Lockman 2010 B

CRC PRESS .
TAYLOR & FRANCIS
Kansei/Affective Engineering
A sampler
www.crcpress.com
Contents
Page 3 - Chapter 8 - Kansei/Affective Engineering and Web Design
Page 28 - Chapter 9 - Kansei/Affective Engineering for the European

Fast-Moving Consumer Goods Industry
Page 50 - Chapter 11 - Kansei, Quality, and Quality Function Deployment
https://www.crcpress.com/KanseiAffective-Engineering/Nagamachi/p/book/9781439821336
20% Discount Available*

Enjoy a 20% discount across all books featured here. Simply add the discount code
ACT39 at the checkout.
*Please note: This discount code cannot be combined with any other discount or offer and is only valid on print titles
purchased directly from www.crcpress.com. This discount is valid until 30/06/2018.
Publishing with CRC Press/Taylor and Francis

If you’ve ever considered writing a book in your area of expertise, this is your chance.
Please contact: Cindy Renee Carelli, Executive Editor, CRC Press – Taylor & Francis Group, LLC
Email: cindy.carelli@taylorandfrancis.com
www.crcpress.com
8
and Web Design
Anitawati Mohd Lokman
Contents
8.1 Introduction to Web Design...................................................................... 228
s
8.1.1 Positioning Kansei/Affective Engineering in Web Design...... 228
ci
8.2 Engineering Kansei in Web Design......................................................... 230
an
8.2.1 Structuring the KAE Method into Kansei Design Model........ 230
8.2.1.1 L1: Synthesizing the Specimen....................................... 231
8.2.1.2 L2: Establishment of Kansei Checklist.......................... 232
Fr
8.2.1.3 L3: Determination of Kansei Concept

and Requirement.............................................................. 232
r&
8.2.1.4 L4: Prototyping/Testing................................................. 232

8.3 Demonstration of the Kansei Design Model Implementation in
Web Design.................................................................................................. 233
o
8.3.1 Synthesizing the Specimen........................................................... 233

8.3.1.1 Identification of Initial Specimens................................. 233
yl
8.3.1.2 Investigation of Design Elements.................................. 233

Ta
8.3.1.3 Classification of Item and Category..............................234

8.3.1.4 Finalizing Valid Specimen.............................................. 235
8.3.2 Establishment of Checklist............................................................ 236
8.3.2.1 Checklist Development................................................... 237
8.3.2.2 Synthesize Kansei Words................................................ 237
8.3.2.3 Development of Kansei Checklist................................. 238
8.3.3 Determination of Kansei Concept and Requirement................ 238
8.3.3.1 Kansei Measurement....................................................... 238
8.3.3.2 Requirement Analysis..................................................... 239
8.3.4 Prototyping/Testing....................................................................... 245
8.3.4.1 Prototyping....................................................................... 245
8.3.4.2 Testing................................................................................ 246
8.4 Summary...................................................................................................... 250
References.............................................................................................................. 251
227
228 Kansei/Affective Engineering
8.1 Introduction to Web Design

A Web site is a collection of related Web pages accessible using a domain
name or IP (Internet protocol) address via the Internet or a private local area
network. A Web page consists of contents such as text, images, videos, or other
kinds of digital assets. Designing a Web site requires skill in creating pre-
sentations of content that is delivered ubiquitously through the World Wide
Web. Typically, the basic components of Web design include content, usability,
appearance, and visibility. Web site design, or Web design, may involve mul-
tiple disciplines in information systems, information technology and commu-
nication design, graphic design, human–computer interaction, information
architecture, interaction design, marketing, photography, search engine opti-
mization, and typography. Thus, to successfully produce a good Web site, a
s
proper approach before one can start creating Web sites is required.
ci
Web site architecture (WA) is an organized approach in planning and
designing a Web site. Like the traditional architecture, WA involves techni-
an
cal, aesthetic, and functional criteria. Similarly, the main focus is on users
and user requirements. Technical aspects of WA address the back-end
Fr
components that deal with the underlying technology such as source code,
data warehouse, and server-side component. It also addresses the business
plan and the information architecture (IA), which deals with the structural
r&
design, method of organizing and labeling Web pages, and ways of demon-
strating the Web site to the digital landscape. The aesthetic aspect addresses
the user friendliness of the Web site and the visual and audible impres-
o
sions. Functional aspects deal with the front-end components, which include
yl
usability, visibility/searchability, and accessibility.

Ta
8.1.1 Positioning Kansei/Affective Engineering in Web Design

In the process of designing Web sites, the aesthetic and functional crite-
ria require particular attention to user experience (UX), the term used to
describe the experience that a user has as a result of his or her interac-
tions with the Web site. The consequences of UX in Web design result
from the user’s perception toward the elements of functional and non-
functional qualities. Perception of functional qualities particularly lies in
the usability and usefulness of the Web site. On the other hand, percep-
tion of nonfunctional qualities particularly lies in the aesthetic aspect,
which includes visual and audible quality of the Web site. The most
important component that cannot be neglected is the affective qualities
that address the aspect of feelings, reaction, appraisal, and behavior from
users. Affective qualities encompass both the components of functional
and nonfunctional qualities.
Kansei/Affective Engineering and Web Design 229
UX
Affective Qualities
(Subjective feelings, physiological reaction, cognitive appraisal,
behavioral tendency)
Non-functional qualities
Functional qualities - Aesthetic
- Usability - Visual
- Usefulness - Audible
- Accessibility - Symbolic
- Motivational
s
Figure. 8.1
Elements of UX in Web design.
ci
an
In the process of designing Web sites, designers have been putting effort
into what the user wants in order to come out with the desired result.
Fr
However, by its very nature and similar to any kind of software system,
Web design always produces conflicts regarding designer specification and
user conformance. In realizing the concept in terms of Web site appearance,
r&
designers often misunderstand the description provided by users or user

advocates, who in turn do not really understand the concept that they actu-
ally want. Designers possibly come out with the specification by their own
o
intuition and creativity. Another problem is that even though the users may
yl
have well described the concept that they want, the designer’s side has no
clue on the design requirements to produce the kind of Web site that the
Ta
users have described.

Kansei/affective engineering (KAE) provides a systematic way of under-
standing the insights of user perceptions toward artifacts via several physi-
ological and psychological measurement methods. These insights are then
translated to the design characteristic of the artifact. This approach of KAE
matches the concern of requirement specifications to correspond to the foun-
dation of UX in Web design, providing the possibility for users to express
their concept and providing clues to designers in the form of design require-
ments so that the designer can objectively develop the desired Web design.
In every aspect that contributes to the desired UX, where the design needs to
match the insights of the users’ experience, KAE is seen as a possible require-
ment generation technique. The implementation of KAE in every possible
aspect of UX in Web design will enable the identification of the concept of
Kansei in Web design and the contributing design specification to the con-
cept in the form of design requirements.
8.2 Engineering Kansei in Web Design

This chapter attempts to demonstrate the implementation of KAE in Web
design, focusing on the visual quality of design that shapes one of the ele-
ments that have consequences on the user experience. The detail of the imple-
mentation is described in the following sections, beginning with the building
of a reference model for the implementation of KAE in Web design.
8.2.1 Structuring the KAE Method into Kansei Design Model

From the review of KAE literature, a gap in terms of the description of steps
to be performed in the implementation method of Kansei engineering was
identified. There are many types of techniques in different kinds of imple-
s
mentation, but the description of the method is largely narrative. For this
ci
reason, based on previous literature involving the adoption of KAE, setting
the foundation to the basic principles of KAE, the method is structured into
an
a model called the Kansei design model. In structuring the model, careful
attention was given to the capacity and availability of infrastructure, facili-
Fr
ties, and cost. The structured model employs a self-reporting method in the
measurement of Kansei, allowing KAE implementation in a basic environ-
ment where no special equipment and skills are required. With this model,
r&
the audience can have a useful guide to the implementation of KAE.

o
yl
L4 - Prototyping/Testing Kansei
Ta
design
specification
L3 - Determination of
Kansei Requirement
Kansei Concept and
measurement analysis
Requirement
Selection of
Synthesize Development
L2 - Establishing domain specific
Kansei words of checklist
of Checklist Kansei words
New
Determine Classify physical Build
L4 - Synthesizing product
elements trait samples
concept
Specimen
Identification of Investigation Classification Determination
specimen from of design of design of valid
existing product elements elements specimen
Figure 8.2
The Kansei design model.
The model is developed to provide a systematic approach to the imple-

mentation of KAE in designing Kansei products. The model presented is a
useful mechanism for industries, designers, academic researchers, and other
stakeholders in discovering Kansei concept, and design requirements for the
development of Kansei products. The model is divided into four levels: L1,
L2, L3, and L4. The following sections describe details of each level.
8.2.1.1 L1: Synthesizing the Specimen

L1 is the level of synthesizing the specimen. The level is subdivided into two
different procedures, PI and PII. These procedures differ in the process of
synthesizing the specimen. There are four steps in both PI and PII, which are
essential in determining a valid specimen. The procedure can be decided
according to one’s objective.
s
PI is applicable to products that already exist in the market where the
ci
maker needs to improvise the design. The procedure begins with a collec-
tion of samples with visible differences from existing products in the market
an
within a specific domain. KAE emphasizes controlling the domain, as the
consumer’s response is unique with different domains (Nagamachi, 2003;
Fr
Ishihara et al., 2005). Previous KAE studies have suggested different tech-
niques for determining specimens from using actual products and using
pictures of product. The choice of specimen depends on its suitability to the
r&
experimental design.
Then, the following procedure is the process of investigating design elements
in all samples. Determination of the number of design elements depends on the
o
level of detail that needs to be included in one study. Controlling the number of
yl
elements enables a more objective measurement. On the other hand, including

Ta
all identifiable elements from the consumer’s point of view enables a more accu-
rate measurement. The latter ensures the accuracy of design requirements as an
outcome of a study, as consumers are assessing a product as a whole. To match
the consumer’s emotional response to design elements, this chapter suggests
that controlling the number of elements will produce a less accurate result.
The next procedure is the classification of design elements. The identified
design elements are further classified into item and category. Item is the type
of common physical traits of all specimens such as background color, body
shape, and text alignment. Category is the specific attribute of the item in
each specimen, such as red as a background color of specimen A, and blue as
a background color of specimen B. The process is crucial since the findings
will be the essence in the success of the requirement analysis stage. Finally,
based on a set of rules, a valid specimen for Kansei measurement can be
synthesized among all the initial samples.
PII, on the contrary, is designed for application when a company or
designer plans to design a new concept of a product based on their objec-
tives. This is applicable to the development of product that has yet existed
on the market. In this case, designers and experts have to determine product
specifications based on their inspiration in relation to the target concept. For
instance, to design an elegant mobile phone, the process begins with syn-
thesizing words related to the concept of elegant within the domain. Then,
designers or experts have to determine design elements that have a connec-
tion with an elegant feeling, classify the physical traits, and build a number
of prototypes based on the technical specification. This prototype will then
be used as a specimen at the following level of the model to confirm their
design with consumers.
8.2.1.2 L2: Establishment of Kansei Checklist

L2 describes preparation and establishment of Kansei checklist. The level
is divided into three steps: (1) synthesizing Kansei words, (2) selection of
s
domain-specific Kansei words, and (3) development of checklist. The level
ci
synthesizes Kansei words, from a larger number of possible Kansei words
to focused Kansei words that highly related to the product domain. Kansei
an
words can be adjectives, such as calm, sophisticated, and natural, or nouns.
These Kansei words can be synthesized from pertinent literature, technical
Fr
magazines, or even consulting experts. Finally, utilizing the Kansei words,

L2 produces a Kansei checklist in the form of the semantic differentials (SD)
scale as a measurement tool for Kansei measurement in the next level.
r&
8.2.1.3 L3: Determination of Kansei Concept and Requirement

o
yl
L3 describes determination of the Kansei concept and requirement. This

level is divided into two steps: (1) Kansei measurement, and (2) requirement
Ta
analysis. In the first step, Kansei measurement is performed using expert or

ordinary consumers as test subjects. The subjects are required to rate their
impressions toward product specimen on the Kansei checklist. Results from
the evaluation will be analyzed to interpret links between subjects’ Kansei
and design elements identified in L1. The outcome can be used to determine
design requirements for the development of the Kansei product.
8.2.1.4 L4: Prototyping/Testing

L4 describes prototyping/testing. In this final level, the results from L3
will be used as the foundation to build a prototype of the Kansei product.
The process will involve the employment of the Kansei concept and design
requirements identified in L3. To develop a successful Kansei product,
experts’ creativity should be included in the design process. Testing may be
performed to validate the design requirements.
8.3 Demonstration of the Kansei Design Model

Implementation in Web Design
This section demonstrates the implementation of the constructed Kansei
design model in Web design to explore the potential of Kansei engineering
in Web design.
8.3.1 Synthesizing the Specimen

In order to implement KAE in Web site interface design, the first procedure,
PI, of the Kansei design model is adopted. The procedure is applicable in
engineering the Web site interface designs and provide guidelines for design-
ers to improvise Web site interface design into Kansei design. This section
s
focuses on the context of interface design that is visible to users of clothing
ci
Web sites.
Four stages are involved in implementing level 1 of the model:
an
1. Identification of initial specimens
Fr
2. Investigation of design elements

3. Classification of item and category
r&
4. Finalizing valid specimen

o
8.3.1.1 Identification of Initial Specimens

yl
Table 8.1 shows the controls and criteria used to select initial specimens as
part of preparation of the instruments to be used in the Kansei measurement
Ta
process.
In identifying the initial specimens, 163 Web sites were selected based
on their visible design differences in both content and layout context (color,
typography, layout, etc.). These Web sites were chosen according to their
listing on the Apparel Search Web site (http://www.apparelsearch.com).
Apparel Search is the leading online clothing directory and has the categori-
zation structure that helped in selecting Web sites.
8.3.1.2 Investigation of Design Elements

