An analytical Kano model for customer

need analysis
Qianli Xu, Roger J. Jiao, Xi Yang and Martin Helander, School of Mechanical
and Aerospace Engineering, Nanyang Technological University, Singapore
Roger J. Jiao, Woodruff School of Mechanical Engineering, Georgia Institute
of Technology, USA
Halimahtun M. Khalid, Damai Sciences Sdn Bhd, Malaysia
Anders Opperud, Volvo Technology Corporation, Sweden
In an effort to address the inherent deficiencies of traditional Kano method, this
paper proposes an analytical Kano (A-Kano) model with focus on customer
need analysis. Kano indices in accordance with the Kano principles are
proposed to incorporate quantitative measures into customer satisfaction.
Accordingly, two alternative mechanisms are proposed to provide decision
support to product design, (1) the Kano classifiers are used as tangible criteria
for categorizing customer needs, and (2) the configuration index is introduced
as a decision factor of product configuration design. The merit of product
configurations is justified using a Kano evaluator, which leverages upon both the
customer’s satisfaction and the producer’s capacity. A case study of dashboard
in automotive design is also presented. It is demonstrated that the A-Kano
model can effectively incorporate customer preferences in product design, while
leading to an optimal tradeoff between customer’s satisfaction and producer’s
capacity.
Ó 2008 Elsevier Ltd. All rights reserved.
Keywords: product design, automotive design, perception, evaluation, Kano
M
anufacturing enterprises are increasingly focusing on satisfying
individual customer needs (CNs) in a highly competitive global
market. A constant challenge for manufacturers is how to deal
with the customer satisfaction, which in turn largely determines the customers’
willingness to buy the products. Understanding and fulfilling CNs have been
well recognized as one of the principle factors for product design and develop-
ment to succeed in the market place (McKay et al., 2001).
Analysis of customer need information is an important task with focus on the
interpretation of the voice of customers and subsequently derivation of ex-
plicit requirements that can be understood by marketing and engineering
(Jiao and Chen, 2006). In general, it involves three major issues, namely
(1) understanding of customer preferences, (2) requirement prioritization,
Corresponding author:
Roger J. Jiao
jiao@ieee.org
www.elsevier.com/locate/destud
0142-694X $ - see front matter Design Studies 30 (2009) 87e110
doi:10.1016/j.destud.2008.07.001 87
Ó 2008 Elsevier Ltd. All rights reserved. Printed in Great Britain
and (3) requirement classification. Among many approaches that address cus-
tomer need analysis, the Kano model has been widely practiced in industries
as an effective tool of understanding customer preferences owing to its con-
venience in classifying CNs based on survey data (Kano et al., 1984). Never-
theless, traditional Kano methods are not equipped with quantitative
assessment. And, Kano classification provides limited decision support in
engineering design. Moreover, it inherently emphasizes the customer and
market perspectives only, with limited consideration of the producer’s capac-
ity to fulfill the CNs.
To enhance the above aspects in relation to customer need analysis, this paper
scrutinizes the theoretical foundation of the original Kano model and pro-
poses an analytical Kano (A-Kano) model. The A-Kano model extends tradi-
tional Kano model (Kano et al., 1984; Berger et al., 1993) by introducing (1)
Kano indices, which are quantitative measurements of customer satisfaction
derived from Kano questionnaires and surveys; (2) Kano classifiers, which con-
sist of a set of criteria to classify CNs based on the Kano indices; (3) Config-
uration index, which provides a decision factor for selecting the functional
requirements (FRs) that contribute to product configurations; and (4) Kano
evaluator, which is a shared surplus-based performance indicator leveraging
upon both the customer’s satisfaction and the producer’s capacity. A compre-
hensive process model is proposed to integrate these techniques for customer
need analysis.
The subsequent sections proceed as follows. A critical review of the Kano
model is presented in Section 1. The fundamental issues of A-Kano are scru-
tinized in Section 2. The A-Kano process model is proposed in Section 3 by
implementing the proposed techniques that address the fundamental issues.
In Section 4, the A-Kano model is applied to car dashboard design. The results
are compared with those derived from the traditional Kano model. Discus-
sions and conclusions are presented in Sections 5 and 6, respectively.
1 Critical review of Kano model
The Kano model of customer satisfaction is a useful tool to classify and prior-
itize customer needs based on how they affect customer’s satisfaction (Kano
et al., 1984). It captures the nonlinear relationship between product perfor-
mance and customer satisfaction. In practice, four types of product attributes
are identified: (1) must-be attributes are expected by the customers and they
lead to extreme customer dissatisfaction if they are absent or poorly satisfied,
(2) one-dimensional attributes are those for which better fulfillment leads to
linear increment of customer satisfaction, (3) attractive attributes are usually
unexpected by the customers and can result in great satisfaction if they are
available, and (4) indifferent attributes are those that the customer is not inter-
ested in the level of their performance.
88 Design Studies Vol 30 No. 1 January 2009
The Kano model is constructed through customer surveys, where a customer
questionnaire contains a set of question pairs for each and every product attri-
bute. The question pair includes a functional form question, which captures
the customers’ response if a product has a certain attribute, and a dysfunc-
tional form question, which captures the customers’ response if the product
does not have that attribute. The questionnaire is deployed to a number of
customers, and each answer pair is aligned with the Kano evaluation table
(Berger et al., 1993), revealing an individual customer’s perception of a product
attribute. The final classification of a product attribute is made based on a sta-
tistical analysis of the survey results of all respondents.
1.1 Qualitative vs. quantitative evaluation
The Kano diagram provides a rough sketch of the customer’s satisfaction in
relation to the product performance level. In such a sense, it only allows qual-
itative assessment of product attributes (Wassenaar et al., 2005; Rivie` re et al.,
2006). A convenient way to incorporate quantitative measures is to assign
some scales in terms of the levels of customer satisfaction/dissatisfaction
(Matzler and Hinterhuber, 1998). However, the resulting Kano category is still
qualitative in nature, which could not precisely reflect the extent to which the
customers are satisfied (Berger et al., 1993). Hence, the application of Kano
model in engineering design is primitive in comparison with other quantitative
methods, such as conjoint analysis (Green and DeSarbo, 1978), stated choice
methods (Louviere et al., 2000), and decision-based design (Hazelrigg, 1998).
1.2 Classification criteria
Different criteria have been adopted for classifying customer requirements,
such as empirical observation, mode statistics, and customer satisfaction coef-
ficient (Berger et al., 1993). Timko (Berger et al., 1993) proposes a two-
dimensional representation of Kano quality category based on the customer
satisfaction coefficients. In particular, a positive number is used to represent
the relative value of meeting the respective customer requirement, whilst a neg-
ative number is used to reflect the relative cost of not meeting this customer
requirement. However, classification criteria are not explicitly defined in this
model. DuMouchel (Berger et al., 1993) proposes a graphical Kano diagram
that is based on predefined scales related to the customer’s satisfaction and
dissatisfaction. Each customer requirement can be represented as a pair of sat-
isfaction and dissatisfaction values. The nature of a customer requirement can
be delineated by the quadrant into which that point falls. However, it is very
subjective to classify the customer requirements into four quadrants because of
a lack of logical classification criteria.
1.3 Decision support
The ultimate goal of customer need analysis is to provide decision support to
product design. Although the Kano categories may enhance designers’ under-
standing of customer needs, they fall short as concrete decision criteria. In
general, the categories of the Kano model signify different priorities in
An analytical Kano model 89
designing products. For example, a designer is directed toward fulfilling all
must-be FRs, staying competitive with market leaders on the one-dimensional
FRs, and including some attractive FRs (Berger et al., 1993). However, such
approaches do not distinguish FRs within the same category. Hence, the
Kano classification is deemed to be inadequate to facilitate decisions in prod-
uct design.
1.4 Producer’s capacity
The Kano model is inherently customer-driven, i.e., it focuses exclusively on
addressing the concerns of customers (Sireli et al., 2007). As a decision-making
tool used by engineers, the Kano model fails to account for the producer’s con-
cerns in terms of the capacity to fulfill the CNs. Cost constraints are usually
accommodated by the product development team based on expertise, such
that the product will only include those features that are affordable to the pro-
ducers (Matzler and Hinterhuber, 1998). In practice, without accurate cost es-
timation, such an assumption seldom holds true. Lai et al. (2004) propose
a linear cost function to constrain the selection of product features in product
development. Tan and Shen (2000) propose an approximation transformation
function based on Kano analysis to adjust the improvement ratio of customer
attributes. However, these cost models are primitive and are inadequate to re-
flect the complexities of design and manufacturing costs. From an engineering
perspective, engineers have to seek for a leverage of the customer-perceived
value and the producer’s capacity.
2 A-Kano model
The A-Kano model entails a series of interactions between the customers and
the producers, as shown in Figure 1. In general, the CNs tend to be imprecise
and ambiguous due to their linguistic origins (Jiao and Chen, 2006). And
hence it is difficult to apply analytical tools for CN analysis. As a quick fix,
the CNs are translated into explicit and objective statements, namely the
FRs. The distinction between CNs and FRs is in line with the domain mapping
principle proposed by Suh (2001). Essentially, what a customer de facto per-
ceives is the CNs in the customer domain, rather than FRs in the functional
domain. While providing customer-perceived diversity in CNs, the producer
must seek for an economy of scale in product fulfillment, which is meant by
FRs. In this research, analysis is carried out in the functional domain, and
the FRs are assigned different priorities for product fulfillment. From the cus-
tomer’s perspective, the priority assignment corresponds to different customer
perceptions. From the producer’s perspective, the FRs are mapped onto
various product attributes, which represent the physical form of a product
in fulfillment of the FRs.
Customer perception can be explicitly measured as the overall customer satis-
faction with respect to a combination of product attributes. On the other hand,
the legacy producer’s capacity has to be assessed in terms of delivery of the
90 Design Studies Vol 30 No. 1 January 2009
desired product attributes. The value chain converges when the customer’s sat-
isfaction overlaps with the legacy producer’s capacity. Ultimately the business
success is achieved by maximizing such an overlap, either by expanding the
producer’s capacity to meet the CNs or by directing the customers to the total
capacity of the producer so that customers are better served (Jiao et al., 2003a).
2.1 Kano indices for quantification of customer satisfaction
Kano survey is carried out within specific market segments that consist of cus-
tomers with similar demographic information. Let s denote the market segment
which contains a total of J customers (respondents), i.e., shft
j
jj ¼ 1; 2; .; Jg.
Aset of FRs is identified as Fhffj
i
i ¼ 1; 2; .; Ig. Surveys are carried out to col-
lect the respondents’ evaluation of f
i
ðci ¼ 1; 2; .; IÞ according to the func-
tional and dysfunctional forms of Kano questions (Table 1). The preliminary
category of the FR is determined using the Kano evaluation table (Table 2).
For each respondent t
j
˛sðcj ¼ 1; 2; .; JÞ, the evaluation of f
i
ðci ¼ 1; 2; .; IÞ
is represented as e
ij
¼(x
ij
,y
ij
,w
ij
), where x
ij
is the score given to an FR for the
dysfunctional form question, y
ij
is the score given to an FR for the functional
form question, and w
ij
is the self-stated importance, which is the respondent’s
perception of the importance of an FR. Similar to the method proposed by
DuMouchel (Berger et al., 1993), this research adopts a scoring scheme that
defines customer’s satisfaction and dissatisfaction as shown in Table 3. The
scale is designed to be asymmetric because positive answers are considered to
be stronger responses than negative ones. Hence, the scaling has the effect of
diminishing the influence of negative evaluations (Berger et al., 1993). Further-
more, the self-stated importance score is normalized such that it falls within
a range of 0e1, as shown in Table 4.
Next, for each FR (f
i
), the average level of satisfaction for the dysfunctional
form question within market segment s is defined as X
i
, and the average level
Figure 1 Customereproducer
interactions along the product
value chain
An analytical Kano model 91
of satisfaction for the functional form question within the same market seg-
ment is defined as Y
i
, i.e.,
X
i
¼
1
J

