Customer Value Creation by Brand Development 263
Regular
Knowledge Brand Brand
Un- Brand Brand
Awareness Preference Repurchase Repeat
awareness Trial
of Features
Purchase
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.
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
BrandCustomer Relationship
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.
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
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
Beliefs Proud
and Self-esteem
values Confident
Happiness
Fulfilled
Associated with
Benefits most respected designer
High status customers
feel like they are in style
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).
Customer Value Creation by Brand Development 275
Brand Identity
Signal Brand
Transmitted Image
Organization
Factors
(controllable) Market Factors
(Competition
and Noise,
Uncontrollable)
Brand identity
Marketing mix factors, or the four Ps: product, price, promotion, and
place
Time to market
Value Proposition
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.
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
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
Evaluation Phase
Measuring the
Return on Brand
Investment
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.
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
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
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:
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:
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.
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
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.
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.
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
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
Copyright 2008 by The McGraw-Hill Companies. Click here for terms of use.
298 Chapter Nine
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.
1. Information phase
2. Creative phase
3. Evaluation phase
4. Planning phase
5. Reporting phase
6. Implementation phase
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.
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.
Cost Visibility
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.
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.
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
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
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.
Value Engineering 309
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.
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.
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
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
Value Engineering 311
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 (continued)
List All Functions
Verb Noun Basic Second Remarks
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
Why How
Eliminate
Deviancy
Modify Identify
Behavior 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
Objectives
or
Specifications
High Lowest
Order Basic Dependent Order
Function Function Function Function
(Output) (Input)
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
Activity
Construct
Pool
Build Contract
Swim Club Builder
Construct
Club House
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.
Review
Proposals
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:
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.
Monitor
Identify Performance
Discrepancies
Apply
Skill
Inform
Customer
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.
Release How?
Fuel
Ignite Produce Ignite
Cigarette Flame Fuel
Produce How?
Spark
How Why
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
Value Engineering 327
3 $4.69
Plan
Treatment
$4.07
Figure 9.16 A fully-marked function
block in a FAST diagram
Function Name Best Cost
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.
Figure 9.17 A portion of a FAST diagram with fully-marked function blocks for youth
assistance program
328 Chapter Nine
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:
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.
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.
Idea 1 Idea 2
Advantage Disadvantage Advantage Disadvantage
Idea 3 Idea 4
Advantage Disadvantage Advantage Disadvantage
332 Chapter Nine
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.
B C D E
A B-2 A-1 D-2 E-2
C D-1 E-3
D E-2
3 Significant
2 Moderate
1 Minimal
Value Engineering 335
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
Establish
Budget
Resolve
Restrictions
When Purchase
Property
Why Scope
Estimate
Cost
Identify
Requirements
Negotiate
Contract
345
Copyright 2008 by The McGraw-Hill Companies. Click here for terms of use.
346 Chapter Ten
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.
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.
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.
Wear
Tear Gums
352 Chapter Ten
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.
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
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
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
Customer Value Creation Through Creative Design (TRIZ) 355
Where
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.
12 Shape 32 Manufacturability
14 Strength 34 Repairability
Explosives
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.
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)
New System
Effort (Time)
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.
Customer Value Creation Through Creative Design (TRIZ) 361
Brush
Toothbrush Teeth
Mech.
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
Physical contradiction
Method: Separation principles
Technical contradiction
Method: Inventive principles
Customer Value Creation Through Creative Design (TRIZ) 363
functions.
There are excessive harmful functions.
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-
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-
What is deteriorated ?
improved?
35. Adaptation
27. Reliability
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)
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)
objects operation
17. Temperature
18. Illumination
fixed object
14. Strength
21. Power
stance
What should be improved?
37. Measurement or
test complexity
29. Manufacturing
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
379
Copyright 2008 by The McGraw-Hill Companies. Click here for terms of use.
380 Chapter Eleven
of random variables, that is, their value will vary with some degree of
uncertainty. Consider the case in Example 11.1.
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
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
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.
9
8
7
6
Frequency
5
4
3
2
1
0
520 540 560 580 600 620
Film thickness
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.
600
Film thickness
580
560
540
520
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
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.
1 n
n 1
s2 = ( yi y)2 . (11.4)
i =1
1
s2 = [(547 572.02)2 + (563 572.02)2 + ... + (575 572.02)2 ]
47 1
= 601.72
1 n
n 1
s = s2 = ( yi y)2 . (11.5)
i=1
s = s2 = 601.72 = 24.53.
p( yi ) = 1.
All yi
f ( y) 0
b
P(a y b) = f ( y)dy
a
+
f ( y)dy = 1.
= E( y) = yp( y) if y is discrete
All y
+
= E( y) = yf ( y)dy if y is continuous.
2 = ( y )2 p( y) if y is discrete
All y
+
2 = ( y )2 f ( y)dy if y is continuous.
2 = Var( y) = E[( y )2 ] .
388 Chapter Eleven
= 2
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:
E(y) =
Var(y) = 2
f(y)
Distribution
with smaller
Distribution
with larger
68.27%
95.45%
99.73%
Figure 11.5 Percentage distribution properties of normal random variable
390 Chapter Eleven
E(y) = = Var(y) = 2 .
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.
y e y = 0, 1, 2,
p( y) =
y!
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
USL LSL
C pk = Min , = Min CPU , CPL .{ }
3 3
LSL = 18 USL = 24
Process center
Center = 22
24 22 22 18
C pk = Min , = Min(0.83, 1.67) = 0.83 .
38 38
Cp = Cpk = 1.25.
LSL USL
6 3 +3 +6
Figure 11.6 Six Sigma quality
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.
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.
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
Copyright 2008 by The McGraw-Hill Companies. Click here for terms of use.
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
403
Copyright 2008 by The McGraw-Hill Companies. Click here for terms of use.
404 Index
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