Each component within the basic structure of the Web site was used as the
basis during the empirical investigation of all design elements that compose
all of the 163 specimens. The design elements are broken down into item and
category so that clear categorization of designs can be organized. In total
77 items have been identified from the empirical investigation. These items
Table 8.1
Control Condition
No. Item Condition
1. Web site criterion Visible differences in design
2. Focus context Design content and layout
3. Screen resolution 1024 × 768 pixels
4. Access/download date June 1–30, 2006
5. Platform Win32
6. Operating system Windows XP
7. Color quality 32 bit
8. Browser Opera 9.00
9. Browser control Encoding = Windows-1252
Default language = English, [en]
s
Default Text Size = Medium
ci
Colors = Windows 32 bit color
10. Encoding Windows-1252
an
11. Language English
12. Default text size Medium (3 pt)
Fr
were then investigated individually from each Web site to identify categories
assigned to each. Table 8.2 summarizes all design elements in the 163 Web
r&
sites that are transparent from the viewpoint of Web site visitors.
o
8.3.1.3 Classification of Item and Category

yl
In the context of the basic structure of a Web page, each specimen may com-
prise all or part of the elements within each section: Body (refers to the layer
Ta
that a Web page may reside on), Page (refers to the page of a Web site), Header
(refers to the head section of the Web page, which may contain the menu),
Top Menu (refers to the menu on the top of the page), Left Menu (refers to the
menu on the left pane of the Web page), Main (refers to the main body of the
Web page), Right Menu (refers to the right menu on the right pane of the Web
page), and Footer (refers to the footer of the Web page, which may contain
a menu). The Web page may also contain pictures and other elements such
as artistic menus and logo. From the set of items identified in the previous
section, the classifications of categories are identified that form the different
characteristics of Web site designs. Table 8.3 gives examples of item and cate-
gory to be investigated from all specimens. A total of 77 item and 249 category
of Web site design elements were identified.
To simplify the organization of the huge amount of data, all the identi-
fied design elements (item and category) were organized into specimens
by design elements matrix. Each specimen was carefully investigated to
check the item and category that make up the characteristic of the specimen.
Table 8.2
Item Identified from the Initial Specimens
Section Item
Body Background color, background style
Page Shape, menu shape, style, orientation, color, size, border existence
Header Existence, background color, background picture existence, font size, menu
existence, menu link style, menu background color, menu font size, menu font
family, menu font style
Main Background color, background picture existence, shape, adv. Existence, text
existence, text alignment, font color, font size, font family, font style
Top menu Existence, location, link style, background color, font color, font size, font
family, font style
Right menu Existence, style, font size
Left menu Existence, link style, background color, font color, font size, font family, font style
s
Footer Existence, menu existence, menu link style, menu background color, menu font
ci
color, menu font size, menu font family, menu font style, shape
Picture Existence, size, dimension, focus, arrangement, style, image used?, No of
an
people in 1 picture, body representation type, face expression, face facing?
Empty space? Other images? Product display style, product try on? Product
view style
Fr
Others Dominant item, artistic menu used? Discount advertisement existence, logo
existence, logo location
r&
Table 8.3
o
Example of Design Item and Category

yl
Item Category
Ta
Page background color Blue

Left menu style Button
Main text size Medium
When the category matched the investigated specimen, then the matrix was
checked. The process is repeated until investigation of the 249 categories in
all specimens is completed. Although the construction of a matrix does not
substantially reduce the amount of work, which is impossible, it offers easy
management of the knowledge by providing orderly data organization. The
matrix data also alleviate the screening procedure, involving 249 categories
in 163 specimens, for identifying a valid specimen.
8.3.1.4 Finalizing Valid Specimen

Results from the item/category classification stage were then examined
according to the following rules in conformance to KAE methodology
Page Bg Color Top Menu Location Font Size Picture Size ...
Specimen
No. None Green Blue Grey Brown Mix Left Right Center S M L S M L
2
Rule No. 1
.
.
.
Rule No. 2 163
Rule No. 3
s
ci
Figure 8.3
Rules in screening specimens.
an
1. For each sample, only one category under one Item is ticked.
Fr
2. Only one sample will be taken if exactly the same category under
each Item is ticked for more than one Web site.
3. Take two or more sample Web sites where same category is ticked.
r&
Figure 8.3 illustrates how the rules are executed in screening the Web site
specimen to identify valid samples.
o
Conforming the first rule, for Page Bg Color every specimen must only
yl
have one color checked. Secondly, specimens 4 and 163 have exactly the
Ta
same result, so only one can be included as a valid sample. Finally, two or
more samples of the same category, for example Top Menu Location and
Center in specimens 2 and 3, must be included. This simple set of rules has
helped narrow down the previously identified 163 initial specimens into
a smaller number to be used as valid specimens in the empirical studies.
Although the rules followed are simple, the screening of 249 categories of
more than 163 specimens was enormously demanding. With careful atten-
tion, 35 Web site specimens were finally concluded. The specimens were
coded numerically from 1 to 35. A snapshot of the specimens is shown in
Figure 8.4.
8.3.2 Establishment of Checklist

This section develops the Kansei checklist as one of the instruments to be
used in Kansei measurement and describes the stages of the establishment
of the checklist.
s
ci
an
Fr
r&
Figure 8.4
The 35 valid specimens.
o
yl
Ta
8.3.2.1 Checklist Development

A checklist is developed as one of the instruments used as measurement
tools. The checklist comprises Kansei words that are identified according to
the steps described in the following sections. The Kansei words are used as
the measure of strength of the Kansei, that is, the emotional responses that
subjects feel when looking at the Web site.
8.3.2.2 Synthesize Kansei Words

A set of Kansei words was selected based on frequency of appearance in Web
design guidebooks, Web sites, research papers, and journals. Additionally,
general Kansei words were added according to relevance in describing Web
sites. A total of 40 words were then finalized to be used in the experimen-
tal procedure. Some of the synthesized words were adorable, classic, creative,
elegant, lovely, masculine, and sophisticated.
8.3.2.3 Development of Kansei Checklist

The 40 Kansei words compiled in the earlier section were then organized
into a 5-point SD scale to form the Kansei checklist. This checklist will be
used as a measurement tool for investigating users’ Kansei.
8.3.3 Determination of Kansei Concept and Requirement

In determining the Kansei concept and design requirement in Web design,
two stages of activities were performed. The following subsections describe
the stages.
8.3.3.1 Kansei Measurement

The first stage describes the process of Kansei measurement that is performed
s
to evaluate visitors’ Kansei responses when interacting with Web sites. This
ci
stage consists of the core activities involved in engineering emotion in Web
an
site design. The instruments and equipments used in the Kansei measure-
ment procedure are the Kansei checklist, 120 test subjects, the screenshot of
the 35 valid specimens, a computer, and one large LCD screen.
Fr
1. Instrumentation. The instruments involved in the experimental pro-

cedure are as follows:
r&
a. The 35 valid specimens. The specimens were coded numeri-

cally from 1 to 35. A snapshot of all specimens can be found in
o
Figure 8.4.
yl
b. The Kansei checklist, which consists of 40 Kansei words orga-

nized in a 5-point SD scale. The order of Kansei words was
Ta
organized into five different arrangements to minimize bias in

the evaluation process. A full Kansei checklist can be found in
Appendix 5.7.
c. One hundred and twenty test subjects. Subjects for the empirical
study were employed with equal distribution of male and female
subjects of 30 people in four groups. A total of 120 undergraduate
students from
• Information Technology and Quantitative Science Faculty (IT)
• Architecture, Building, Planning, and Survey Faculty (AD)
• Business and Management Faculty (BM)
• Engineering Faculty (ER)
from the author’s university were recruited for the empirical
evaluation. All of them were in their 20s, Internet users, and
familiar with online shopping. The suitability of the employment
of young students as subjects in this research is supported by the
literature in information system studies that suggests students

and youngsters represent the majority of e-commerce consum-
ers (Saarenpää and Tiainen, 2005). Therefore, they are the best
consumer demographic group to be studied. On the other hand,
the population of subjects in this research was decided based
on the suggested number in KAE methodology. Although the
population of test subjects varies from a minimal number such
as five to more than a thousand in different KAE implementa-
tion, depending on objectives and measurement tools used, the
suggested number for this kind of consumer research is around
30–50 subjects (Nagamachi, 2003; Nagamachi and Lokman, 2010).
In the research, in an effort to explore differences in educational
background, a total of 120 students from four different academic
backgrounds were chosen. Additionally, to facilitate balance of gen-
s
der population, equal distribution of subject numbers—15 females
ci
and 15 males—was employed in each group.
an
2. Evaluation. Four Kansei evaluation sessions were held separately for
each group. During each session a briefing was given before the sub-
jects began their evaluation exercise. The 35 Web site specimens were
Fr
shown one by one on a large white screen to all subjects in a systematic

and controlled manner. Subjects were asked to rate their feelings on the
checklist according to the given scale. Subjects were given 3 minutes
r&
to rate their feelings toward each specimen. They were given a break
after the 15th Web site specimen, to refresh their minds. The order of
o
the checklist was also changed to eliminate bias. Each Kansei evalua-
tion session took approximately 2 hours to complete.
yl
Ta
8.3.3.2 Requirement Analysis

This stage is performed to conceptualize Kansei in Web site design and to
analyze design requirements. Multivariate analyses were performed to find
empirical evidence toward the goal of engineering emotion in Web design.
The outcomes of the analyses were used to propose a guideline to the design
of a Kansei Web site. We calculated the average Kansei evaluation value of
each sample obtained from all subjects from the experimental procedure.
These averaged data from the evaluation results were used in the calculation
of the multivariate analyses.
Factor analysis (FA) was used to analyze Kansei concept, and partial least
squares analysis (PLS) was used to analyze relationships of design and Kansei.
8.3.3.2.1 The Kansei Concept

The research performed FA to determine the concept of Kansei in Web
design. Table 8.4 shows the result of FA after varimax rotation. Varimax
rotation, which was originated by Kaiser (1958), is the most popular
Table 8.4
Factor Contribution
Cumulative
Factors Variance Contribution Contribution
Factor 1 16.09262 40.23% 40.23%
Factor 2 12.29421 30.74% 70.97%
Factor 3 3.427578 8.57% 79.54%
Factor 4 1.856272 4.64% 84.18%
Factor 5 1.810882 4.53% 88.70%
Factor 6 0.923415 2.31% 91.01%
Factor 7 0.370649 0.93% 91.94%
Factor 8 0.250962 0.63% 92.57%
rotation method that simplifies the interpretation of variables. In the
s
table, it is evident that the first factor explains 40.23% of the data and
ci
the second factor explains 30.74% of the data. Both factors represent the
an
majority of factor contributions. This shows that Factor 1 and Factor 2
have a dominant influence on Kansei words. The first two factors together
represent 70.97% of the variability, while three factors explain 79.54% of
Fr
the variability. Thus, the research considered including the third factor
to increase the proportion that represents most of the data. The propor-
r&
tion of variability explained by the fourth factor and the rest are minimal
(4.64% and less), and they probably can be ignored as they can be consid-
ered insignificant.
o
Table 8.5 shows factor loading results after varimax rotation. The table
yl
shows factor results in ascending order. The structure of Kansei words is

observable from the table. Variables that have a high score are perceived
Ta
as significant factors in Web site design. The research set approximately

0.7 as the reference score and cross-checked with the result of correlation
coefficient analysis (CCA) to draw conclusions. The fourth and fifth factors
were included since they can be regarded as having a clear image on the
Web site Kansei, even though the score is slightly lower. Henceforth, the con-
clusion from the results of FA was that the Kansei concepts of Web design
are structured by five factors: sophistication, elegant-beauty, simplicity, lightness,
and tidiness. These five factors altogether explain 88.70% of the total data.
The first factor, sophistication, consists of mystic, futuristic, masculine,
luxury, sophisticated, surreal, impressive, gorgeous, cool, and professional.
The second factor, elegant-beauty, consists of feminine, chic, beautiful, cute,
sexy, charming, adorable, and elegant. The third factor, simplicity, consists of
simple and plain. The fourth factor, lightness, consists of light. The fifth factor,
tidiness, consists of neat and natural.
As can be seen from the result, the first and second factors explain most
of the data; sophistication and elegant-beauty represent 70.97% of data. Thus,
these two factors are very important Kansei concepts. This indicates that all
Table 8.5
Ta
Factor Loading for Kansei Words
Variable Factor 1 Variable Factor 2 Variable Factor 3 Variable Factor 4 Variable Factor 5 Variable Factor 6
yl
Professional 0.805803 Pretty 0.689458 Sexy
o0.272922 Impressive 0.202914 Lively 0.201622 Adorable 0.046115
Cool 0.811333 Lovely 0.690027 Classic 0.275441 Masculine 0.211368 Lovely 0.207907 Creative 0.055077
Gorgeous 0.812754 Elegant 0.703414 Boring 0.308598 Adorable 0.21181 Beautiful 0.239447 Classic 0.063075
r&
Impressive 0.822734 Adorable 0.713039 Light 0.313839 Cool 0.284239 Relaxing 0.247409 Light 0.066423
Surreal 0.846445 Charming 0.763686 Neat 0.319281 Interesting 0.308149 Light 0.274663 Plain 0.066736
Kansei/Affective Engineering and Web Design
Sophisticated 0.848426 Sexy 0.787619 Calm 0.339163 Comfortable 0.320447 Calm 0.302602 Chic 0.131081
Fr
Luxury 0.878831 Cute 0.794058 Relaxing 0.348516 Lively 0.328505 Comfortable 0.328556 Old-fashioned 0.144072
Masculine 0.899118 Beautiful 0.816958 Natural 0.424887 Refreshing 0.390941 Refreshing 0.354777 Fun 0.222652
an
Futuristic 0.913165 Chic 0.93916 Plain 0.839005 Fun 0.49998 Natural 0.604973 Cute 0.276694
Mystic 0.941857 Feminine 0.948707 Simple 0.9241 Light 0.610599 Neat 0.738318 Childish 0.633318
ci
s
241
Web sites should have these two factors in order to produce optimum results.
Simplicity, lightness, and tidiness are also important but have weaker influ-
ence. Therefore, it is suggested that these factors be used as background/
supporting elements in Kansei Web site design.
8.3.3.2.2 Analyzing the Relationship of Design and Kansei

In analyzing design requirements, PLS analysis is used to link the Kansei
responses with design elements. In this analysis, there are three sets of
data used:
1. The dependant (objective) variables, y, that is, the 40 sets of Kansei

responses by 120 subjects.
2. The sample, s, that is, 35 Web sites.
3. The independent (explanatory) variables, x, the design elements
s
(design category).
ci
an
As mentioned earlier, the investigation of design elements has resulted
in 77 design items composed of 249 categories. For PLS analysis purposes,
all these categorical variations were converted into dummy variables. The
Fr
result of averaged data, for example, adorable data, was then appended
to the next column right after the last column of design category. This
research has 40 predictors and therefore the analysis was repeated 40
r&
times, exchanging the predictor into the last column. An instance from the
result for the Kansei score by design category can be found in Table 8.6.
o
The research analyzed the result of the PLS coefficient score to determine
relations between Kansei and design elements. In order to determine the
yl
influence of design elements to Kansei, PLS Range for each Kansei was cal-
Ta
culated. The calculation of Range enables the identification of design influ-

ence, good design and bad design. Range is calculated using maximum and
minimum value, where
Range = PLS Max − PLS Min
Mean of Range is calculated, where
n
1
Range =
n ∑ Range i
i=1
Each Kansei has means of Range, and if the mean value of a category is
larger than Range, the item is considered to have good influence in design.
Range for every category having value larger than Range implies a best fit
category that highly influences users’ Kansei in Web site design.
Table 8.6
PLS Coefficient Values
Kansei
Category Adorable Appealing Beautiful Boring Calm Charming
BodyBgColor-White −0.0365510 −0.03699 −0.01674 0.024457 −0.02534 −0.0355
BodyBgColor-Black 0.0065448 0.011992 −0.01374 −0.00265 0.028478 0.005989
Ta
BodyBgColor-DKBrown 0.0604354 0.067045 0.018645 −0.03459 0.034535 0.062087
BodyBgColor-LtBrown 0.0132480 0.011571 −0.00476 0.006006 0.017753 0.021147
BodyBgColor-Gray 0.0293157 0.036547 0.050832 −0.044 0.006308 0.033964
yl
BodyBgColor-LtBlue 0.0214068
o 0.004199 0.01559 −0.01155 −0.01207 0.005272
PageMenuShape-Curve 0.0005942 −0.01064 0.006802 −0.00788 −0.00697 −0.01168
PageMenuShape-Sharp −0.0122160 −0.00199 −0.02235 0.021834 0.006232 −0.00213
r&
PageMenuShape-Mix 0.0286988 0.024137 0.042154 −0.03895 −0.00256 0.026368
PageStyle-Frame 0.0340362 0.025436 0.027154 −0.03955 0.005176 0.005524
Kansei/Affective Engineering and Web Design
PageStyle-Table −0.0420300 −0.03508 −0.02236 0.04195 −0.01811 −0.01117