J
j¼1
w
ij
x
ij
; Y
i
¼
1
J

J
j¼1
w
ij
y
ij
ð1Þ
The value pair ðX
i
; Y
i
Þ can be plotted in a two-dimensional diagram, where the
horizontal axis indicates the dissatisfaction score and the vertical axis stands
for the satisfaction score. Most ðX
i
; Y
i
Þ should fall in the range of 0e1 because
the negative values are results of either Questionable or Reverse categories. A
questionable category will not be included in the averages, and a Reverse cat-
egory can be transformed out of the category by reversing the sense of func-
tional and dysfunctional of questions (Berger et al., 1993). Accordingly, the
classification of an FR can be defined based on the corresponding location
of the value pair in the diagram, as shown in Figure 2.
From the customer’s perspective, the characteristics of an FR (f
i
) can be rep-
resented as a vector, i.e., f
i
wr
.
i
hðr
i
; a
i
Þ, where r
i
¼ jr
.
i
j ¼
ffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
X
2
i
þY
2
i
_
is the
magnitude of r
.
i
and a
i
¼ tan
À1
ðY
i
=X
i
Þ is the angle between r
.
i
and the hori-
zontal axis. The rationale of representing the satisfaction and dissatisfaction
as a vector r
.
i
is that it becomes equivalent to a polar form, i.e., the magnitude
Table 1 Kano questionnaire
Kano question Answer
Functional form of the question
(e.g., if the car has air bags, how do you feel?)
,I like it that way
,It must be that way
,I am neutral
,I can live with it that way
,I dislike it that way
Dysfunctional form of the question
(e.g., if the car does not have air bags, how do you feel?)
,I like it that way
,It must be that way
,I am neutral
,I can live with it that way
,I dislike it that way
Table 2 Kano evaluation table
Dysfunctional form of the question
Like Must-be Neutral Live with Dislike
Functional form of the question Like Q A A A O
Must-be R I I I M
Neutral R I I I M
Live with R I I I M
Dislike R R R R Q
A, Attractive; O, One-dimensional; M, Must-be; I, Indifferent, R, Reverse; Q, Questionable.
92 Design Studies Vol 30 No. 1 January 2009
of the vector denotes the overall importance of f
i
to the customers belonging to
segment s, and the angle a
i
determines the relative level of satisfaction and dis-
satisfaction. Therefore, the magnitude of the vector (r
i
) is called the impor-
tance index; and the angle (a
i
) is called the satisfaction index. Both
0 r
i