Fr
PageOrientation-BC −0.0447260 −0.04155 −0.03098 0.026269 −0.03299 −0.0298
PageOrientation-Content 0.0326846 0.037362 0.004381 0.011088 0.024659 0.034727
an
PageOrientation-Header −0.0546260 −0.05817 −0.03232 0.037364 −0.03451 −0.06225
PageOrientation-HF 0.0157878 0.015311 0.026275 −0.01806 0.00773 0.014688
ci
PageOrientation-H Split 0.0228924 0.02168 0.004236 −0.01846 s 0.016584 0.022154
PageOrientation-V Split 0.0189942 0.034075 0.030549 −0.02336 0.016852 0.028757
PageOrientation-Plain 0.0241808 0.019132 0.007982 −0.01443 0.015297 0.005532
DominantItem-Pict 0.0467300 0.048044 0.030602 −0.04358 0.014746 0.050762
DominantItem-Adv. −0.0296800 −0.03225 −0.01741 0.019399 −0.00744 −0.02078
DominantItem-Text −0.0561230 −0.04549 −0.02663 0.050166 −0.00284 −0.04293
243
Table 8.7
Design Influence in Kansei
Adorable Appealing
Design
Influence No. Design Element Range Design Element Range
1 Page Color 0.11488 Header Bg Color 0.12338
2 Product Display Style 0.10644 Face Expression 0.12216
3 Header Menu Bg Color 0.10612 Header Menu Bg Color 0.12077
4 Left Menu Font Color 0.10370 Product Display Style 0.10646
5 Header Bg Color 0.10218 Body Bg Color 0.10574
6 Face Expression 0.10024 Page Color 0.10091
7 Body Bg Color 0.10015 Left Menu Font Color 0.10085
8 Dominant Item 0.09980 Picture Style 0.09771
9 Header Font Size 0.09651 Page Orientation 0.09182
s
10 Main Text Existence 0.08813 Dominant Item 0.09141
ci
11 Main Bg Color 0.08587 Main Text Existence 0.08811
an
In every category, maximum value shows best fit value of design elements
that influence the Kansei, and minimum value shows the worst value of
Fr
design elements that influence the Kansei.

Table 8.7 shows an example of the result of the selected category for
which Range has value larger than Range, for each Kansei. The results are
r&
sorted in descending order to illustrate dominant design category for

each Kansei.
o
The result shows that in designing an adorable Web site, the designer must
yl
set priorities to design elements according to higher order of influence, such

as page color, product display style, header menu background color, left
Ta
menu font color, and so forth. On the other hand, in designing an appealing
Web site, the designer must set priorities to design elements according to
higher order of influence, such as header background color, face expression,
header menu background color, and so forth.
8.3.3.2.3 Proposing Kansei Web Design Guidelines

All the above analyses have enabled the research to propose Kansei Web
Design Guideline, a guideline to the design of Kansei Web sites. Results of
Kansei structure from FA are used to conceptualize Kansei, and results from
PLS scores are used to compose the design requirement. The design require-
ments included in the guidelines are the elements that have high influence
in eliciting each Kansei. Table 8.8 shows examples of the results of the estab-
lished guideline.
To effectively utilize the guideline, the audience and especially designers
are advised to select the best combination possible from a Kansei concept
that may consist of one or more Kansei elements. It should be noted that it
is important to blend designers’ creativity with the guideline to ensure the
Table 8.8
Customer Needs Direction Effect Measure (E)
Customer Needs Direction Supp Cer Eff
Simple and Advanced and Beauty 41 0.569 1.387
Simple and Advanced and Young 41 0.569 1.387
Advanced and Original 77 0.535 1.303
Handiness and Original and Young and Functional 50 0.521 1.269
success of the Kansei product (Nagamachi, 2008), in this case an e-commerce

Web site. To illustrate an example of the guideline, design elements for the
Kansei mystic should be interpreted as follows:
Body background color should be in black.
s
Body background style should be in texture of color tone.
ci
Page shape should be not specific.
Page orientation should be plain.
an
Dominant item should be picture, and so on.
Fr
In KAE, the success of a Kansei design product relies on the idea implied
from the result of Kansei evaluation, blended with technical expertise of the
designer. In the design process, the guideline should be referenced by and
r&
technical specifications must be provided by the expert.

o
8.3.4 Prototyping/Testing
yl
The chapter performed two stages of activities for the purpose of valida-
Ta
tion of the successful implementation of the Kansei Design Model in engi-

neering Kansei in Web design. The chapter developed several prototypes
using the proposed guideline, and then conducted Kansei evaluation to see
if there are any difference in Kansei responses after the implementation of
the guideline.
8.3.4.1 Prototyping
In the KAE approach, a designer will design a new product based on a Kansei
concept identified from FA results. In the case of this research, in performing
confirmatory study, the research attempts to design a Kansei Web site to be
used as a specimen. Although a combination of Kansei factors are ideal, in
the case of Web site design, due to the large number of design variables, it
is almost impossible to combine Kansei factors. For confirmatory purposes,
such attempt was conducted where five Kansei were selected and the indi-
vidual guideline then was used as a foundation for design specimens. The
five Kansei were selected from factors 1, 2, and 3 and referred to the pro-
Table 8.9
The Selected Kansei Guideline
Kansei
Design Element Cute Feminine Luxury Masculine Simple
Body Bg Color Light Blue Light Blue Black Black Dark Brown
Body Bg Style Texture Texture Color Tone Color Tone Picture
Page Shape Sharp Sharp Sharp N/S Sharp
Page Menu Shape Mix Mix Sharp Sharp Sharp
Page Style None None None None None
Page Orientation Footer Footer Vertical Split Header Content
Dominant Item Picture Picture Picture Picture N/S
Page Color Gray Pink Black Blue Brown
Page Size Small Small Small Medium Medium
s
Other Images? Animal Animal Animal Animal Kids
ci
Product Display Style Filmstrip Filmstrip Filmstrip Filmstrip Catalog
Product Try On Yes Yes Yes Yes Yes
an
Product View Angle Rear Rear Mix Side None
Artistic Menu? Yes Yes Yes Yes No
Fr
Empty Space? Less Less Less Less More

Discount Ad. Existence No No No Yes Yes
r&
posed guideline presented earlier. Table 8.9 shows an example of the selected
Kansei and guideline.
The guidelines were carefully followed in the development of prototypes
o
to be used as specimens in the confirmatory study. Figure 8.5 shows an

yl
example of the developed cute Web site to give an illustration of how the
Ta
guideline is used in the formation of the Web site. It should be noted that
the guideline is to be used to support the designer’s creativity by providing
design requirements to the anticipated Kansei.
Other than the above Web site, the research has developed four prototypes,
and altogether five prototypes were developed to be used as a specimen in
the experimental procedure. Figure 8.6 shows a snapshot of the developed
five specimens.
8.3.4.2 Testing
To test the success of the guideline implementation in the prototype, a com-
parative study was conducted using two sets of data:
1. The exploratory data—The data extracted from initial study described

in earlier sections.
2. The confirmatory data—The data obtained from Kansei evaluation
performed using the prototype.
Texture background Kids picture
Image of animal
Product try-on
s
ci
an
Fr
3D image Mix menu shape
Figure 8.5
r&
Example of design requirement for cute Web site.

o
yl
Ta
Figure 8.6
Confirmatory Web site specimen.
8.3.4.2.1 Extracting the Exploratory Data

Exploratory data are the data extracted from exploratory study results. In
the process of identifying which data to extract, the research performed the
following procedures:
1. Data coding to segregate all data from 15 good subjects

2. Random generator to select 5 specimens
3. Calculate averaged evaluation value from the 15 good subjects
In the process of extracting data, the Microsoft Excel random generator

was used to select the five specimens. Exploratory data were finalized by
filtering all the data using the generated sample ID number. Average data
were calculated using the filtered data and labeled as the exploratory data to
be used in the comparison procedure.
s
8.3.4.2.2 Obtaining Confirmatory Data
ci
an
To obtain confirmatory data, Kansei evaluation was performed with the
newly developed prototype. A checklist comprising five sets of the selected
Kansei was developed as a measurement tool to be used in the experiment.
Fr
Using the five prototypes, the Kansei checklist, and 15 good subjects, the
research performed a confirmatory Kansei measurement. The 15 good sub-
r&
jects were selected among good respondents from the initial study. One
Kansei evaluation session was conducted to measure Kansei responses
from all subjects. During the session a briefing was given before the subjects
o
began their evaluation exercise. The five Web site specimens were shown one
yl
by one on a large white screen to all subjects in a systematic and controlled

manner. Subjects were asked to rate their feelings on the checklist according
Ta
to the given scale. Subjects were given 3 minutes to rate their feelings toward
each specimen. The Kansei evaluation session took approximately 15 min-
utes to complete.
The average evaluation result obtained from the Kansei evaluation process
was calculated, and the data were labeled confirmatory data.
8.3.4.2.3 Comparison of the Exploratory and Confirmatory Data

To provide proof that the implementation of the proposed guideline is suc-
cessful in producing a Kansei Web site, the chapter performed PCA on both
exploratory data and confirmatory data. Comparison of the Kansei structure
formed by the two groups of data also helped confirm the validity of the
Kansei design model. Both purposes can be achieved when improvement in
the structure of Kansei is evident.
In performing the comparative study, the averaged data from the two
groups, exploratory data and confirmatory data, were combined. PCA was then
performed to investigate the structure of Kansei before and after the guide-
line implementation. Figure 8.7 illustrates the procedure. In the procedure,
Prototype Kansei specimen Exploratory specimen selected

according to guideline by random generator
Confirmatory Data Exploratory Data
Comparative Analysis
Figure 8.7
The testing method.
s
the confirmatory data were appended at the end of the exploratory data to
make it easier to recognize specimens in the investigation process. In the
ci
an
combination, exploratory specimens were coded from 1 to 5, and confirma-
tory specimens were coded from 6 to 10. The overall data of the 10 sam-
ples were analyzed using PCA to investigate relations between all samples
Fr
and Kansei.
Figure 8.8 shows PC vector for the comparison data. The vector plot shows
the implied Kansei structure by all specimens in two-dimensional spaces.
r&
It is evident from the vector plot that specimens 1, 2, 3, 4, and 5, which are
specimens from the exploratory study, are concentrated toward the center
o
yl
2.7
Masculine
Ta
1.4 Luxury
Simple
8
10
Component 2
0.0 3
45
2 7
1
Feminine
–1.4
6
Cute
–2.7
–2.7 –1.4 0.0 1.4 2.7
Component 1
Figure 8.8
PC vector for comparison data.
space. This indicates that the specimens have poor influence on Kansei in
comparison to the other groups. Specimens 6, 7, 8, 9, and 10, which are newly
developed Kansei Web site prototypes designed based on the guideline, are
very well spread all over the Kansei vector. This result provides evidence
that the newly designed Web site has a good fit to Kansei. Thus, it can be con-
cluded that the Kansei in Web site design has been improved and that the
Kansei Web Design Guideline that was referenced is valid and justified.
8.4 Summary
This chapter has explored the potential of KAE implementation in Web
s
design. Literature about product emotion has described the need to design
ci
products that capture the user’s attention by captivating emotional connec-
tivity with the interface of the product. Many products that were designed
an
geared to the user’s emotion have been produced successfully in the mar-
ket (Nagamachi, 2008). With this motivating factor, the chapter explored
Fr
the possibility of embedding users’ emotion in Web site design. When

addressing the diverse aspects of UX in Web design, KAE can be seen to
provide a systematic approach to discover the concept of user Kansei in
r&
Web design, and thus enable designers to strategize the design of a Web
site that caters to the insight of user emotions and feelings in its design by
using KAE, the emotion.
o
Until recently, the design requirement for Web sites has been focused on
yl
the aspect of cognitive functionality and usability (Backlund, 2001; Garret,

2003; Ivory and Hearst, 2001; Krug, 2000; Lederer et al., 1998; Marcus and
Ta
Gould, 2001; Nielsen, 2000). Although recent research has paid increasing
attention to the emotional aspect of Web site design (Kim et al., 2003; Li and
Zhang, 2005; Norman, 2002; Overbekee et al., 2004; Thielsch, 2005), research
is lacking in terms of determining the aspects of emotional design concept
and requirements. Consequently, paradigm designs still must include the
emotional aspect of Web site design according to designer’s interpretation
and inspiration. Unfortunately this does not take into account the user’s
implicit needs and emotions. Thus, this chapter has attempted to engineer
emotion in Web site design to fill in the gap of design requirements geared
to users’ emotional responses.
The chapter has provided evidence that KAE can be used to determine the
implicit needs and emotion of users and shape the concept of Kansei in Web
design. The chapter has also formulated design requirements to create the
desired concept of Kansei Web design. Users or customers will then have
a guide to objectively describe the concept of Web design that they view or
want, and the designer will have clues for designing such a Web site.
References
Backlund, J. (2001). Web Interfaces and Usability. Stockholm: Center for User Oriented
Design, ISSN 1403-0721.
Garrett, J. J. (2003). The Elements of User Experience. New York: New Riders.
Ishihara, I., Nishino, T., Matsubara, Y., Tsuchiya, T., Kanda, F., Inoue, K. (2005). Kansei
and Product Development, in M. Nagamachi (Ed.), Vol. 1. Tokyo: Kaibundo
Publishing.
Ivory, M .Y., Hearst, M. A. (2001). The State of the Art in Automating Usability
Evaluation of User Interfaces, ACM Computing Surveys 33, 4: 470–516.
Kaiser, H. F. (1958). The Varimax Criterion for Analytic Rotation in Factor Analysis,
Psychometrika, 23, 187–200.
Kim, J., Lee, J., Choe, D. (2003). Designing Emotionally Evocative Homepages: An
Empirical Study of the Quantitative Relations between Design Factors and
s
Emotional Dimensions. International Journal of Human-Computer Studies, 56,
ci
6: 899–940.
Krug, S. (2000). Don’t Make Me Think, New York: New Riders.
an
Lederer, A. L., Maupin, D. J., Sena, M. P., Zhuang, Y. (1998). The Role of Ease of Use,
Usefulness and Attitude in the Prediction of World Wide Web Usage. In S. P.
Robbins, Organizational Behavior, 8th edition. Upper Saddle River, NJ: Prentice
Fr
Hall, p. 168.
Li, N., Zhang, P. (2005). Towards E-Commerce Web sites Evaluation and Use: An
Affective Perspective. Post-ICIS ‘05 JAIS Theory Development Workshop, Las
r&
Vegas, NV.
Marcus, A., Gould, E.W. (2001). Cultural Dimensions and Global Web Design: What?
So What? Now What? White Paper, AM+A.
o
Nagamachi, M. (2003). The Story of Kansei Engineering, Tokyo: Japanese Standards

yl
Association.
Nagamachi, M. (2008). A successful statistical procedure on Kansei engineering prod-
Ta
uct, The 2nd European Conference on Affective Design and Kansei Engineering,
(Helsingborg). (CDROM).
Nagamachi, M., Lokman, A. M. (2010). Innovations in Kansei/affective engineering.
Industrial Innovation Series, A. B. Badiru (Ed.). (in press), Boca Raton, FL: CRC
Press.
Nielsen, J. (2000). Designing Web Usability: The Practice of Simplicity. New York: New
Riders Press.
Norman, D. A. (2002). Emotional Design: Attractive Things Work Better. Interactions:
New Visions of Human-Computer Interaction, ix, 36–42.
Overbeeke, K., Djajadiningrat, T., Hummels, C., Wensveen, S. (2004). Beauty in
Usability: Forget About Ease of Use! Human-Computer Interaction.
Saarenpää, T., Tiainen, T. (2005). Empirical Samples of IS Studies on eCommerce Consumers,
IRIS28. Norway.
Thielsch, M. T. (2005). Web-evaluation: Aesthetic Perception of Web sites, University of
Münster, Germany.
9
Kansei/Affective Engineering for the European
Fast-Moving Consumer Goods Industry
Cathy Barnes, Tom Childs, and Stephen Lillford
Contents
9.1 Introduction.................................................................................................254
s
9.1.1 Building a Successful Brand..........................................................254
ci
9.1.2 New Product Development in FMCG.......................................... 255
an
9.1.3 Why Does FMCG Need Kansei/Affective Engineering?.......... 257
9.2 Kansei/Affective Engineering Framework for FMCG Products......... 257
9.2.1 Stage 1: Receive Brief...................................................................... 259
Fr
9.2.2 Stage 2: Define Scope...................................................................... 259

9.2.2.1 Context............................................................................... 259
9.2.2.2 Evaluation Variables........................................................ 260
r&
9.2.2.3 Design Attribute Variables............................................. 261

9.2.3 Stage 3: Select Variables................................................................. 263
o
9.2.3.1 Evaluation Variables........................................................ 264

9.2.3.2 Design Attributes............................................................. 266
yl
9.2.4 Stage 4: Quantify Relationships.................................................... 268

Ta
9.2.4.1 The Bottled Drink Case Study....................................... 269

9.2.4.2 The Moisturizer Case Study........................................... 270
9.2.5 Stage 5: Optimize Product............................................................. 270
9.2.5.1 Target Users...................................................................... 270
9.2.5.2 Design Space..................................................................... 271
9.2.5.3 Design Guidelines............................................................ 272
9.3 Conclusion................................................................................................... 273
Acknowledgments............................................................................................... 273
References.............................................................................................................. 274
253
9.1 Introduction
During the last 15 years the number of fast-moving consumer goods (FMCG)
available in the average European supermarket has grown exponentially.
To be successful in this dynamic environment, companies must constantly
launch many new products and build successful brands.
To support the speed of innovation and brand creation, the industry sec-
tor has many existing methods and tools incorporated within its product
development processes. Thus, to be widely adopted in the FMCG arena,
Kansei/affective engineering must fit within the industry’s current high-
speed product development processes and demonstrate that it addresses
gaps in the existing suite of methods. Perhaps even more important, Kansei/
affective engineering must ensure it takes full account of the needs of the
s
product brand.
9.1.1 Building a Successful Brand

ci
an
One of the distinctive factors of the FMCG industry is the importance of
Fr
the brand. The brand was originally created to reassure the consumer of
a quality product by building awareness and developing positive associa-
tions. Today, the brand is of significant commercial importance and stretches
r&
far beyond having a recognizable logo on the product. Figure 9.1, based on
Brand Association
o
- Type Attributes
yl
Benefits
Ta
Brand Association
- Uniqueness
Brand Attitudes
Image Brand Association
- Strength
Brand Association
Brand
- Favorability
Knowledge
Brand
Recognition
Brand
Awareness
Brand
Recall
Figure 9.1
Dimensions of brand knowledge. (Adapted from Keller, K. L. 1993, Journal of Marketing 57
[January], 1–22.)
Kansei/Affective Engineering for the European Consumer Goods Industry 255
Keller (1993), shows the many different aspects that comprise the nebulous
“brand” concept.
9.1.2 New Product Development in FMCG

Typically most FMCG companies use a variant of the stage-gate process.
Figure 9.2 shows a reduced set of stages, from targeting a need to deploy-
ing in the market, based on a survey of companies as part of a European
Commission (EC)-funded coordination action, the ENGAGE (design for
emotion) project (Childs et al., 2006). The impression of a linear activity that
it gives is, however, misleading. Although failure to pass a gate (between
stages) will result in canceling the development, learning from a later stage
can cause an earlier stage to be revisited. The “learning by doing” view of
Figure 9.3 shows the stages more truly as activities that can be revisited at
s
Target Ideate Develop
Implement
ci Scale Up Deploy
an
& Verify
Create, Develop Prototype at Evaluate

Understand Test
Fr
explore and pre-production against

user and in the
and screen evaluate level needs and
market need market
concepts which to take launch
r&
Figure 9.2
Steps of a stage-gate development process.
o
yl
Ta
Possible ideas.
Create, explore,
screen (reduce)
concepts Concept
development Scale-up
and evaluation and evaluate
to select one against needs
and launch Test in
to become Implement the market
a product and verify. and deploy
Prototype at pre-
Need/target? production level
(Product brief
from consumer
insights and
company values)
Figure 9.3
Learning by doing in a product development funnel.
any time, although that is more common the earlier the stage of development.
Particularly at the start of a development the need is to quickly explore and
screen out a large number of possible ideas and to move on. New insights
during concept development and even during preproduction certainly can
lead to design changes. Feedbacks from the later to intermediate stages more
commonly relate to aspects such as product color or details of labels rather
than, for example, shape changes.
Feedback from the market to the need is usually considered to be part of
continuous improvement, signaling launch of a new development, rather than
modification to an existing project. One aspect of development that is not shown
by either Figure 9.2 or Figure 9.3 is who is involved. Different companies bring
together teams of marketeers, industrial designers, engineers, and sales staff
in different ways at different stages of the process. In some cases a multidisci-
plinary group is formed to take a product forward from need to deployment.
s
In others the product is passed from group to group as it develops. In all cases
ci
there is a need for good interdisciplinary communication tools.
Designing products and packaging for consumers is considered by most
an
businesses to be a “consumer journey,” which is a chronological map of the
consumer/brand interactions (Davis and Longoria, 2003). Figure 9.4 shows
Fr
that there are, in the main, three touchpoints on the journey, and lists some of
the ways by which the consumer experiences a branded product. However,
two of the touchpoints are particularly important to a consumer-centered
r&
design process such as Kansei/affective engineering. These are the Purchase

Experience and the Use Experience, often colloquially called the First Moment
of Truth and the Second Moment of Truth, after a widely reported Procter &
o
Gamble process. The Purchase Experience is particularly important because

yl
it occurs within the retail outlet. There the product needs to quickly and
effectively communicate its benefits. The messages here impact consumer
Ta
expectations, but primarily a good design for this touchpoint will certainly
î Advertising
Pre-Purchase î Direct mail/samples
Experience î Coupons and incentives
î Deals and promotions
î Packaging
Purchase î Point of sale display
Experience î Store and shelf placement
î Salesperson
î Product and pack performance

Use î Customer service
Experience î Newsletters
î Loyalty programs
Figure 9.4
Experiences on the consumer journey.
sell more products in the first instance. The consumer experience at the Use
Experience determines whether the product will be repurchased. A well-
designed product for this stage will build brand loyalty and ensure long-
term profitability.
9.1.3 W hy Does FMCG Need Kansei/Affective Engineering?

It is clear that good design that appeals to consumers throughout the prod-
uct’s life cycle is critical to FMCG. However, it is not just the product that
needs to be considered. The impact of the brand communication is critical,
as is the packaging that is used to transport and contain the product. Thus,
simple but effective tools are needed for companies to be able to quickly
understand whether products in development will satisfy consumers and
thus be successful.
s
Kansei/affective engineering fills a gap in the FMCG manager’s toolbox.
ci
The Kansei/affective engineering variant presented in this chapter can con-
sider all the touchpoints throughout the consumer journey and measure
an
quantitatively whether a concept matches consumer needs. It can also ensure
that the brand equity is incorporated within the design and can also test the
Fr
effectiveness of that communication, critical in this industry. This Kansei/

affective engineering is also of benefit over and above more traditional design
and market research techniques as it can be used to create a set of assessments
r&
from untrained consumers rather than from a trained panel of people. This
can both reduce the costs of testing and ensure a wider population sample can
be used in the research. A Kansei/affective engineering system, as reported
o
earlier in this book, is rarely needed in the FMCG industry, as the test samples
yl
used in the research are not the final shelf-ready design but rather will form a
set of fundamental guidelines for the solution. Finally, Kansei/affective engi-
Ta
neering provides a unique tool by which all the disciplines involved in prod-
uct development can be brought together in a team to use a single process.
9.2 Kansei/Affective Engineering Framework

for FMCG Products
The world of FMCG is very fast moving and consumer focused and so
requires specific tools to ensure these are considered throughout the devel-
opment process. Figure 9.5 shows the five-stage framework of our Kansei/
affective engineering methodology for FMCG products. It has developed
from experience gained through some 30 sponsored projects, most of which
have been carried out under confidentiality agreements. It is described step
by step in this section and illustrated by elements from three case studies
that we can talk about. One is based upon a project completed with a drinks
1. Receive brief
1.RECEIVE brief
2. Define
DEFINEscope
scope
3. Select
SELECT variables
variables
4. Quantify
QUANTIFY relations
relations
s
5. Optimize product
ci
an
Figure 9.5
Fr
Kansei engineering framework for FMCG products.
company that had introduced a new bottled product to the UK market. It

r&
was an extension to a well-developed brand. The aim of the study was to

explore how well the launched product communicated the brand’s values
o
as well as the product’s intended unique benefits, and to recommend any

design changes that would improve user perceptions. The second relates to
yl
a project on a personal skin-care product, namely, the container for a body

Ta
moisturizer. The project was less constrained than the first. Although the
product too was to be a branded one, it had not at the time been released on
the market, although concept work had started. Guidelines relating to shape,
color, and texture (for touch feeling) were requested. The third relates to con-
fectionery product packaging. It was the least constrained of all. Again, it
concerned a branded product, but only a need had been identified. The com-
pany was open to all ideas about what the product’s form should be. These
three examples can be seen to fit into different areas of the development fun-
nel (Figure 9.3).
The methodology has five stages to support the designer:
Stage 1: Receive brief

Stage 2: Define scope
Stage 3: Select variables
Stage 4: Quantify relationships
Stage 5: Optimize product
9.2.1 Stage 1: Receive Brief

The process starts from a high-level design brief concerning the need to be
met, including target audience and imposed brand guidelines. In the case of
the bottled drink project, the brief was for a drink for young males who enjoy
a short social drink with close friends and prefer premium alcoholic drinks.
For the confectionery project, the brief was to create packaging to support
sharable product moments, and a target group was specified. The moisturizer
project was positioned slightly differently. Its brief was to explore what attri-
butes of the container were most important to women who were willing to pay
more for products that were good for their skin and who felt neutrally about
or were dissatisfied with their existing lotion. Out of this, the guidelines for
shape, color, and texture, referred to in this section’s introduction, emerged.
s
9.2.2 Stage 2: Define Scope
ci
The next process stage is to determine exactly what is to be the scope of the
an
investigation, by defining its context and generating the experimental vari-
ables. It could be argued that defining the context is a detail of the first stage.
In this respect the division into stages is artificial. A continuous development
Fr
is involved. Generating the experimental variables is, however, certainly sep-

arate from the first stage. The experimental variables are the independent
and dependent variables of product attributes and evaluation descriptors
r&
that are the common aspects of Kansei/affective engineering activities. At

this stage the object is exhaustively to generate variables, many more than
o
can be followed up in detail. The purpose is as completely as possible to con-

sider the design space. Reduction of variables occurs in Stage 3.
yl
Ta
9.2.2.1 Context
First, which touchpoint is to be investigated needs to be defined. It is pos-
sible to study more than one touchpoint in a study, but this requires separate
experiments to be run for each. Consumer interactions with products are
highly context specific. The most relevant location and environmental condi-
tions must be specified along with determining the exact type of interaction
with the product.
Taking the bottled drinks product as an example, many aspects con-
tribute to the whole experience of it: the flavor, the bottle design, and
the advertising. The flavor is developed through sensory exercises with
trained panels using defined descriptors like sweet, crisp, and so forth.
Advertising is outside the scope of this chapter. The case study was
concerned with the bottle design. To select an appropriate design, what
qualities are communicated through different bottle designs, what attri-
butes of bottles contribute to communicating different qualities, and what
might be done to communicate the ideal qualities more effectively must
be clarified. These aspects require the responses of the specific target

users to be understood rather than those of a skilled, trained panel. It is
here that the sponsoring company found Kansei/affective engineering
to give it insights it did not already have. Several points of interaction
between bottle and user could have been explored: viewing the bottle on
the bar shelf (the product’s social context), holding the bottle, opening
the bottle, drinking from the bottle, and disposing of it. The scope of the
case study was purely on-the-shelf appearance, the purchase experience
of Figure 9.4.
9.2.2.2 Evaluation Variables