ffiffiffi
2
p
and 0 a
i
p=2 are collectively called the Kano indices. In
the extreme situation, a
i
¼0 means that dysfunction of f
i
causes dissatisfac-
tion, while functioning of f
i
does not enhance satisfaction, and hence it is an
ideal must-be element. Conversely, a
i
¼ p=2 means that f
i
is an ideal attractive
element.
2.2 Kano classifiers for categorization of customer needs
Based on the above formulation, the FRs can be classified into four cate-
gories, i.e., indifferent, must-be, attractive and one-dimensional, as shown
in Figure 3. A threshold value of the importance index, r
0
, is used to dif-
ferentiate important FRs from less important ones. If r
i
r
0
, f
i
is consid-
ered as unimportant, and thus called an indifferent FR. The region
defined by the sector OFI in Figure 3, where the radius is smaller than
r
0
, is considered as the indifferent region. Hence r
0
is called an indifference
threshold.
Likewise, a lower threshold value of the satisfaction index is defined as a
L
,
such that for f
i
, if r
i
> r
0
and a
i
a
L
, it is considered as a must-be FR. The
region of the must-be FRs corresponds to the sector DEFG.
A higher threshold value of the satisfaction index is defined as a
H
, such that for
f
i
, if r
i
>r
0
and a
i
>a
H
, it is considered as an attractive FR. The region of the
attractive FRs is shown as sector ABHI.
Table 3 Scores for functional/dysfunctional features
Answers to the Kano question Functional form of the question Dysfunctional form of the question
I like it that way (like) 1 À0.5
It must be that way (must-be) 0.5 À0.25
I am neutral (neutral) 0 0
I can live with it that way (live with) À0.25 0.5
I dislike it that way (dislike) À0.5 1
Table 4 Scores for self-stated importance
Not important Somewhat important Important Very important Extremely important
0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0
An analytical Kano model 93
If r
i
>r
0
and a
L
<a
i
a
H
, f
i
is considered as a one-dimensional FR. The
region of the one-dimensional FRs is shown as sector BCDGH.
The set of thresholds r
0
, a
L
, and a
H
are collectively called Kano classifiers,
denoted as k ¼(r
0
, a
L
, a
H
). According to the Kano principles, the classification
of FRs provides a decision criterion for selecting the FRs that constitute
a product configuration. To do so, this research proposes the following heuris-
tic to simulate the consequence of Kano classification on product configura-
tion design.
Figure 2 Vector representa-
tion of customer perception
on a Kano diagram
Figure 3 Kano classifier and
Kano categories
94 Design Studies Vol 30 No. 1 January 2009
Heuristic 1: A product configuration (p) is generated as a combination of
FRs ðf
i
; ci ¼ 1; 2; .; IÞ. A binary variable, z
i
, is used to indicate whether f
i
is contained in the product or not, such that
z
i
¼
_
1; if z
i
!Randomð0; 1Þ;
0; otherwise:
where the Random(0,1) generates a random number between 0 and 1, and z
i
denotes the probability that f
i
be included in the product configuration. The
probability z
i
is dependent on the Kano category that an FR belongs to, i.e.,
z
i
¼ 1, if f
i
is a must-be FR,
z
i
¼ 0:8, if f
i
is a one-dimensional FR,
z
i
¼ 0:2, if f
i
is an attractive FR,
z
i
¼ Randomð0; 1Þ, if f
i
is an indifferent FR.
Determining appropriate values of Kano classifiers is challenging in that
these threshold values may be problem-specific and context-aware for differ-
ent applications. In practice, it is deemed to be difficult to define universal
thresholds for different products. This research uses sensitivity analysis to
select and testify the appropriate settings of the Kano classifiers. Moreover,
a proper performance indicator must be defined to justify the significance of
the classification with respect to different settings of Kano classifiers, as is
discussed in Section 2.4.
2.3 Configuration index for product configuration design
Decision-making based on the Kano classification inevitably suffers the dis-
continuity problem, i.e., data points located near the borders of two adjacent
regions may be classified as different categories, while their distinction is actu-
ally minor. To alleviate such a problem, this research proposes a configuration
index to indicate the priority of an FR in fulfillment of customer expectations.
The purpose of this strategy is to provide better decision support to product
configuration design. The configuration index ðr
i
Þ is defined as a function of
Kano indices (r
i
, a
i
) to indicate the probability that an FR is contained in a
production configuration.
r
i
¼
2
ffiffiffi
2
p
3
_
1 À
a
i
p
_
r
i
ð2Þ
Given a particular a
i
, the configuration index r
i
is proportional to the impor-
tance index r
i
, which agrees with the observation that an FR with greater
influence on customer’s satisfaction/dissatisfaction is more likely to be
included in the product configuration. At the same time, for a specific value
An analytical Kano model 95
of r
i
, r
i
decreases with an increase of the satisfaction index a
i
, which reflects the
decreasing priorities associated with the Kano categories in the order of must-
be, one-dimensional and attractive.
The selection of FRs that contribute to a product configuration is contingent
on the configuration indices. In particular, a product configuration p is repre-
sented as a combination of FRs ðf
i
; ci ¼ 1; 2; .; IÞ. The following heuristic is
used to simulate the effect of configuration index on product design.
Heuristic 2: A binary variable, z
i
, is used to indicate whether f
i
is contained
in the product or not.
z
i
¼
_
1; if f
i
is contained in product p;
0; otherwise:
The value of z
i
is assigned in the product design stage based on a stochastic
procedure. Let Random(0,1) be a function that generates a random number
between 0 and 1. Variable z
i
¼1, when r
i
! Randomð0; 1Þ; and z
i
¼0, when
r
i
< Randomð0; 1Þ.
Based on this heuristic, an FR with a larger configuration index has a higher
chance to be included in the product. It should be noted that product config-
uration design is subject to other configuration constraints. For example, two
FRs may be incompatible with each other such that only one of them may be
selected in the configuration.
2.4 Kano evaluator for leveraging customer’s satisfaction
and producer’s capacity
The Kano classifiers and configuration indices provide two alternative mecha-
nisms to determine the FRs to be included in a product. While the resulting
product configuration reflects the customer’s preferences, the producers have
tostrive todesignproducts at affordable costs. Therefore, product development
involves two interrelated perspectives, i.e., increasing customer’s satisfaction
andenhancing producer’s capacity. Inthis respect, the A-Kano model explicitly
defines a Kano evaluator to estimate the value of the planned products.
2.4.1 Objective function
This research adopts the concept of shared surplus to leverage upon both cus-
tomer’s and producer’s concerns. The consumer surplus is modeled as the
overall customer’s satisfaction (U) of the product, and the producer surplus
is simulated as the overall cycle time index (C). Accordingly, the Kano evalu-
ator (E) is defined as a shared surplus-based performance indicator, i.e.,
E ¼
U
C
ð3Þ
96 Design Studies Vol 30 No. 1 January 2009
Based on the customer’s responses to the functional/dysfunctional form of
Kano questions, the configuration index reflects the customer’s satisfaction
level for individual FRs. Hence, the customer’s satisfaction (U) is defined as
the summation of the configuration indices ðr
i
Þ of FRs contained in the prod-
uct configuration, i.e.,

I
i¼1
r
i
z
i
ð4Þ
where
z
i
¼
_
1; if f
i
is contained in product p;
0; otherwise
On the other hand, producer’s capacity is related to the technical feasibility of
a design and the engineering cost, which refers to the producer surplus. This
research proposes the cycle time index (C) to estimate the producer’s capacity
as discussed next.
2.4.2 Cycle time index
The cycle time index (C) is based on the concept of process capability index
which provides an economic latitude of process variations due to product cus-
tomization (Jiao and Tseng, 1999). The process capability index gives an indi-
cation of how expensive of a customization is to be if implemented in
production. The idea is to allocate costs to those established time standards
from previously reported work and time studies and thereby circumvent the
tedious tasks of identifying various cost drivers and cost related activities.
To do so, it is necessary to find the mapping relationships between different
attribute levels and their expected consumptions of standard times within
legacy process capabilities. These part-worth standard time accounting rela-
tionships are built into the product and process platforms (Jiao et al.,
2003b). Any product configured from available attribute levels is justified
based on its expected cycle time, which demonstrates the distinctions between
variables that differ as a result of random error and are often well described by
a normal distribution (Tielemans, 1995). The one-side specification limit pro-
cess capability index (P
CI
) can be formulated as:
P
CI
¼
USL
T
Àm
T
3s
T
ð5Þ
where USL
T
, m
T
, and s
T
are the upper specification limit, the mean, and the
standard deviation of the estimated cycle time, respectively. Variations in the cy-
cle time are characterized by m
T
, and s
T
, reflecting the compound effect of mul-
tiple products on production in terms of process variations. The USL
T
can be
determined ex ante based on the worst case analysis of a given process platform,
in which standard routings can be reconfigured to accommodate various prod-
ucts derived from the corresponding product platform (Jiao et al., 2003b).
An analytical Kano model 97
Owing to the close correlation between costs and the cycle time, P
CI
indicates
how expensive a product is expected to be if produced within the existing pro-
cess capabilities. Thus, the cycle time index C corresponding to a product p can
be formulated based on the respective process capability index:
C ¼ l exp
_
1
P
CI
_
¼ l exp
_
3s
T
USL
T
Àm
T
_
ð6Þ
where l is a constant indicating the average dollar cost per variation of process
capabilities. The meaning of l is consistent with that of the dollar loss per devi-
ation constant widely used in Taguchi’s loss functions. It can be determined ex
ante based on the analysis of existing product and process platforms. Such
a cost function produces a relative measure, instead of actual dollar figures,
for evaluating the extent of overall process variations among multiple products.
The estimated cycle time for product p, namely (m
T
, s
T
), is assumed to be a lin-
ear function of the part-worth standard times for the FRs assumed by product
p, i.e.,
m
T
¼