As with all versions of Kansei/affective engineering, a set of variables needs
to be defined that will become the criteria against which all the product con-
s
cepts will be evaluated by consumers. Often they are word variables, usually
ci
adjectives, but they can be phrases. They can also be nonverbal. Hereafter,
the terms evaluations, adjectives, or just words will be used interchangeably as
an
seems best to fit the context. Words must be included that describe the full
range of desirable product qualities.
Fr
Target users’ words are collected from a series of discussion exercises in

which a recruited target group is asked to interact normally with a set of
relevant products and related materials (e.g., images). Their preferences are
r&
interrogated using structured exercises such as
1. Articulation of product preferences and reasons

o
2. Sorting products into groups by both similarities and differences,

yl
similar to Kelly’s repertory grid method (Van Kleef et al., 2005)

Ta
3. Describing ideal or ultimate products
Figure 9.6 shows views of a focus group discussion, both with products as
screen images and a wide range of collected actual products, taken during
the body moisturizer case study. Words also need to be chosen that accurately
describe the essence or personality of the brand. Without them, judgments relat-
ing to the brand image will be lost from the experiment. We have found that
words supplied directly from the brand owners’ resources of brand and product
descriptors are often not suitable. Some problems that can arise are as follows:
• Users may not understand a word, resulting in an incorrect

response.
• The words may have similar meanings to other words; this can cause
results to incorrectly weight a particular response.
• Users can be led or confused by unfamiliar words or by words that
may be difficult to consider in the required context (e.g., Is this wrist-
watch oppressive?).
s
ci
an
Fr
r&
Figure 9.6
o
A focus group session.

yl
We have developed a way to avoid such problems. Words supplied by the

Ta
brand owner are used as seeds to grow more words that are their synonyms,
using the British National Corpus of language (Delin et al., 2007). At Stage 3
(Section 9.2.3) it is decided which of them, with the target user–generated
words, to use. More detail of the Corpus tool is presented in Section 9.2.3.
Returning to the bottled drinks product example, small focus groups of
5 to 10 target users were held. Structured exercises were completed in the
areas of articulating reasons for preferences (e.g., “Describe your preferred
drink and why you like it.”) and stimulating insights using previously devel-
oped mood boards (see Design Attribute Variables, below). Brand and prod-
uct benefit adjectives were collected and expanded using the Corpus tool. A
selection of the many hundreds of resulting words is shown in Table 9.1.
9.2.2.3 Design Attribute Variables

Similarly to the collection of evaluation variables, a large number of products/
product concepts is initially explored from which to generate design attri-
bute variables that might be expected to contribute to the overall affective
Table 9.1
Some of the Valuation Words from the Bottled Drink Case Study
Product and Flavor Qualities Brand Essence
Lime Crisp Refreshing Masculine Modern Urban
Citrus Crisp Arousing Angular Advanced Bustling
Fresh Dry Energizing Bold Contemporary City
Lively Edge Exhilarating Definite Current Dark
Natural Firm Invigorating Heroic Influential Downtown
Piquant Fresh Keen Intense Hip Ghetto
Juicy Piquant Lift Macho Latest Industrial
Zesty Ripe Refreshing Manly Modern Metropolitan
… Sharp Regain Masculine Now Neon
… Regenerate Mature Popular Suburban
Renew Muscular Powerful Urban
Rejuvenating Potent Sophisticated …
s
Reviving Powerful Successful
ci
Rousing Robust Trendy
Stimulating Strong …
an
Tonic Vigorous
… …
Fr
quality of the product. The generation of sample products from which to

extract affective attributes by experiment is straightforward when a large
r&
range of products similar to the target product already exists. Sources are
shops, catalogs, magazine photographs, and company (including competitor
company) Web sites.
o
Figure 9.6 has shown an example of collected products that were used
yl
to stimulate the generation of evaluation variables in the moisturizer case

study. These were also used to analyze the typical shapes of moisturizer
Ta
bottles. From these, images of 48 shape samples were created: jars, cylin-
ders, flasks, downward and upward tapering, waisted, oval, teardrop, and
tubular forms. Some were later made as three-dimensional objects by rapid
prototyping. They are shown in Figure 9.7. In addition, 72 color patches were
generated for color tests. Forty-nine plastic, fabric, paper and board, metal,
and miscellaneous samples were collected for tactile tests.
The position is more complicated the more one has to rely on the genera-
tion of novel concepts as part of the project, that is, for radical new products.
The position is most problematic of all when trying to develop new product
attributes to relate to brand qualities, such as heroic. Such descriptors seem to
be too subjective. Our investigations suggest that transferring brand quality
traits to user perceptions of products is particularly difficult.
To overcome this we have developed a new technique based upon mood
boards (McDonagh et al., 2002), which has the added advantage that it is
aligned with existing practice. Mood boards are common design tools cre-
ated by a multidisciplinary team and use images, colors, forms, and textures
s
ci
an
Fr
r&
Figure 9.7
o
A selection of rapid prototyped moisturizer container shapes.

yl
to communicate the feel of a conceptual product to others. A set of images

Ta
is chosen by the design team to cue the brand qualities. If appropriate, these
are checked with consumers to ensure suitable images are used. These
images are then used as stimuli in a multidisciplinary brainstorming study
to translate them to product attributes. Figure 9.8 is a montage sample from
the confectionery product case study. Typically, many tens of concepts are
generated, and these are sorted to reveal around 20 diverse concepts. The
criterion for selection is team consensus. They would be added to products
collected more straightforwardly from the existing market.
9.2.3 Stage 3: Select Variables

This stage determines which adjectives and product concepts are used in the
main experiment. Kansei engineering typically uses a large number of both
adjectives and concepts. This makes for a very long and costly experiment,
which does not fit well into current FMCG product development processes. It is
essential to find robust ways to reduce the number of adjectives and concepts.
Sharable product moments Packaging concepts

Exciting Friendly
Generous Family fun Togetherness

a treat
Informal
Surprise
Figure 9.8
Mood boards and process of translation to concepts.
9.2.3.1 Evaluation Variables
s
The outcome from Stage 2 is typically several hundred collected adjectives,
ci
from both the target users and brand owners, describing perceived product
qualities. The total from both groups must be reduced to an overall total of
an
around 15 to 20, with some balance kept between them. We have developed
two procedures to support the reduction process. One of the two is a set of
Fr
rules involving a simple set of linguistic guidelines. Table 9.2 lists a selection
of them. Words from both groups are assessed against these rules to deter-
mine whether they will be easily understood in the context of the experi-
r&
ment. Unsuitable words are eliminated from future consideration.

The other procedure relates specifically to the brand-generated words.
Table 9.3 was generated, based on the moisturizer study, to help describe the
o
yl
Table 9.2
Ta
A Selection of Five Rules from a Total of Almost 20 for Rejecting Evaluations

Rule 1 Remove evaluations While “friendly,” for example, may be metaphorically
that are not plausibly extended to apply to inanimates, words such as
related to objects “enthusiastic” and “unbiased” may not be.
Rule 2 Remove evaluations For example “good” and “nice” do not describe products but
that describe purely people’s feelings about them. Feelings of like and dislike
evaluative reactions do not inform us about specific product-related reactions.
Rule 3 Remove ambiguous For example “clear” might be a good word to refer to either
evaluations a concrete or an abstract quality of products that promote
cleanliness (i.e., clear skin) or that are themselves without
color but it is also frequently used to describe the clarity
of an argument or conclusion.
Rule 4 Remove non-gradable For example “unique” is not gradable, something cannot
evaluations be “very unique” or “slightly unique.” So using it for a
semantic differential questionnaire could be ineffective.
Rule 5 Remove evaluations Remove adjectives that relate to untested senses or that
that are out of relate to a prolonged experience with the object rather
context with the than the controlled conditions of the Kansei study. For
research example, “satisfying” for an appearance.
Table 9.3
Translating Brand Words to User Words with the Corpus Tool
Synonyms from Corpus Tool
Tender
Conventional
Fun
Luxurious
Showy
Everyday
Slender
Cozy
Bold
Friendly
Simple
Traditional
Natural
Romantic
Advanced
Amusing
Spiritual
Casual
Ta
Love √ √ √ √ √ √ √ √ 8
Delight √ √ √ √ 4
yl
Happy √ o √ √ 3
Comfortable √ √ √ √ √ √ √ √ 7
Believable √ √ √ √ 4
r&
… √ √ √ √ √ √ √ √ 8
Brand Descriptors
… √ √ √ √ √ √ 6
√ √ √ 3
Fr
Skin kindness √ √ √ 3
Seed Words
an
Smoothness √ √ √ √ 4
Balanced √ √ √ √ √ √ 6
Moisturize √ 2
ci
Clarifying √ √
s √ 3
Number of synonyms the seed word relates to
… √ √ √ √ √ √ √ 7
Product Descriptors
… √ √ √ √ 4
Kansei/Affective Engineering for the European Consumer Goods Industry
6 6 5 5 5 5 5 4 4 4 4 4 3 3 3 2 2 2
Number of seed words synonyms relate to
265
Table 9.4
The Final Word Selection for the Bottled Drink Study
Intense Powerful Sophisticated Crisp Lively
Rough Refreshing Masculine Urban Modern
procedure. Its left-hand column contains the seed words (Section 9.2.2) from
the brand owner. Its top row holds some of the synonyms generated from
the Corpus tool introduced in Section 9.2.2. These words are placed from
left to right in the order of their importance as judged by how many of the
seed words they relate to. For example, the word tender relates to six of
the seed words, as indicated by the ticks in the table elements and totaled
in the bottom row. How many times the seed words are related to the syn-
onyms is also totaled along the rows of the matrix and entered in the right-
s
hand column. Synonyms are selected for the main experiment on the basis
ci
of their relevance (lower row scores) and inclusiveness (ensuring that seed
words remain represented), up to the maximum number that is judged
an
appropriate for the experiment considering that target user words must also
be chosen. Normally they would be found from among the 10 top ranked in
Fr
terms of their relevance.

Usually there are too many target user words remaining after reduction by
the linguistic rules. These may be reduced by thinning clusters, in the same
r&
way as can be used for reducing samples, as considered next under Design
Attributes. If it is found impossible to reduce the number of words to the
15 to 20 range, it is an indication that the planned experiment is too ambi-
o
tious. Table 9.4 has the final words used in the bottled drinks study.
yl
Ta
9.2.3.2 Design Attributes

The last aspect of the experiment that needs to be defined is which design
attributes should be considered and how to vary them appropriately. We use
two methods for making these choices:
9.2.3.2.1 Semantic Mapping
This method is suitable when it is easy to define just two key evaluations
for a product. These are chosen as the axes of a two-dimensional map that
physically is a large horizontal board. A small number of consumers, 5 to
10, place the large number of product concepts on the map. Figure 9.9 shows
a semantic map being created for the moisturizer shape experiment. Each
numbered counter represents one sample, typically the counter’s cluster.
Analyzing the clustering enables the key design attributes to be identified.
Any suitable technique may be used (Bech-Larsen and Nielsen, 1999); how-
ever, the Repertory Grid technique has yielded good results. Sample reduc-
tion is also based on thinning out the clusters.
s
Figure 9.9
ci
an
A semantic mapping experiment in progress.
9.2.3.2.2 Pilot Semantic Differential Survey

Fr
This is used when it is not possible to identify just two key evaluations to test.
A small number of users, around 10, are asked to rate many concepts against
r&
a small number of key evaluations on a 7-point Likert scale. Figure 9.10 shows
the results of a principal component analysis for the same samples as were
used for the semantic mapping, Figure 9.9. The significant design attributes
o
of these concepts can then be extracted using any suitable technique, as with
yl
the semantic mapping method.

Ta
6.0
4.0
2.0
Smooth
0.0
–6.0 –4.0 –2.0 0.0 2.0 4.0 6.0
–2.0
–4.0
Relaxing
Figure 9.10
Outcome of the moisturizer pilot semantic differential test, with 48 samples in place.
Table 9.5
The Factorial Design for the Bottled Drink Study
Design Body Neck Surface Body
Attribute Shoulder Shape Length Color Detail Width
Embossed
Straight
Attribute
Angled
Round
Brown
Curvy
Other
None
Short
None
Clear
Wide
Long
Style
Thin
1 1 0 0 1 0 0 1 1 0 0 1 0 1 0
2 1 0 0 0 1 0 1 1 0 0 1 0 1 0
3 1 0 0 1 0 1 0 1 0 0 1 0 0 1
4 1 0 0 1 0 0 1 0 1 0 1 1 1 0
5 1 0 0 0 1 0 1 1 0 1 0 0 1 0
6 1 0 0 1 0 0 1 1 0 0 1 1 1 0
s
7 0 0 1 1 0 0 1 1 0 0 1 0 1 0
ci
Sample
8 1 0 0 1 0 0 1 0 1 1 0 1 0 1
an
9 1 0 0 0 1 0 1 0 1 1 0 1 1 0
10 1 0 0 1 0 0 1 1 0 0 1 0 0 1
11 0 1 0 0 1 0 1 0 1 1 0 0 1 0
Fr
12 0 1 0 1 0 1 0 1 0 1 0 0 1 0
13 0 0 1 1 0 0 1 1 0 0 1 0 1 0
r&
14 0 1 0 1 0 0 1 0 1 0 1 0 1 0
15 0 0 1 0 1 0 1 1 0 0 1 0 0 1
o
After sample reduction, by thinning clusters and/or ignoring areas of the

yl
design space that might not be of interest, a factorial experimental design is

constructed to test the affect of the selected attributes. Table 9.5 shows the
Ta
bottled drink study design. Fifteen bottle samples were required to repre-
sent the total of 14 attributes (three shoulder types, two body shapes, two
neck lengths, two colors, three surface details, and two body widths) to be
taken forward to Stage 4. Among the samples there were at least two cases
of each attribute style.
9.2.4 Stage 4: Quantify Relationships

This stage comprises a standard semantic differential survey with some sup-
plementary demographic and overall liking/preference questions that allow
for additional correlations and consumer groupings. Samples are rated in
terms of the adjectives on a 7-point scale, using a semantic questionnaire.
The order of the adjectives is randomized in the questionnaire to reduce con-
ditioned responses and any systematic effects of participants’ fatigue. An
overall liking score is also collected on a separate 7-point scale.
The aggregated data from a sample of users representing the target market
are used to identify optimal designs, but before results are interpreted the
response distributions are examined to judge the level of consensus. If an

evaluation result from a group of users shows a double peak or random dis-
tribution, then it is removed and analyzed separately. This is because these
distributions could be an indication of disagreement in the meaning/use of
the evaluation or that it is not suitable for relation to the concepts (despite the
various linguistic tests and rules that have been applied to avoid this).
Outlier users are also identified from their responses: These are users who
have scored outside two standard deviations from the aggregated response.
Their responses are removed too and analyzed separately.
The cleaned data set is used to display the profile of each of the concepts
against all of the adjectives. To reduce the complexity of the results, principal
component analysis is used to identify which adjectives have been perceived
similarly in the exercise. Those with similar meanings are given a new cluster
heading, by making a judgment about the common meaning among them.
s
The concept set can then be displayed in the principal component space to
ci
make easy comparisons.
As aggregate scores have been used for the points in the profiles and PCA,
an
a Kruskal–Wallis test (Kruskal and Wallis, 1952) for significant differences
is carried out to see if differences between scores of different concepts are
Fr
important or whether there are overlaps between users. Multiple regression

analysis is performed as described in earlier chapters of this book to identify
and rank which attributes are contributing to which affect.
r&
9.2.4.1 The Bottled Drink Case Study

o
Thirty-seven users matching the target consumer profile took part in a

yl
semantic differential survey to rate the 15 bottles against the 10 evaluations.

The data were collated and outliers were removed. In this case two users
Ta
were removed. A principal component analysis of the data showed three

significant independent components. However, the first two accounted for
more than 77% of the variance so they will be focused on here. Based on
the adjectives’ loading on each component, these were called flavor intensity
and masculine strength. Figure 9.11a shows the type of bottle that made up
the study. Figure 9.11b shows how the 15 bottles of the study loaded in the
PC1/PC2 principal component space. It is seen that bottles 15 and 11 com-
municated flavor intensity most; 3 and 10 communicated masculine strength
most; and 2, 5, and 8 communicate both qualities to some extent.
There were several bottles that did not have significantly different adjec-
tive scores when considering the raw rather than aggregated data. These
have been circled in the principal component space to show that, although
they locate differently, this is not significant.
A regression analysis between the attributes and the adjectives aligned
with the principal components showed that the design of the shoulder
shape was the most important attribute for communicating flavor intensity
and that angled shoulders scored highest. Shoulder shape was also the most
11
15
14
9
4 Optimal
space
7 8
2
PC1, ‘Flavor Intensity’
5
6
12
10
13
1
s
3
ci
an
–5.0 –2.5 0.0 2.5 5.0 7.5
PC2 ‘Masculine Strength’
Fr
(a) (b)
Figure 9.11
r&
(a) A standard drink bottle product, and (b) a range organized by Kansei engineering methods.
important attribute for communicating masculine strength; however, round

o
shoulders scored highest.

yl
9.2.4.2 The Moisturizer Case Study

Ta
Sixty-five women took part in each of three semantic differential experi-

ments, one each for the shape, color, and texture aspects of the design attri-
butes. Figure 9.12a is a partial composite view of the loadings of the samples
in the PC space. Further detail is in Childs et al. (2006).
9.2.5 Stage 5: Optimize Product

9.2.5.1 Target Users
Cluster analysis can be used on the overall liking scores to determine if the
preferences of the participants in the experiment were homogenous or not.
For example, the dendrogram of Figure 9.13 shows that in the bottled drinks
case study there were three distinct participant groups within the sample.
There are two large groups: one of 25 users and the other of 8 users, and one
small group of two users. The scores showed that the group of 8 liked wide
bottles whereas the group of 25 did not. The other two users liked both. This
result suggests not a single but two possible optimal design solutions.
12
Oval
Fresh, Clean
6
8 Al. strip
PET 5P2 4
5GY4
4 2
Taper
PC2
Satin 0
5YR2
0
PC2
Tracing Velvet –2
paper –4
5Y4
–4
–10 –5 0 5 10 15 –6 –4 –2 0 2 4
PC1 Calm, Relaxing, Gentle PC1
(a) (b)
Figure 9.12
s
(a) Composite result of moisturizer shape, color, and texture experiments; (b) combined valida-
ci
tion experiment. an
Dendrogram
P23 cut off
P16
P34
Fr
P20
8 users P2
P15
P32
P27
2 users P6
r&
P33
P8
P36
P19
P37
P9
P1
o
P17
P10
P28
yl
P31
P22
25 users P35
Ta
P7
P25
P24
P21
P4
P26
P5
P30
P3
P12
P29
P13
P11
1.2 1.0 0.8 0.6 0.4 0.2 0.0 –0.2

Similarity
Figure 9.13
A dendrogram from the bottled drinks case study.
9.2.5.2 Design Space

To correctly interpret how the principal component (PC) space can be used
as a visual design guide, the construction of the space must be checked care-
fully to determine whether evaluations are correlating negatively or posi-
tively with the space’s axes. If all evaluations correlate positively and all are
equally desirable to the project owner, then those concepts located highly on
all the relevant PCs should be recommended. If there is no concept in this
location, then a new design could be proposed that is developed from com-
paring concepts that locate highly on individual components. Integrating
consideration of liking score with PC space provides an insight into which
qualities relate highly with user preference. Concept liking scores can be
overlaid or correlated with PCs but should not be included in the PC analysis
itself as they have a different type of semantic meaning.
Continuing with the example from the bottled drink study, Figure 9.11,
supposing the aim was to generate a bottle to support both intense flavor
and strong masculinity, the optimal concept location in the PC space is indi-
cated as in the top right corner of Figure 9.11b. The location of the existing
bottles shows that the wide clear bottle (8) is the most suitable from the set,
with the initial concept bottle (5) second. From the locations of the set and
s
visual differences between the bottles we could speculate that change in the
ci
initial concept glass color and width might increase affective communica-
tion of the product.
an
Fr
9.2.5.3 Design Guidelines

Through a combination of both PC multivariate and regression analyses,
design guidelines can be created. Many cases will require the designer to
r&
make compromises between the optimal concepts for each PC to achieve a

best overall design. The regression analysis gives an order of attribute sig-
o
nificance to inform this decision making or can provide guidance for further
attributes to test that have not been included.
yl
Again from the bottled drink case study, the regression results that show
Ta
different shoulder designs are responsible for transmitting flavor intensity

and masculine strength explain why there is not a bottle in the top right-hand
corner of the PC space. Therefore to increase communication of both PCs
beyond the potential of the variables tested requires a new design of shoul-
der shape or a new attribute to be tested. From the range of attributes tested
we are able to speculate that the shape of bottle 5 communicates the com-
bined qualities fairly well, and a possible alternative would be a wide, clean
bottle (based on bottle 8).
In the moisturizer case study, a verification experiment was carried out,
combining the shapes and colors expected to create a most clean and gentle
effect. Figure 9.12b shows the result. In this case the experiment’s samples
were simpler than the finally launched product, inset in the figure. However,
its oval and downward tapering shapes and heather coloring (code 5P2 in
Figure 9.12a) are reflected in the final product. It helped confirm the product
development direction.
9.3 Conclusion
Within the FMCG industry the Kansei/affective engineering methods that
are the subject of this chapter are widely seen by new product development
teams to offer a sufficient number of benefits to make them useful. They are
seen as not standing alone but fitting in with and filling gaps between other
new product development tools. In particular, they are not used to provide
complete design solutions but to define key parameters of a solution to be
fleshed out in more detail by other means. In an industry that frequently
relies on expert panels for opinions, they also provide a way to involve
untrained users (closer to the general consumer) in the development process.
They also generate rational reasons for decision making that all the different
disciplines involved in new product development can respond to. This com-
s
munication role lasts after a product’s launch, through archival documenta-
ci
tion that can be referenced for future developments.
However, Kansei/affective engineering is not universally accepted. It can
an
generate confusion and skepticism. It is perceived by some designers as con-
flicting with or restricting their own skills and intuition. It is also perceived
Fr
by some designers and innovation managers as constrictive to the design

process itself. By bringing untrained users into the development process,
seen by supporters as one of its advantages, it certainly results in decision
r&
making relying less on the flair of individuals. The more Kansei success
stories can be told, for example, through books such as this, the more the
negative perceptions may be countered. However, what may be a best way to
o
balance individual flair and group assessment within a structured process is

yl
an area that could benefit from further development.

Ta
Acknowledgment
The work reported in this chapter includes work carried out as part of
Knowledge Transfer Partnership No 6179, to “apply Kansei Engineering to
the European packaging industry” in conjunction with Faraday Packaging
Partnership and PIRA International, Ltd. The authors of this chapter would
also like to thank all the participating company members of the Faraday
Packaging Partnership who had significant input into the definition of this
process. Special thanks must go to the client companies who provided the
case studies reported here.
References
Bech-Larsen, T., and Nielsen, N. A. (1999). A comparison of five elicitation techniques
for elicitation of attributes of low involvement products. Journal of Economic
Psychology, 20, 315–341.
Childs, T. H. C., Agouridas, V., Barnes, C. J., and Henson, B. (2006). Controlled appeal
product design: A life cycle role for affective (Kansei) engineering. In Proc. 13th
Int. Conf. Life Cycle Engineering, Leuven, Belgium, pp. 537–542.
Davis, S., and Longoria, T. (2003). Harmonising your touchpoints. Brand Packaging,
Jan/Feb.
Delin, J., Sharoff, S., Lillford, S., and Barnes, C. (2007). Linguistic support for concept
selection decisions. AI EDAM 21, 123–135.
Keller, K. L. (1993). Conceptualizing, measuring, and managing consumer based
brand equity. Journal of Marketing 57 (January), 1–22.
s
Kruskal, W. H., and Wallis, W. A. (1952). Use of ranks in one-criterion variance analy-
ci
sis. Journal of the American Statistical Association 47, 583–621.
McDonagh, D., Bruseberg, A., and Haslam, C. (2002). Visual product evaluation:
an
exploring users’ emotional relationships with products. Applied Ergonomics, 33,
231–240.
Van Kleef, E., van Trijp, H. C. M., and Luning, P. (2005). Consumer research in the
Fr
early stages of new product development: A critical review of methods and

techniques. Food Quality and Preference, 16, 181–201.
o r&
yl
Ta
11
Kansei, Quality, and Quality
Function Deployment
Ricardo Hirata Okamoto
Contents
11.1 Introduction................................................................................................. 295
s
11.2 New Product Development Phases and Tools........................................ 296
ci
11.3 Kansei/Affective Engineering.................................................................. 298
an
11.4 Quality Function Deployment.................................................................. 302
11.5 QFD and Kansei..........................................................................................305
11.6 Conclusions.................................................................................................. 307
Fr
References..............................................................................................................308
r&
11.1 Introduction
o
Because of the drastic change in customer needs, the trend for shorter life
yl
cycles is increasing in organizations in terms of the importance of finding

product development methods and technologies that can reduce devel-
Ta
opment cycle time and cost. This influences a direction to reduce the gap
between true needs and expectations of the customers versus the developed
product itself. One key indicator to improve is the success rate of newly devel-
oped products in the market (sales and profits). In the comparative perfor-
mance assessment study regarding new products (PDMA Foundation 2004),
the “best” companies have a success rate of 75.5%, while the “rest” of the
companies average only a 53.8% success rate for newly introduced products.
The rate of ideas that become successes in the market is 25% (1 out of 4) in the
“best” companies and of 11% (1 out of 9.2) in the “rest” of the companies. The
current challenges are related to improving the product development pro-
cesses and reducing the gap between the final products and services versus
the market true needs.
Throughout the past 70 years, the quality standards have reached such a
high and competitive level that the product or service differentiation based
on quality, delivery times, quality in design, production efficiency, or costs
is not enough in current competitive markets. Companies and institutions
295
have developed the ability to translate functional requirements and customer

usability needs, and as a consequence, they have developed their capability
to design and construct the corresponding products, services, and environ-
ments that meet these requirements. But many of these product attributes
are mandatory and are already required as well as expected by the custom-
ers who do not feel fully satisfied. We are evolving from the satisfaction of
the obvious and evident needs, from the functional and usability demands
toward the satisfaction of profound, emotional, and affective, customer and
market needs (Jordan, 2000; Green and Jordan 2002).
The translation of profound needs is a complex task because the needs
are not necessarily known by the customer himself. On the other hand, the
design elements of the products as well as the technical requirements to
make them feasible are not necessarily known by the designers, engineers,
and team of experts, so current organizations require new technologies, pro-
s
cesses, and methodologies that allow (1) the detection of these new needs
ci
and requirements of the client, and (2) the parametric translation of these
needs and requirements into the design elements, its specifications and stan-
an
dards, and most important, adding more value to the market.
The pursuit of this relationship between the consumer, the design, and the
Fr
development of new products should be the centerpiece of a new deployment

of the quality and customer satisfaction approaches. Competitive organiza-
tions are clearly oriented to the customers and have found that the trans
r&
lation of the customer’s affective as well as emotional needs (Kansei voice

of the customer) is critical in the development of a new product or service.
Product quality must fit the customer’s Kansei value; he/she simply wants to
o
have an enjoyable life, and it is our responsibility to deliver the means for full
yl
satisfaction (Nagamachi 2007).