I
i¼1
_
z
i
m
T
i
z
i
þu
_
ð7aÞ
s
T
¼
ffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi

I
i¼1
ðs
T
i
z
i
Þ
2
¸
¸
¸
_
ð7bÞ
where z
i
and u are regression coefficients, m
T
i
and s
T
i
are the mean and the
standard deviation of the part-worth standard time associated with f
i
, respec-
tively, and z
i
is defined the same as in Eq. (4).
3 A-Kano decision-making
The product planning stage features a series of processes including elicitation,
analysis, and fulfillment of the CNs. The A-Kano model can assist decision-
making in the process by prioritizing the FRs according to their impacts on
the customers and producers. A coherent A-Kano process model is developed
to combine the various analytical techniques presented in Section 2, and a gen-
eral roadmap is shown in Figure 4.
3.1 Identification of functional requirements
The A-Kano model requires the survey results of customer’s satisfaction using
the Kano questionnaire. In general, the questionnaire is designed according to
a set of CNs. To allowfor unambiguous understanding, the CNs are translated
into FRs, Fhff
i
ji ¼ i; 2; .; Ig. For example, visibility of the gauges in a car
dashboard is a CN, and is manifested by the size and lighting conditions of
the gauges, namely the FRs. In this research, the FRs are typically related
to customizable product attributes from which the customers can choose
according to his/her preferences.
98 Design Studies Vol 30 No. 1 January 2009
3.2 Division of market segments
Customers are grouped into different market segments based on their demo-
graphic information, such as age, gender, income, etc. If the division of market
segments is not evident, it becomes necessary to carry out market investiga-
tions to differentiate the customer groups. Many methods and tools are avail-
able to assist the process, such as conjoint analysis (Green and DeSarbo,
1978), perceptual mapping (Moore and Pessemier, 1993), and data mining
(Jiao and Zhang, 2005).
3.3 Kano survey
With respect to the FRs, the Kano questionnaire is fabricated and surveys are
conducted to acquire the customers’ assessment of the FRs according to the
functional and dysfunctional forms of questions. In addition, the customers’
immediate perceptions of the importance of the FRs are extracted according
to the self-stated importance (Table 4).
3.4 Computation of Kano indices
For each market segment, the customer’s average satisfaction/dissatisfaction
with respect to the functional/dysfunctional form question is computed ac-
cording to Eq. 1. Next, the Kano indices are computed and plotted in the
A-Kano diagram. Accordingly, the configuration index for each FR is com-
puted, which indicates the priority of the FR in product configuration design.
3.5 Product configuration design and evaluation
Two alternative methods are adopted for product configuration design,
namely (1) the Kano classification-based method, and (2) the configuration
index-based method. In each method, the candidate product configuration is
evaluated against the Kano evaluator. As a prerequisite, part-worth standard
time ðm
T
i
; s
T
i
Þ is identified from the standard routing that constitutes the
Figure 4 A design process
model of analytical Kano
An analytical Kano model 99
process platform. Subsequently, the customer’s satisfaction (U) and cycle time
index (C) of the products are evaluated, leading to an expected value of the
Kano evaluator (E). The optimal product configuration is generated as the
one that produces the largest Kano evaluator.
3.5.1 Kano classification-based configuration design
The threshold values of the Kano classifiers, namely k ¼(r
0
, a
L
, a
H
) are tested
such that the FRs are classified as four categories. Different threshold values
lead to different classification schemes, which in turn influence the design of
product configuration and the expected value of the product. Sensitivity anal-
ysis is carried out to examine different settings of Kano classifiers, using the
Kano evaluator as the performance indicator. Sensitivity analysis entails an it-
erative process of evaluating the Kano evaluator based on specific FR classi-
fication schemes. This research adopts a strategy where r
0
changes from 0.1 to
0.9 with an increment of 0.1, a
L
changes from 5