This chapter deals with the relation between the Kansei/affective engi-
Ta
neering approaches and the new product development phases and tools,
especially that known as QFD (quality function deployment), created by
Dr. Joji Akao and Dr. Shigeru Mizuno in the 1960s (Akao 1994; Akao and
Mizuno 1994) as a planning process to develop new products, services, pro-
cesses, and technologies, as well as innovative concepts.
11.2 New Product Development Phases and Tools

New product development refers to the overall disciplined and standardized
process of a company for the definition of the steps and activities to convert
ideas and concepts into salable products and services. The process usually
considers the concept generation, strategy setting and planning, research-
ing, organization, resourcing, product and marketing plan creation and
Kansei, Quality, and Quality Function Deployment 297
evaluation, as well as the commercialization of the developed individual or

portfolio of products or services. Most of the developers are adopting best
practices (Griffith 1997) and organization of the body of knowledge of the
Product Development & Management Association (Katz 2007; PDMA 2009)
which defines three macrophases of the total product life cycle as follows:
The discovery phase covers all the process of searching and identifying
the customer’s problems, needs, and benefits; defining the concep-
tual features, functions, and attributes to be built or created; as well
as all the planning activities and the strategies to achieve these mar-
ket opportunities. The discovery phase ends with the explicit defini-
tion of the formal product or service specification documents and
the elaboration of the business case (plan).
The second, named the development phase, covers all the process of
s
converting the product or service specifications into designs as
ci
well as the definition of all the activities to accomplish this, such
an
as required processes, parts or components, technologies, methods,
and resources. It usually includes the design, resource management,
test and validation, information, and engineering technology. The
Fr
development phase ends when the product or service is commer-

cially available.
r&
Finally, the commercialization phase includes the whole process of prod-

uct or service production, launching into the market, postlaunching
review, process improvements, performance and evaluations, man-
o
agement of demand, and achievement of financial goals. The com-

yl
mercialization phase ends when the product or service has reached

the end of its life cycle and decisions are to be made regarding its
Ta
retirement, renewal, or regeneration.
Different knowledge areas in the discovery phase are important in order

to capture and understand all external insights from clients, buyers, users,
channels, competitors, and substitute products, and anything having to do
with the understanding of the voice of the customer and its translation into
design and technical elements. The development phase must translate all the
information captured and defined in the discovery phase into the product
characteristics, as well as all the operational dimension of product innova-
tion, processes and tools for the development and management of technical
requirements, design, manufacturing, supply chain, and other dimensions
for the product creation, process standardization, and improvement.
Various tools and methodologies are used along each one of the phases and
also depending on the main objectives or the area of knowledge, such as cus-
tomer and market research, technology and intellectual property, strategy
and planning, people and teams, alliances, and finally, process, execution
Table 11.1
Phases and Tools in the Development of a New Product
Discovery Phase Development Phase Commercialization Phase
Growth share model Design automation tools Advertisement
Benchmarking Design of experiments Customer service
Business case/business plan FMEA ERP
Ethnography (Mariampolski 2006) KAIZEN activities KAIZEN activities
Competitive intelligence (Kahaner Kansei/affective Management
1998) engineering Market research
Conjoint analysis (Green et al. 1999) Market testing Outsourcing
Kansei/affective engineering QFD Management systems
(Nagamachi 1999, 2004) Simulation
Patent mapping and mining (Kahn Technology road mapping
2003) (García and Bray 1997)
s
Pugh analysis Toyota production system
QFD (Akao 1994)
TRIZ (Altshuller 1999)
ci
an
Voice of the customer analysis
(Katz 2004, Shillito 2001)
Source: Hirata, R. (2009). Traducción de las emociones y sensaciones del cliente en productos y
Fr
servicios, Ph.D. Dissertation, UNAM, Graduate School of Management Science, Mexico

City, Mexico.
r&
and metrics. Table 11.1 shows an example of the tools and techniques com-
monly used in each of the product development phases (Hirata, 2009).
Of all these tools, two of them cross the discovery phase and are used in
o
the development phase articulating the market needs, definition of techni-

yl
cal requirements, and design elements, as well as the definition of the final
Ta
product or service attributes or quality characteristics. These tools are QFD

and Kansei/affective engineering.
11.3 Kansei/Affective Engineering
Kansei is a Japanese word with no direct and precise translation to English or
Spanish, but its meaning is nearer to a psychological feeling rather than an
emotion. In the new product development context, it can be defined as the
image a person has of a determined product, environment, or situation, when
sensed through the senses of sight, hearing, taste, smell, and touch. Kansei is
the consumer’s psychological feeling and mental image regarding a product
or service (Nagamachi 2004). Kansei needs are not easy to measure because
they depend on the individual experience and environment, but also because
their meaning depends on the context, the time, and the culture of the region
or country. The meaning of elegant or masculine can deliver different mental
images in Japan and other countries and 5 years ago versus today.
Kansei/affective engineering was founded by Mitsuo Nagamachi at

Hiroshima University about 35 years ago (Nagamachi, 1989; 1995; 1999;
2007). Kansei/affective engineering aims at the discovery and translation of
the customer’s affective and emotional needs and is considered one of the
best-structured methodologies in the world to translate the Kansei needs
into attributes of the new product, its characteristics, and its functions (i.e.,
design elements or decision rules). It is also known by the names of affective
engineering, affective ergonomics, and emotional engineering; and it is an
evolving (i.e., enriched through other tools and methods) customer-oriented
product development technology that invariably leads to a better satisfaction
as products and services match and/or exceed customers’ profound needs,
desires, and feelings (Hirata, 2008).
Different types of Kansei/affective engineering have been defined by
Dr. Nagamachi, and they depend on successful applications in new fields
s
based on the following elements: Totally new product or innovation from an
ci
existing product; clear definition of customer needs (explicit needs); clear def-
inition of technical characteristics that make customer satisfaction possible;
an
computer system databases with a knowledge base that controls the system
and modeling; virtual imaging; methodological approach to define design
Fr
rules; and finally, product, service, or community-development orientation.

The customer’s psychological responses (Kansei) are more general qualita-
tive characteristics, and in consequence, difficult to measure; but in order
r&
to transfer Kansei into design elements, qualitative psychological phenom-

ena should be changed to quantified characteristics (linking Kansei with
design technical specifications). For example, in the beverage industry, if the
o
developmental target is the design of a new beverage flavor, we must not

yl
only define the market’s specific segment and its functional needs regarding
quality, quantity, cost and delivery, safety, and service, but also understand
Ta
its Kansei needs (e.g., “I want an urbanlike drink,” or “I want a attractive

drink”), translate these profound voices of the customer, and deploy them
into technical characteristics or design elements that satisfy them (e.g., can
color, brand name, letter font, percentage of CO2, dryness, liquid color).
The applications of Kansei/affective engineering can be found in the auto-
motive industry, construction machinery, electric home appliances, office
automation machinery, audiovisual equipment, home construction materi-
als, shoe and garment industry, cosmetic industries and laboratories, statio-
nery products, community design projects, and food industry.
It is important to note that in highly competitive markets, the Kansei/
affective engineering approach has a purpose to enhance quality of life through
customer satisfaction.
The general model of Kansei/affective engineering has the following phases:
1. Selection of the product or service domain and definition of the

strategy, which includes the selection of the product or service
(existing or totally new one), definition of the market and current
competition with its solutions, potential market segments, senses

to be used in the study (sight, taste, smell, touch, hearing) and their
combinations, and finally, the general definition of the strategy and
plan of the project. This phase shall include potential concepts as
well as solutions not yet developed in order to cover a larger scope
of the domain.
2. Definition of the semantic space and its structure, which include the
collection of adjectives that describe the product or service domain
and the potential Kansei needs (i.e., profound needs of the market)
called Kansei words (e.g., elegant, masculine, sober, attractive,
urbanlike, sexy, heavy), their categorization, definition of the hierar-
chical structure, and data collection. The Kansei words are collected
from various sources such as the team of experts, designers, expe-
rienced users, advertisements, magazines, ideas, direct observation,
s
and interviews. The list can go from 50 to more than 500 Kansei
ci
words and is commonly categorized (i.e., in groups) in a manual or
statistical approach. In the manual form, a group of experts hierar-
an
chically organize the Kansei words depending on how specific or
general the adjective is (Figure 11.1 shows an example of the hierar-
Fr
chical structure of Kansei words). The statistical approaches include

factor analysis, principal component analysis, cluster analysis, and
others (Figure 11.2 shows an example of a principal component anal-
r&
ysis graph to reduce the dimensionality of Kansei words). The main

objective is to determine the most representative Kansei needs.
o
3. Definition of product or service properties. The objective is to deter-

mine potential properties or design elements (i.e., technical and
yl
design requirements) of the future product or service, which include

Ta
the collection of existing products, creation of new concepts, identi-

fication of potential customer and company images and priorities, as
well as the definition of properties, elements (i.e., attributes or char-
acteristics) and design categories.
4. Data collection. This phase is where the semantic space or the Kansei
needs are related to the potential product or service properties
through evaluations made by user surveys, direct observation, or
physiological measurements. Kansei need is a response variable,
and the potential properties or design elements are the independent
variables of the model.
5. Data analysis. Data are analyzed and can be processed through
manual (e.g., category classification method), statistical (e.g., regres-
sion analysis), or nonstatistical methods (e.g., rough sets theory;
Nishino, 2005) in order to obtain the best approximation for the rela-
tion between the profound Kansei needs and the design elements
(Figure 11.3 shows an example of the translation of Kansei needs
into design elements).
3rd. level
Non fruity
2nd. level
Dry Flavor
1st. level
Bitter Heaviness
Full bodied
Drinkability
Dense
General Specific
Figure 11.1
Hierarchical Kansei word structure.
s
1.0
ci
Component 2
Attractiveness
an
0.5
Fr
r&
0.0
o
yl
–0.5
Ta
Simpleness
–1.0 –0.5 0.0 0.5 1.0

–1.0
Component 1
Heaviness Lightness
Figure 11.2
Principal component analysis of Kansei words (Hirata, R., Nagamachi, M., Ishihara, S. (2004a).
Satisfying emotional needs of the beer consumer through Kansei Engineering Study, Proceedings
of the 7th International QMOD Conference 2004, University of Linköping and ITESM; Hirata, R.,
Nagamachi, M., Ishihara, S. (2004b). Satisfying emotional needs of Mexican beer consumers mar-
ket through Kansei Engineering Study, Proceedings of 10th ICQFD Conference 2004, Mexico.).
6. Validation and prototype construction. All the results shall be tested

for validity and discussed with the team of experts (e.g., technicians,
designers, expert users) before proceeding to the construction of
prototypes and market testing, as well as planning for the next pro-
duction and market introduction processes (i.e., development and
commercialization phases).
Design elements
Illustration or logo
1st. level Kansei need
Package main color
Kansei/Affective
Heaviness
Engineering analysis Number of lines
Package texture
Need Technical/Design requirement
Figure 11.3
Kansei need translation to design elements.
Representative Kansei/Affective Engineering

Kansei Needs Analysis
s
3rd level 2nd level 1st level 1st level 2nd level 3rd level
ci
an
Definition Collection Hierarchical Kansei needs structure Category classification Design
1.0
Fr
of product of target elements

domain & market or
Component 2
0.5
goals needs Data Technical
0.0
analysis reqs.
r&
–0.5 (statistical
–1.0 or non
–1.0 –0.5 0.0 0.5 1.0 statistical
Component 1 methods)
o
Statistical definition of
representative Kansei needs
yl
Ta
Figure 11.4
Translation of market needs and Kansei/affective engineering.
Kansei/affective engineering is a structured approach for the identification,

categorization, and translation of profound Kansei needs of the market into
physical design elements or technical requirements, using a diverse variety
of quantitative and qualitative tools. Figure 11.4 shows a general process for
Kansei/affective engineering.
11.4 Quality Function Deployment

Quality function deployment (QFD) was originated in Japan by Dr. Joji Akao
while linking the critical points of quality assurance to be carried out through
the design and manufacturing processes. Around 1972, he and Dr. Shigeru
Mizuno developed a matrix of customer demands and quality characteristics
that was used at the Kobe Shipyards of Mitsubishi Heavy Industries. Today,
QFD is a comprehensive method for the product or service design and plan-
ning ensuring customer satisfaction through meeting and translating his
requirements and demands throughout each stage of the product or service
development process (Akao 1994; Akao and Mizuno 1994). Its main function
is to translate customer needs and desires into the technical requirements
of the product or service, as well as deploying its features, components,
required technologies, process characteristics, required capability and reli-
ability, and deliverable actions, among others. It links the demanded quality
of the market with design, development, engineering, production, and ser-
vice functions, aligning all company departments to the construction of the
benefits the customer requirements and desires (i.e., value from the customer
perspective) rather than elimination of errors and claims of defects.
The QFD process gathers and organizes the customer needs and then tai-
s
lors a specific strategy in order to translate or deploy the market require-
ci
ments into means to accomplish them using matrix relationships (e.g.,
target–means matrix) where rows represent the requirements and columns,
an
the alternative or potential means. Figure 11.5 presents an example of how
QFD can deploy demanded quality from the market into technical char-
Fr
acteristics (quality deployment, matrix 1), demanded quality into product

functions (function deployment, matrix 2), demanded quality into failures
to avoid (reliability deployment, matrix 3), as well as further deployments
r&
such as the relation of functions and technology (technology deployment,

matrix 4).
o
Quality deployment Function deployment Reliability deployment Technology deployment

yl
Technical Functions Failure

Matrix 1 requirements Matrix 2 Matrix 3 modes
Ta
Demanded Demanded Demanded

quality quality quality
Functions Technology
Matrix 4
Technical Technical
reqs. reqs.
Failure Technology
modes
Functions Functions
Technical Functions Failure Technology

requirements modes
Parts list Parts list Parts list Parts list
Figure 11.5
Comprehensive QFD (example of deployments).
Vision deployment Core competencies deployment Market Segmentation deployment
Goals and Core

measures competencies
Company’s Company’s
vision vision
Core Customer
competencies segments
Goals and Goals and

measures measures
Voice of the
customer
analysis
Customer
segments
Core
competencies
s
Figure 11.6
Vision deployments with QFD matrix (example of deployments).
ci
an
As a planning and design process, a modern QFD matrix is also used to
bridge the gap between the company’s priorities and needs with product
Fr
designs. QFD links to strategic planning deploying the company’s vision into
its business plan, strategy definition, market segmentation, and customer
selection, where it connects into the traditional QFD with the definition of
r&
customer needs and demands (Shillito 1994). Figure 11.6 shows an example
of the vision deployment into customer needs analysis.
One of the key matrices in the QFD methodology is the relationship
o
matrix in the quality deployment stage (Figure 11.7, matrix 1), where cus-
yl
tomer needs are related to technical requirements by a group of experts and

Ta
generally built as the first relationship for product planning. QFD requires
the demanded quality to be prioritized, in order to determine key technical
requirements and its targets. For this purpose a quality planning table (QPT)
is constructed, which includes customer priorities, an evaluation between
the company’s current products and the competitive products and other ele-
ments such as sales opportunities (Terninko 1997). Other relationship matri-
ces such as the product planning table (PPT) are added and create what is
commonly known as the House of Quality (HoQ) (Figure 11.7 is an image of
the HoQ with its matrices).
Customer needs are collected, interpreted, and organized through vari-
ous methods such as focus groups, consumer brainstorming, nominal
group technique, ethnography (Mariampolski 2006), voice of the customer
(VOC) analysis (Katz 2004; Shillito 2001), in-depth interviews, customer vis-
its (McQuarrie 1998), lead user analyses (Von Hippel 1988, 2005), and oth-
ers. The key point is to clearly differentiate and separate the customer’s
expected features (functions) and the required and expected benefits, named
House of Quality (HoQ)
Technical reqs.
relations
Technical
requirements
QPT
Demanded Quality
quality Relationship Planning
matrix 1 Table
PPT
Product
Planning Table
(target values)
s
Figure 11.7
Quality deployments and the HoQ.
ci
an
House of Quality
Tech.
Fr
Functional Needs Demanded Quality (DQ) reqs. relations

(Features) (Benefits)
Technical
1st level 2nd level 3rd level 1st level 2nd level 3rd level Requirements
r&
Definition QPT
of product Analysis Quality
domain & of VOC Relationship Planning
target Matrix 1 Table
o
Breakdown of features Breakdown of DQ Target values

yl
and standards
Ta
Figure 11.8
From VOC to DQ for the HoQ construction.
demanded quality (DQ) for the correct input of the relationship matrix and
HoQ. Figure 11.8 shows a general process for the definition of the DQ.
11.5 QFD and Kansei