to 45

with an increment of
5

, and a
H
changes from 50

to 90

with an increment of 5

. A full-factorial
experiment requires 9 Â9 Â9 ¼729 runs of experiment. In each run, the
FRs are classified according to each set of Kano classifier k ¼(r
0
, a
L
, a
H
),
and a set of N product configurations are generated based on Heuristic 1. Sub-
sequently, the customer satisfaction (U) and cycle time index (C) of the set of N
products are evaluated according to Eqs. (4) and (6), leading to an expected
value of the Kano evaluator (E). The optimal values of Kano classifiers are
selected as those that produce the largest Kano evaluator.
3.5.2 Configuration index-based configuration design
Product configuration is generated using Heuristic 2. This is a typical combi-
natorial optimization problem where a large search space needs to be explored.
Genetic algorithms (GAs) have been proven to excel in solving combinatorial
optimization problems in comparison with traditional calculus-based or
approximation optimization techniques. In this regard, a heuristic GA (Jiao
and Zhang, 2005) is adopted to solve such a combinatorial optimization
problem.
4 Case study
The proposed method is applied to car dashboard design in an automobile
company. The dashboard is an important subassembly in the automobile
with a number of customizable components and features (Figure 5). These
components can be represented as FRs in the design stage. The actual selection
of the FRs collectively constitutes the car interior environment, which influ-
ences the driver’s perceptions, and ultimately determines his/her satisfaction.
However, the exact correspondence between FRs and customer’s satisfaction
is not evident. Moreover, the relationship between the selection of FRs and the
producer’s production capability is not considered in product design. There-
fore, it is necessary to gain insight of the FRs with respect to the customer’s
satisfaction and the producer’s capacity.
100 Design Studies Vol 30 No. 1 January 2009
The company has conducted extensive market studies to collect the CNs which
are represented according to well-known FR descriptors. This case study
focuses on a set of 10 CNs, which are refined as 25 FRs, as shown in Table 5.
4.1 Kano survey
A total of 50 car drivers constituted the Kano survey respondent set. The
respondents were divided into three groups based on their age, gender and
income levels, representing three market segments as shown in Table 6.
Each respondent was required to answer the Kano questions with respect to
each and every FR, and give the perception of the importance of an FR using
the self-stated importance (Table 4). The form of Kano questions used in this
survey is shown in Table 1, including both the functional and dysfunctional
forms.
Upper instrument
panel
Gauge display Air bag (inside)
Storage
compartment
Radio/CD control
Lower instrument
panel
Steering wheel
column
Vent
Figure 5 Main components
and features of car dashboard
subassembly
Table 5 Functional requirements of car dashboard
Customer need Functional requirement Code
Gauge display type Analog display f
1e1
Digital display f
1e2
Gauge display size Large (15 inch) f
2e1
Small (10 inch) f
2e2
Gauge display lighting Back light f
3e1
External light f
3e2
Instrument panel material Plasticewooden f
4e1
Plastic f
4e2
Plasticemetal f
4e3
Instrument panel color Cream f
5e1
Gray f
5e2
Black f
5e3
Steering wheel material Plasticewooden f
6e1
Plastic f
6e2
Plasticeleather f
6e3
Steering wheel height High (24 inch) f
7e1
Low (20 inch) f
7e2
Position of CD/radio control Integrated in steering wheel f
8e1
On control panel f
8e2
Storage configuration Top f
9e1
Side f
9e2
Front f
9e3
Passenger side f
9e4
Air bag Front f
10e1
Side f
10e2
An analytical Kano model 101
Next, for each FR, the average level of satisfaction/dissatisfaction for the func-
tional/dysfunctional form question within each market segment is computed,
resulting in the data point ðX
i
; Y
i
Þ as shown in Figure 6. Accordingly, the
Kano indices r
.
i
hðr
i
; a
i
Þ are derived, which is further used to compute the
configuration index. The configuration index with respect to three market seg-
ments is summarized in Table 7.
Table 7 also shows the part-worth standard times of the FRs. The company
fulfills customer orders through assembly-to-order production while import-
ing all components and parts via global sourcing. With assembly-to-order pro-
duction, the company has identified and established standard routings as basic
constructs of its process platform. The part-worth standard time of each FR
is established based on time and motion studies of the related assembly and
testing operations.
4.2 Comparative study
A comparative study is carried out to validate the effectiveness of the A-Kano
method. The study involves three independent procedures to design products
based on the same set of survey data. In the first procedure, the FRs are
Table 6 Customer groups in Kano survey
Market segment Age Gender Income (10
3
S$/year)
Segment 1 (s
1
) 46þ M/F 120þ
Segment 2 (s
2
) 31e45 M/F 60e119
Segment 3 (s
3
) 21e30 M 30e59
Figure 6 Kano model of
functional requirement
classification
102 Design Studies Vol 30 No. 1 January 2009
classified using the Kano method proposed by DuMouchel (Berger et al.,
1993). In the second procedure, the respective FRs are classified using the
Kano classifiers proposed in the A-Kano method. In both methods, the FR
classification is applied to assist the generation of product configurations. In
the third procedure, the configuration indices of the FRs are used to guide
the design of product configurations following the decision framework pre-
sented in Section 3. The merits of the three methods are compared according
to the product performance in terms of the Kano evaluator. In this case study,
the results of only one market segment (s
1
) are presented for purpose of brev-
ity. The analysis of the other market segments can be carried out following the
same procedures.
4.2.1 Kano-based product configuration design
Based on the method proposed by DuMouchel (Berger et al., 1993), a two-
dimensional Kano diagram is established according to customers’ satisfaction
and dissatisfaction levels, as shown in Figure 6. According to this model, the
diagram is divided into four quadrants, and accordingly the FRs are classified
into four categories. Heuristic 1 is adopted to generate product configurations
based on the Kano classification. The product configuration, called Product I,
is generated manually and is shown in Table 8, where a number ‘1’ indicates
Table 7 Configuration index and part-worth standard time
FR Configuration index (market segment) Part-worth standard time
s
1
s
2
s
3
m
T
i
(s) s
T
i
(s)
f
1e1
0.30 0.2 0.31 498 21
f
1e2
0.38 0.33 0.2 557 35
f
2e1
0.55 0.3 0.56 505 22
f
2e2
0.36 0.33 0.3 493 20
f
3e1
0.42 0.35 0.1 90 6
f
3e2
0.54 0.74 0.6 50.5 4
f
4e1
0.48 0.44 0.3 462 15.5
f
4e2
0.46 0.38 0.42 465 18
f
4e3
0.32 0.4 0.35 411 13
f
5e1
0.51 0.22 0.61 50 5.5
f
5e2
0.38 0.56 0.29 47.5 5.5
f
5e3
0.66 0.55 0.69 45 6
f
6e1
0.54 0.46 0.53 442 28
f
6e2
0.53 0.48 0.45 365 30.4
f
6e3
0.74 0.7 0.63 483 32
f
7e1
0.60 0.43 0.52 365 30.4
f
7e2
0.53 0.23 0.35 483 32
f
8e1
0.26 0.19 0.08 99 9.4
f
8e2
0.14 0.09 0.08 99 9.4
f
9e1
0.39 0.55 0.47 69 8
f
9e2
0.66 0.69 0.74 54 5.6
f
9e3
0.57 0.49 0.64 80.3 5.6
f
9e4
0.55 0.56 0.49 40.6 4
f
10e1
0.85 0.55 0.74 216 33
f
10e2
0.67 0.59 0.95 247 40
An analytical Kano model 103
that the respective FR is included in the product; otherwise, it is left blank. The
Kano evaluator is computed using Eqs. (3)e(7), such that Product I results in
a Kano evaluator of 0.48.
4.2.2 A-Kano-based product configuration
design e Kano classifiers
The Kano classification is implemented based on the Kano indices and Kano
classifiers. To select the appropriate values of Kano classifiers, an iterative pro-
cess of sensitivity analysis is carried out. Each set of Kano classifiers, i.e.,
k ¼(r
0
, a
L
, a
H
) results in a scheme of FR classification, which in turn guides
the generation of the product configurations following Heuristic 1. Once the
product configurations are created, the Kano evaluator is computed.
The sensitivity analysis involves three variables, namely r
0
, a
L
and a
H
. Hence,
it is difficult to visualize the response surface using a three-dimensional graph.
A graphical illustration can be made by setting one variable at a fixed value
while changing the values of the other two variables. For example, Figure 7
Table 8 Product configurations generated based on three methods
FR Product I
(Kano-based)
Product II
(A-Kano: classifier)
Product III
(A-Kano: configuration index)
f
1e1
1
f
1e2
1 1
f
2e1
1 1 1
f
2e2
f
3e1
1
f
3e2
1 1 1
f
4e1
f
4e2
1
f
4e3
1 1
f
5e1
1 1 1
f
5e2
1
f
5e3
1 1
f
6e1
1
f
6e2
1 1
f
6e3
f
7e1
1
f
7e2
1 1
f
8e1
1
f
8e2
1 1
f
9e1
1
f
9e2
1 1 1
f
9e3
1 1 1
f
9e4
1
f
10e1
1 1 1
f
10e2
1 1 1
Customer satisfaction 7.03 7.33 7.71
Cycle time index 14.63 7.23 6.63
Kano evaluator 0.48 1.01 1.16
104 Design Studies Vol 30 No. 1 January 2009
shows the result of sensitivity analysis when the Kano classifier r
0
is fixed as
r
0
¼0.3, a
L
varies from 5

to 45

, and a
H
varies from 50

to 90

. The Kano
evaluator is normalized for purpose of illustration. Base on the sensitivity
analysis, the largest Kano evaluator is 1.01, which is obtained when the
Kano classifiers are r
0
¼0.5, a
L
¼25

, and a
H
¼65

, respectively. The A-
Kano diagram for FRs classification is shown in Figure 8, and the product
configuration derived from this classification is called Product II, which is
shown in Table 8.
4.2.3 A-Kano-based product configuration
design e configuration index
The A-Kano method proposes using the category index to guide the genera-
tion of product configurations. A heuristic GA is adopted to generate a set
of solutions, each of which is evaluated against the Kano evaluator. The prod-
uct configuration (Product III) derived from this method is shown in Table 8,
which leads to a Kano evaluator of 1.16.
4.3 Results and analysis
The major difference between the three methods lies in the criteria to select the
FRs that constitute the product configuration. In the traditional Kano
method, FRs are selected manually based on the Kano category. Although
the process is straightforward, it does not reflect the differences among FRs
within the same category. For example, both f
2e2
and f
8e2
are classified as in-
different FRs. However, these two FRs represent different customer percep-
tions considering their importance and relative customer satisfaction. Such
a difference is effectively captured by the configuration index in the A-Kano
method, where the configuration indices are 0.36 and 0.14, respectively.
Interestingly, the polar form A-Kano classification assigns two FRs into
one-dimensional and indifferent categories, respectively, which reflects the
difference between them.
Figure 7 Sensitivity analysis
of Kano evaluator w.r.t. a
L
and a
H
An analytical Kano model 105
On the other hand, both the traditional Kano method and A-Kano method
based on Kano classifiers suffer the discontinuity problem, i.e., they make
clear-cut distinctions between FRs that fall into different categories, even
when these FRs are actually located close to each other in the Kano diagram.
For example, FRs f
2e2
and f
5e2
which are located closely to each other are
classified as indifferent and attractive FRs in the traditional Kano. However,
such a distinction is not evident based on the Kano indices. A slight change of
the classification threshold value, e.g., moving the horizontal dashed line
higher, would make both f
2e2
and f
5e2
fall into the indifferent category. Hence,
it is subjective to assign the threshold values. On the other hand, in the
A-Kano method, the configuration indices of these two FRs are 0.36 and
0.38, thus indicating identical customer preferences. Interestingly, the classifi-
cation using A-Kano classifiers reflects such similarities so that the two FRs
both fall in the one-dimensional region.
A further comparison can be made between the product configurations re-
sulted from the traditional Kano model and the A-Kano (Table 8). The merit
of the product configurations is evaluated against the Kano evaluator. As can
be seen from the results, the product configurations lead to different custom-
er’s satisfaction and cycle time index, and in turn the Kano evaluator. The
overall performance of the Product I is 0.48, while those of Products II and
III are 1.01 and 1.16, respectively. The latter two products offer larger expected
shared surplus values, which indicate better strategies to leverage upon
customer’s satisfaction and the producer’s capacity. As can be seen from the
results, the overall customer satisfaction values are 7.03, 7.33 and 7.71 for
the three products, which are almost identical. The major difference in the
Figure 8 A-Kano model of
functional requirement
classification
106 Design Studies Vol 30 No. 1 January 2009
Kano evaluator lies in the varied cycle time indices, which are 14.63, 7.23, and
6.63, respectively. Such a distinction reflects that Products II and III effectively
account for the producer’s capacity in the configuration design. It should be
noted the absolute value of the Kano evaluator is not significant. But rather,
it denotes the relative superiority of a classification in terms of the shared
surplus.
5 Discussions
The A-Kano method extends traditional Kano method by adopting quantita-
tive measures of customer’s satisfaction. The Kano classifiers and configura-
tion indices enhance decision support in product design. Furthermore, it
proposes a performance indicator that gives dual considerations of customer’s
satisfaction and producer’s capacity to fulfill customer needs. The main contri-
bution of this research is to extract useful customer need information from
Kano survey for decision support in product design. In comparison with tra-
ditional Kano model that relies on qualitative Kano categories, the A-Kano
method establishes a coherent decision framework that deals with the interac-
tions between the customer domain and the producer domain. The polar form
of Kano indices r
.
i
hðr
i
; a
i
Þ excels the Cartesian form ðX
i
; Y
i
Þ proposed by
DuMouchel (Berger et al., 1993) as a quantification measurement of customer
preferences, in the sense that it facilitates intuitive understanding of the FRs.
In particular, the magnitude of the vector, r
i
, represents the importance of an
FR and the angle a
i
represents the relative level of satisfaction and dissatisfac-
tion, thus representing different aspects of customer perceptions.
The quantification of customer’s satisfaction based on Kano indices distin-
guishes the A-Kano method from other research techniques, such as conjoint
analysis (Green and DeSarbo, 1978), stated choice methods (Louviere et al.,
2000), and discrete choice analysis (Wassenaar and Chen, 2003; Wassenaar
et al., 2005). The major difference lies in the different denotations of customer
preference. In classical conjoint analysis as well as the stated choice methods,
customer preference/demand is modeled as a utility function of various prod-
uct attributes. In discrete choice analysis, the economic benefits to the
producer are modeled as a utility function of customer attributes, prices
and socioeconomic and demographic background of the market population
(Wassenaar and Chen, 2003; Wassenaar et al., 2005). In contrast, the A-
Kano method models customer preference based on customer’s satisfaction
and dissatisfaction. Apparently, the meanings of satisfaction and dissatisfac-
tion in the A-Kano model are different from those used in the utility theory.
For example, the questions ‘How satisfied are you with this product?’ and
‘How useful/valuable is this product to you?’ are in fact quite different. An
outdated product may be very unsatisfying, but very useful because one
may rely on it to carry out important work. In such a respect, the Kano
model implies two different perspectives e proper functioning of a product
feature arouses customer satisfaction, while the dysfunction of a product
An analytical Kano model 107
feature generates dissatisfaction, as shown in Figure 9. This is a useful exten-
sion to the one-dimensional utility measurement. However, this research does
not assert that the Kano’s perspective is more reasonable than that used in
the utility theory. Nevertheless, it provides an alternative measurement of
customer preferences.
6 Conclusions
Product development for customer’s satisfaction with considerations of pro-
ducer’s capacity necessitates logical prioritization of customer needs. A few
major challenges remain unaddressed for customer need analysis, such as
quantitative measure of customer satisfaction, decision support, and capac-
ity assessment of the producers. This paper presents an analytical Kano
model for customer need analysis, following the basic principles of tradi-
tional Kano model, while consolidating the theoretical foundation. The
A-Kano model adopts the Kano indices for measuring customer’s satisfac-
tion and dissatisfaction. By adopting a polar form representation scheme,
the Kano indices are conducive to exploring the nature of customer satisfac-
tion. In accordance with the Kano indices, the Kano classifiers are deemed
to outperform traditional classification criteria based on the logical polar
form. As an alternative decision process, the configuration index is defined
to prioritize the functional requirements so as to provide decision support
to product configuration design. The merit of the designed product is eval-
uated against the Kano evaluator, leveraging upon both customer’s satisfac-
tion and the producer’s capacity. The configuration index in combination
with the Kano evaluator is deemed to be useful extensions to traditional
Kano where the classification criteria are subjective and unjustifiable. The
A-Kano method defines systematic procedures to elicit the customer needs,
and to conduct surveys for customer need analysis. The overall A-Kano
framework address the customer need analysis from a broader scenario,
such that the customizable product features can be better managed at the
product planning stage.
Figure 9 Comparison of the
utility theory and Kano’s
theory
108 Design Studies Vol 30 No. 1 January 2009
Acknowledgment
This research is partially sponsored by the European Commission (DG Infor-
mation Society and Media) in the framework of the 6th Research Program
(FP6) under grant IST-5-035030-STREP (www.cater-ist.org). Also acknowl-
edged is support from Singapore Agency for Science, Technology and
Research (A*STAR) Science and Engineering Research Council (SERC)
Thematic Strategic Research Programme grant on human factors engineering
(#062 131 0066).
References
Berger, C, Blauth, R, Boger, D, Bolster, C, Burchill, G, DuMouchel, W, Pouliot, F,
Richter, R, Rubinoff, A, Shen, D, Timko, M and Walden, D (1993) Kano’s method
for understanding customer-defined quality, Center for Quality of Management
Journal Vol 2 No 4 pp 3e35
Green, P E and DeSarbo, W S (1978) Additive decomposition of perceptions data
via conjoint analysis, Journal of Consumer Research Vol 5 No 1 pp 58e65
Hazelrigg, G A (1998) A framework for decision based engineering design, Journal
of Mechanical Design Vol 120 pp 653e658
Jiao, J and Chen, C H (2006) Customer requirement management in product
development: a review of research issues, Concurrent Engineering: Research and
Applications Vol 14 No 3 pp 173e185
Jiao, J and Tseng, M M (1999) A pragmatic approach to product costing based on
standard time estimation, International Journal of Operations & Production Man-
agement Vol 19 No 7 pp 738e755
Jiao, J and Zhang, Y (2005) Product portfolio planning with customereengineer-
ing interaction, IIE Transactions Vol 37 No 9 pp 801e814
Jiao, J, Ma, Q and Tseng, M M (2003a) Towards high value-added products and
services: mass customization and beyond, Technovation Vol 23 pp 809e821
Jiao, J, Zhang, L and Pokharel, S (2003b) Process platform planning for mass
customization in Proceedings of the Second Interdisciplinary World Congress
on Mass Customization and Personalization, Technical University, Munich
CD-ROM
Kano, N, Seraku, N, Takahashi, Fand Tsuji, S(1984) Attractive quality and must-be
quality, The Journal of the Japanese Society for Quality Control 14 No 2 pp 39e48
Lai, X, Xie, M and Tan, K C (2004). Optimizing product design using the Kano
model and QFD, in 2004 IEEE International Engineering Management Confer-
ence, 18e21 Oct., pp 1085e1089
Louviere, J, Hensher, D and Swait, J (2000) Stated choice methods, analysis and
application Cambridge University Press, Cambridge, UK
McKay, A, de Pennington, A and Baxter, J (2001) Requirements management:
a representation scheme for product, Computer Aided Design Vol 33 No 7 pp
511e520
Matzler, K and Hinterhuber, H H (1998) How to make product development
projects more successful by integrating Kano’s model of customer satisfaction
into quality function deployment, Technovation Vol 8 No 1 pp 25e38
Moore, W L and Pessemier, E A (1993) Product planning management: designing
and delivering value McGraw-Hill, New York
Rivie` re, P, Monrozier, R, Rogeaux, M, Page` s, J and Saporta, G (2006) Adaptive
preference target: contribution of Kano’s model of satisfaction for an optimized
An analytical Kano model 109
preference analysis using a sequential consumer test, Food Quality and Preference
Vol 17 pp 572e581
Sireli, Y, Kauffmann, P and Ozan, E (2007) Integration of Kano’s model into
QFD for multiple product design, IEEE Transactions on Engineering Management
Vol 54 No 2 pp 380e390
Suh, N P (2001) Axiomatic design-advances and applications Oxford University
Press, New York
Tan, K C and Shen, X X (2000) Integrating Kano’s model in the planning ma-
trix of quality function deployment, Total Quality Management Vol 11 No 8
pp 1141e1151
Tielemans, P F J (1995) Lead time performance in manufacturing systems, PhD
thesis, Eburon, Delft.
Wassenaar, H J and Chen, W (2003) An approach to decision-based design with
discrete choice analysis for demand modeling, ASME Journal of Mechanical
Design Vol 125 pp 490e497
Wassenaar, H J, Chen, W, Cheng, J and Sudjianto, A (2005) Enhancing discrete
choice demand modeling for decision-based design, ASME Journal of Mechanical
Design Vol 127 pp 514e523
110 Design Studies Vol 30 No. 1 January 2009