Since QFD for product planning generally focuses on technical solutions,
Kansei needs are difficult to identify, collect, or analyze from the gathered
data. Additionally, common tools for traditional QFD grasp explicit customer
needs and link them to technical requirements and design elements known
by a group of experts (i.e., explicit potential solutions, measures, features).
Explicit
Customer Needs
QFD
Kansei/Affective
Tacit
Engineering
Tacit Explicit
Solutions
Figure 11.9
Relation between Kansei/affective engineering and QFD scope (Hirata, R. (2005). Understand
ing emotional needs of the Japanese and Mexican beer consumer market through Kansei
Engineering Study, presented at International Congress on Quality, Tokyo, 2005.).
Kansei/affective engineering is very useful and an effective technology
s
from the outset, when customer needs are profound and need to be dis-
ci
covered, gathered, and evaluated before deploying solutions (i.e., tacit or
unconscious needs). Second, it translates the Kansei needs into technical
an
requirements or design elements not necessarily known by the group of
experts (i.e., tacit or unknown potential solutions, measures, features).
Fr
Once the Kansei needs and the technical requirements are placed in the
explicit arena, QFD is a powerful and profitable technology for the deploy-
ment of the product or service. The relation between Kansei/affective engi-
r&
neering and QFD regarding its scope is shown in Figure 11.9 (Hirata, 2005).
Kansei/affective engineering is useful and can articulate with QFD in the
identification and categorization of profound, tacit, emotional, and affective
o
needs (i.e., Kansei needs) and is complementary (or prior) to the QFD (VOC)
yl
analysis. Both approaches will feed the relationship matrix of the HoQ with
categorized and organized explicit demanded quality (DQ) elements (e.g.,
Ta
breakdown of DQ). In QFD, a group of experts defines the possible techni-

cal requirements for the DQ elements, and the strengths of their relation-
ships are also determined by distinguishing between a strong, medium, or
weak relationship, which accelerates the deployment and the decision mak-
ing processes under the assumption that all the technical requirements are
known, as well as necessary and sufficient. In case the assumption cannot be
assured, then further quantitative studies shall be done.
Regarding the determination of the technical requirements or design ele-
ments for Kansei needs, Kansei/affective engineering offers quantitative
evidence through statistical and nonstatistical approaches (e.g., regression
analysis, rough sets) and as a consequence, it also provides a more detailed
quantification of the relationships between the DQ and the technical require-
ments giving more certainty to the design or experts’ team. Kansei/affective
engineering also feeds the HoQ with the definition of technical require-
ments or physical design elements, which then can be deployed into other
QFD functions. A full approach with QFD and Kansei/affective engineering
is shown in Figure 11.10 (Hirata, 2005).
Analysis
1st level 2nd level 3rd level
Design
Representative Elements
Kansei Needs Category classification or
Technical
3rd level 2nd level 1st level Data analysis reqs.
(statistical or
non statistical
methods)
Kansei:
Hierarchical Kansei needs structure
Definition Collection 1.0
of product of target
Component 2
domain & market 0.5

goals needs
0.0
s
–0.5
Tech.
ci
–1.0
–1.0 –0.5 0.0 0.5 1.0 Demanded Quality (DQ) reqs. relations
from Kansei & QFD
Component 1 Technical
an
Statistical definition of 1st level 2nd level 3rd level requirements
representative Kansei needs
Functional Needs
Fr
(Features)
QPT
1st level 2nd level 3rd level Quality
Planning
Table
r&
QFD:
Analysis Relationship
of Matrix 1
VOC Breakdown of features Breakdown of DQ
o
Target values
and standards
yl
Figure 11.10
Ta
Kansei/affective engineering and QFD interaction toward HoQ (Hirata, R. (2005). Under
standi ng emotional needs of the Japanese and Mexican beer consumer market through Kansei
Engineering Study, presented at International Congress on Quality, Tokyo, 2005.).
11.6 Conclusions
Translation of affective and emotional needs of the markets into design ele-
ments and attributes of products and services is necessary to satisfy pro-
found human needs. The discovery phase in the new product development
process needs tools and methods to identify, organize, and translate Kansei
needs into product or service specifications, and companies shall integrate
Kansei/affective engineering as a method for this purpose. In product plan-
ning, Kansei/affective engineering is useful to identify the Kansei needs of
the target market, especially in the current competitive environments where
customer satisfaction is evolving from the satisfaction of the obvious and
explicit needs, mostly functionality and usability demands, toward the satis-
faction of profound, tacit, emotional, and affective needs.
Kansei/affective engineering, as well as QFD, is a flexible and evolving

model that allows the addition and integration of other tools and methods
throughout their various stages. Therefore, both can articulate in the pro-
cess of product and service development providing the tools and processes to
translate Kansei needs into design elements, as well as the numerical evidence
to support the decision making processes bringing shorter product develop-
ment cycle times, reduced development costs, and increased satisfaction of
functional, usability, affective, and emotional customer needs and in conse-
quence, a better success rate of new products and services in the market.
References
s
ci
Altshuller, G. S. (1999). The Innovation Algorithm, Technical Innovation Center, Worcester,
MA.
an
Akao, J. (1994). Quality Function Deployment, Productivity Press, New York.
Akao, J., and Mizuno, S. (1994). QFD: Customer Driven Approach to Quality Planning
and Deployment. Asian Productivity Organization, Tokyo.
Fr
Cooper, R., and Edgett, S. (2008). Ideation for product innovation: What are the best
methods? Product Innovation Best Practices Series, Reference paper #29, Stage-
Gate Inc. & Product Development Institute.
r&
Green, P., Wind, J., and Rao, V. R. (1999). Conjoint analysis: Methods and applica-
tions. In Dorf, R. C., (Ed.), The Technology Management Handbook, CRC Press,
New York.
o
Garcia, M. L., and Bray, O. H. (1997). Fundamentals of Technology Roadmapping, Sandia

yl
National Laboratories, Albuquerque, NM.

Green, W. S., and Jordan, P. W. (Eds.) (2002). Pleasurable Products: Beyond Usability,
Ta
CRC Press, New York.

Griffith, A., and Belliveau, P. (1997). Drivers of NPD Success: The 1997 PDMA Report,
Product Development & Management Association PDMA, Mount Laurel, NJ.
Hirata, R. (2005). Understanding emotional needs of the Japanese and Mexican beer
consumer market through Kansei Engineering Study, presented at International
Congress on Quality, Tokyo, 2005.
Hirata, R. (2009). Traducción de las emociones y sensaciones del cliente en produc-
tos y servicios, Ph.D. Dissertation, UNAM, Graduate School of Management
Science, Mexico City, Mexico.
Hirata, R., Nagamachi, M., Ishihara, S. (2004a). Satisfying emotional needs of the beer
consumer through Kansei Engineering Study, Proceedings of the 7th International
QMOD Conference 2004, University of Linköping and ITESM.
Hirata, R., Nagamachi, M., Ishihara, S. (2004b). Satisfying emotional needs of Mexican
beer consumers market through Kansei Engineering Study, Proceedings of 10th
ICQFD Conference 2004, Mexico.
Hirata, R., Nagamachi, M., Ishihara, S., Nishino, T. (2008). Translation of customer
Kansei and emotional needs into products, The 2nd International Conference on
Applied Human Factors and Ergonomics (AHFEI) 2008, Las Vegas.
Jordan, P. W. (2000). Designing Pleasurable Products, CRC Press, London.

Kahaner, L. (1998). Competitive Intelligence, Touchstone, New York, and European
University Association, Brussels.
Kahn, E. (2003). Patent mining in a changing world of technology and product devel-
opment, Intellectual Asset Management, July/August.
Katz, G. M. (2004). The vVoice of the customer, Chapter 7. In P. Belliveau, A. Griffin,
and S. Somermeyer (Eds.), PDMA Toolbook 2 for New Product Development, John
Wiley & Sons, Hoboken, NJ.
Katz, G. (2007). The PDMA’s body of knowledge. In PDMA Toolbook 3 for New
Product Development, Griffin, A., and Somermeyer, S. (Eds.), John Wiley & Sons,
Hoboken, NJ.
Mariampolski, H. (2006). Ethnography for Marketers: A Guide to Consumer Immersion,
Sage Publications, Thousand Oaks, CA.
McQuarrie, E. F. (1998). Customer Visits, Sage Publications, European University
Association, Brussels.
s
Nagamachi, M. (1989). Kansei Engineering, Kaibundo Publishing, Japan.
Nagamachi, M. (1995). The Story of Kansei Engineering, Japanese Standards Association,
Tokyo.
ci
an
Nagamachi, M. (1999). Kansei engineering: A new consumer oriented technology for
product development. In W. Karwowski and W. S. Morris (Eds.), The Occupational
Ergonomics Handbook, CRC Press, New York.
Fr
Nagamachi, M. (2004). Kansei engineering. In N. Stanton and A. Hedge et al. (Eds.),

Handbook of Human Factors and Ergonomics Methods, CRC Press, New York.
Nagamachi, M. (2007). Perspectives and new trend of Kansei/affective engineering,
r&
The 1st European Conference on Affective Design and Kansei Engineering & 10th
QMOD Conference, University of Linkoping and Lund University, Helsingborg.
Nagamachi, M., and Imada, A. (1995). Kansei engineering: An ergonomic technology
o
for product development, International Journal of Industrial Ergonomics, 15 (1),

yl
January.
Nishino, T. (2005). Rough sets and Kansei rules definition. In M. Nagamachi (Ed.),
Ta
Product Development and Kansei, Chapter 9, Kaibundo, Tokyo.

PDMA Foundation (2004). Comparative Performance Assessment Study, PDMA, USA.
PDMA (2009). Product Development and Management Association, Body of
Knowledge, http://pdmabok.arcstone.com/(accessed September 30, 2009).
Schifferstein, H. N. J., and Hekkert, P. (Eds.) (2008). Product Experience, Elsevier,
Amsterdam.
Shillito, L. (1994). Advanced QFD, John Wiley & Sons, New York.
Shillito, L. (2001). Acquiring, Processing and Deploying Voice of the Customer, CRC Press,
New York.
Terninko, J. (1997). Step by Step QFD, 2nd ed., St. Lucie Press, Boca Raton, FL.
Von Hippel, E. (1988). The Sources of Innovation, Oxford University Press, Oxford, U.K.
Von Hippel, E. (2005). Democratizing Innovation, The MIT Press, Cambridge, MA.

Lockman 2010 B

Uploaded by

Document Informationclick to expand document information

Copyright:

Available Formats

Lockman 2010 B

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Lockman 2010 B

Uploaded by

Copyright:

Available Formats

CRC PRESS .

TAYLOR & FRANCIS

Page 28 - Chapter 9 - Kansei/Affective Engineering for the European

Page 50 - Chapter 11 - Kansei, Quality, and Quality Function Deployment

20% Discount Available*

Publishing with CRC Press/Taylor and Francis

Anitawati Mohd Lokman

8.2.1.3 L3: Determination of Kansei Concept

8.2.1.4 L4: Prototyping/Testing................................................. 232

8.3.1 Synthesizing the Specimen........................................................... 233

8.3.1.2 Investigation of Design Elements.................................. 233

8.3.1.3 Classification of Item and Category..............................234

8.1 Introduction to Web Design

usability, visibility/searchability, and accessibility.

8.1.1 Positioning Kansei/Affective Engineering in Web Design

designers often misunderstand the description provided by users or user

users have described.

8.2 Engineering Kansei in Web Design

8.2.1 Structuring the KAE Method into Kansei Design Model

the audience can have a useful guide to the implementation of KAE.

The model is developed to provide a systematic approach to the imple-

8.2.1.1 L1: Synthesizing the Specimen

elements enables a more objective measurement. On the other hand, including

8.2.1.2 L2: Establishment of Kansei Checklist

magazines, or even consulting experts. Finally, utilizing the Kansei words,

8.2.1.3 L3: Determination of Kansei Concept and Requirement

L3 describes determination of the Kansei concept and requirement. This

analysis. In the first step, Kansei measurement is performed using expert or

8.2.1.4 L4: Prototyping/Testing

8.3 Demonstration of the Kansei Design Model

8.3.1 Synthesizing the Specimen

2. Investigation of design elements

4. Finalizing valid specimen

8.3.1.1 Identification of Initial Specimens

8.3.1.2 Investigation of Design Elements

8.3.1.3 Classification of Item and Category

Example of Design Item and Category

Page background color Blue

8.3.1.4 Finalizing Valid Specimen

Rule No. 2 163

8.3.2 Establishment of Checklist

8.3.2.1 Checklist Development

8.3.2.2 Synthesize Kansei Words

8.3.2.3 Development of Kansei Checklist

8.3.3 Determination of Kansei Concept and Requirement

8.3.3.1 Kansei Measurement

1. Instrumentation. The instruments involved in the experimental pro-

a. The 35 valid specimens. The specimens were coded numeri-

b. The Kansei checklist, which consists of 40 Kansei words orga-

organized into five different arrangements to minimize bias in

literature in information system studies that suggests students

shown one by one on a large white screen to all subjects in a systematic

8.3.3.2 Requirement Analysis

8.3.3.2.1 The Kansei Concept

rotation method that simplifies the interpretation of variables. In the

shows factor results in ascending order. The structure of Kansei words is

as significant factors in Web site design. The research set approximately

8.3.3.2.2 Analyzing the Relationship of Design and Kansei

1. The dependant (objective) variables, y, that is, the 40 sets of Kansei

8.1 Introduction to Web Design

8.1.1 Positioning Kansei/Affective Engineering in Web Design

8.2 Engineering Kansei in Web Design

8.2.1 Structuring the KAE Method into Kansei Design Model

8.2.1.1 L1: Synthesizing the Specimen

8.2.1.2 L2: Establishment of Kansei Checklist

8.2.1.3 L3: Determination of Kansei Concept and Requirement

8.2.1.4 L4: Prototyping/Testing

8.3 Demonstration of the Kansei Design Model

8.3.1 Synthesizing the Specimen

8.3.1.1 Identification of Initial Specimens

8.3.1.2 Investigation of Design Elements

8.3.1.3 Classification of Item and Category

8.3.1.4 Finalizing Valid Specimen

8.3.2 Establishment of Checklist

8.3.2.1 Checklist Development

8.3.2.2 Synthesize Kansei Words

8.3.2.3 Development of Kansei Checklist

8.3.3 Determination of Kansei Concept and Requirement

8.3.3.1 Kansei Measurement

8.3.3.2 Requirement Analysis

8.3.3.2.1 The Kansei Concept

8.3.3.2.2 Analyzing the Relationship of Design and Kansei

8.3.3.2.3 Proposing Kansei Web Design Guidelines

8.3.4.2.1 Extracting the Exploratory Data

8.3.4.2.3 Comparison of the Exploratory and Confirmatory Data

9.1.1 Building a Successful Brand

9.1.2 New Product Development in FMCG

9.1.3 W hy Does FMCG Need Kansei/Affective Engineering?

9.2 Kansei/Affective Engineering Framework

9.2.1 Stage 1: Receive Brief

9.2.2.2 Evaluation Variables

9.2.2.3 Design Attribute Variables