and (3) requirement classification. Among many approaches that address customer need analysis, the Kano model has been widely practiced in industries as an effective tool of understanding customer preferences owing to its convenience in classifying CNs based on survey data (Kano et al., 1984). Nevertheless, traditional Kano methods are not equipped with quantitative assessment. And, Kano classification provides limited decision support in engineering design. Moreover, it inherently emphasizes the customer and market perspectives only, with limited consideration of the producer’s capacity to fulfill the CNs.

To enhance the above aspects in relation to customer need analysis, this paper scrutinizes the theoretical foundation of the original Kano model and proposes an analytical Kano (A-Kano) model. The A-Kano model extends traditional Kano model (Kano et al., 1984; Berger et al., 1993) by introducing (1) Kano indices, which are quantitative measurements of customer satisfaction derived from Kano questionnaires and surveys; (2) Kano classifiers, which consist of a set of criteria to classify CNs based on the Kano indices; (3) Configuration index, which provides a decision factor for selecting the functional requirements (FRs) that contribute to product configurations; and (4) Kano evaluator, which is a shared surplus-based performance indicator leveraging upon both the customer’s satisfaction and the producer’s capacity. A comprehensive process model is proposed to integrate these techniques for customer need analysis.

The subsequent sections proceed as follows. A critical review of the Kano model is presented in Section 1. The fundamental issues of A-Kano are scrutinized in Section 2. The A-Kano process model is proposed in Section 3 by implementing the proposed techniques that address the fundamental issues. In Section 4, the A-Kano model is applied to car dashboard design. The results are compared with those derived from the traditional Kano model. Discussions and conclusions are presented in Sections 5 and 6, respectively.

1

Critical review of Kano model

The Kano model of customer satisfaction is a useful tool to classify and prioritize customer needs based on how they affect customer’s satisfaction (Kano et al., 1984). It captures the nonlinear relationship between product performance and customer satisfaction. In practice, four types of product attributes are identified: (1) must-be attributes are expected by the customers and they lead to extreme customer dissatisfaction if they are absent or poorly satisfied, (2) one-dimensional attributes are those for which better fulfillment leads to linear increment of customer satisfaction, (3) attractive attributes are usually unexpected by the customers and can result in great satisfaction if they are available, and (4) indifferent attributes are those that the customer is not interested in the level of their performance.

88

Design Studies Vol 30 No. 1 January 2009

. stated choice methods (Louviere et al. the application of Kano model in engineering design is primitive in comparison with other quantitative methods. and customer satisfaction coefficient (Berger et al. Riviere et al.3 Decision support The ultimate goal of customer need analysis is to provide decision support to product design. 2006). 1993). the categories of the Kano model signify different priorities in An analytical Kano model 89 . 1998). 2000).. which could not precisely reflect the extent to which the customers are satisfied (Berger et al.1 Qualitative vs. 2005. The final classification of a product attribute is made based on a statistical analysis of the survey results of all respondents. In such a sense.. whilst a negative number is used to reflect the relative cost of not meeting this customer requirement. and decision-based design (Hazelrigg. which captures the customers’ response if a product has a certain attribute. Each customer requirement can be represented as a pair of satisfaction and dissatisfaction values. which captures the customers’ response if the product does not have that attribute. the resulting Kano category is still qualitative in nature. In particular. classification criteria are not explicitly defined in this model. 1998). The nature of a customer requirement can be delineated by the quadrant into which that point falls. revealing an individual customer’s perception of a product attribute.The Kano model is constructed through customer surveys. The questionnaire is deployed to a number of customers. However. However. DuMouchel (Berger et al. such as empirical observation. Hence.. it only allows qual` itative assessment of product attributes (Wassenaar et al. such as conjoint analysis (Green and DeSarbo. 1993). quantitative evaluation The Kano diagram provides a rough sketch of the customer’s satisfaction in relation to the product performance level. 1993). 1. In general. where a customer questionnaire contains a set of question pairs for each and every product attribute. 1978). The question pair includes a functional form question. mode statistics. it is very subjective to classify the customer requirements into four quadrants because of a lack of logical classification criteria. and a dysfunctional form question. they fall short as concrete decision criteria. 1993) proposes a graphical Kano diagram that is based on predefined scales related to the customer’s satisfaction and dissatisfaction. A convenient way to incorporate quantitative measures is to assign some scales in terms of the levels of customer satisfaction/dissatisfaction (Matzler and Hinterhuber. However. a positive number is used to represent the relative value of meeting the respective customer requirement... 1.. Although the Kano categories may enhance designers’ understanding of customer needs.2 Classification criteria Different criteria have been adopted for classifying customer requirements. 1993) proposes a twodimensional representation of Kano quality category based on the customer satisfaction coefficients. Timko (Berger et al.. 1. and each answer pair is aligned with the Kano evaluation table (Berger et al.

such that the product will only include those features that are affordable to the producers (Matzler and Hinterhuber. such an assumption seldom holds true.e. what a customer de facto perceives is the CNs in the customer domain. 2007). without accurate cost estimation. the producer must seek for an economy of scale in product fulfillment. namely the FRs. which represent the physical form of a product in fulfillment of the FRs. As a quick fix. and including some attractive FRs (Berger et al. The distinction between CNs and FRs is in line with the domain mapping principle proposed by Suh (2001). the priority assignment corresponds to different customer perceptions.. a designer is directed toward fulfilling all must-be FRs. Cost constraints are usually accommodated by the product development team based on expertise. it focuses exclusively on addressing the concerns of customers (Sireli et al. Customer perception can be explicitly measured as the overall customer satisfaction with respect to a combination of product attributes. From an engineering perspective. the legacy producer’s capacity has to be assessed in terms of delivery of the 90 Design Studies Vol 30 No.. (2004) propose a linear cost function to constrain the selection of product features in product development. engineers have to seek for a leverage of the customer-perceived value and the producer’s capacity. Lai et al. analysis is carried out in the functional domain. 2006). Essentially. Tan and Shen (2000) propose an approximation transformation function based on Kano analysis to adjust the improvement ratio of customer attributes. 1998). 1.. From the customer’s perspective. 2 A-Kano model The A-Kano model entails a series of interactions between the customers and the producers. these cost models are primitive and are inadequate to reflect the complexities of design and manufacturing costs. While providing customer-perceived diversity in CNs. and the FRs are assigned different priorities for product fulfillment. And hence it is difficult to apply analytical tools for CN analysis. which is meant by FRs. the CNs are translated into explicit and objective statements. As a decision-making tool used by engineers. the Kano classification is deemed to be inadequate to facilitate decisions in product design. From the producer’s perspective.4 Producer’s capacity The Kano model is inherently customer-driven. i. the FRs are mapped onto various product attributes. In this research. In practice. In general. 1 January 2009 . However. However.designing products. such approaches do not distinguish FRs within the same category. the CNs tend to be imprecise and ambiguous due to their linguistic origins (Jiao and Chen. as shown in Figure 1. staying competitive with market leaders on the one-dimensional FRs. For example. the Kano model fails to account for the producer’s concerns in terms of the capacity to fulfill the CNs. 1993). On the other hand. Hence. rather than FRs in the functional domain.

the self-stated importance score is normalized such that it falls within a range of 0e1. 2.. . shftj jj ¼ 1. 2.wij).. and the average level An analytical Kano model 91 . either by expanding the producer’s capacity to meet the CNs or by directing the customers to the total capacity of the producer so that customers are better served (Jiao et al. IÞ is represented as eij ¼ (xij. IÞ according to the functional and dysfunctional forms of Kano questions (Table 1).. Furthermore. the scaling has the effect of diminishing the influence of negative evaluations (Berger et al.. Hence. Surveys are carried out to collect the respondents’ evaluation of fi ðci ¼ 1.1 Kano indices for quantification of customer satisfaction Kano survey is carried out within specific market segments that consist of customers with similar demographic information. 2003a). where xij is the score given to an FR for the dysfunctional form question.yij. Ig. for each FR (fi). 1993). Ultimately the business success is achieved by maximizing such an overlap.. . and wij is the self-stated importance. For each respondent tj ˛sðcj ¼ 1. the average level of satisfaction for the dysfunctional form question within market segment s is defined as Xi .. 2. i. .. Jg. 2. as shown in Table 4. the evaluation of fi ðci ¼ 1. this research adopts a scoring scheme that defines customer’s satisfaction and dissatisfaction as shown in Table 3. . yij is the score given to an FR for the functional form question..Figure 1 Customereproducer interactions along the product value chain desired product attributes. 2. The scale is designed to be asymmetric because positive answers are considered to be stronger responses than negative ones.e. The value chain converges when the customer’s satisfaction overlaps with the legacy producer’s capacity. A set of FRs is identified as Fhffji i ¼ 1. which is the respondent’s perception of the importance of an FR. Next. JÞ. Let s denote the market segment which contains a total of J customers (respondents). The preliminary category of the FR is determined using the Kano evaluation table (Table 2).. 2. 1993). . Similar to the method proposed by DuMouchel (Berger et al.

M. Q. 92 Design Studies Vol 30 No. Xi ¼ J 1X wij xij .g.. One-dimensional. R.. the magnitude Table 2 Kano evaluation table Dysfunctional form of the question Like Functional form of the question Like Must-be Neutral Live with Dislike Q R R R R Must-be A I I I R Neutral A I I I R Live with A I I I R Dislike O M M M Q A. O.e. and a Reverse category can be transformed out of the category by reversing the sense of functional and dysfunctional of questions (Berger et al. the characteristics of an FR (fi) can be repqffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi . The rationale of representing the satisfaction and dissatisfaction . how do you feel?) of satisfaction for the functional form question within the same market segment is defined as Yi . and ai ¼ tanÀ1 ðYi =Xi Þ is the angle between . i. A questionable category will not be included in the averages.I can live with it that way . i. as a vector ri is that it becomes equivalent to a polar form. J j¼1 Yi ¼ J 1X wij yij J j¼1 ð1Þ The value pair ðXi . as shown in Figure 2.I dislike it that way Dysfunctional form of the question (e. Must-be.It must be that way . . Yi Þ can be plotted in a two-dimensional diagram. Yi Þ should fall in the range of 0e1 because the negative values are results of either Questionable or Reverse categories. Accordingly. Questionable.I dislike it that way .I like it that way .g.I am neutral . fi wri hðri . Most ðXi . Reverse. Attractive.e. ai Þ. and the horiri zontal axis. 1993). From the customer’s perspective. 1 January 2009 . resented as a vector. I. Indifferent. where ri ¼ jri j ¼ X2 þ Y2 is the i i ri magnitude of ..I like it that way . the classification of an FR can be defined based on the corresponding location of the value pair in the diagram. how do you feel?) Answer . if the car does not have air bags.I am neutral .. i..I can live with it that way . if the car has air bags.It must be that way . where the horizontal axis indicates the dissatisfaction score and the vertical axis stands for the satisfaction score.e.Table 1 Kano questionnaire Kano question Functional form of the question (e..

Table 3 Scores for functional/dysfunctional features Answers to the Kano question I like it that way (like) It must be that way (must-be) I am neutral (neutral) I can live with it that way (live with) I dislike it that way (dislike) Functional form of the question 1 0.5 Dysfunctional form of the question À0. the FRs can be classified into four categories. i. A higher threshold value of the satisfaction index is defined as aH. 2.7 0.6 0.5 0 À0. attractive and one-dimensional.25 À0. must-be.25 0 0. and thus called an indifferent FR.9 1. The region of the must-be FRs corresponds to the sector DEFG. and hence it is an ideal must-be element. it is considered as a must-be FR. Therefore.5 0. as shown in Figure 3. Table 4 Scores for self-stated importance Not important Somewhat important Important Very important Extremely important 0. In 0 ri the extreme situation. The region defined by the sector OFI in Figure 3.4 0. fi is considered as unimportant.2 Kano classifiers for categorization of customer needs Based on the above formulation. while functioning of fi does not enhance satisfaction.0 An analytical Kano model 93 . and the angle (ai) is called the satisfaction index. is used to differentiate important FRs from less important ones. if ri > r0 and ai > aH.3 0. and the angle ai determines the relative level of satisfaction and dissatisfaction. Hence r0 is called an indifference threshold. If ri r0.2 0. where the radius is smaller than r0. a lower threshold value of the satisfaction index is defined as aL. such that for fi. ai ¼ p=2 means that fi is an ideal attractive element. ai ¼ 0 means that dysfunction of fi causes dissatisfaction. is considered as the indifferent region. The region of the attractive FRs is shown as sector ABHI. such that for fi.e. Conversely.1 0. indifferent. A threshold value of the importance index. Both pffiffiffi 2 and 0 ai p=2 are collectively called the Kano indices. Likewise.8 0. the magnitude of the vector (ri) is called the importance index.5 À0. r0.5 1 of the vector denotes the overall importance of fi to the customers belonging to segment s. if ri > r0 and ai aL.. it is considered as an attractive FR.

denoted as k ¼ (r0.Figure 2 Vector representation of customer perception on a Kano diagram If ri > r0 and aL < ai aH. the classification of FRs provides a decision criterion for selecting the FRs that constitute a product configuration. aL. and aH are collectively called Kano classifiers. this research proposes the following heuristic to simulate the consequence of Kano classification on product configuration design. aH). According to the Kano principles. aL. The region of the one-dimensional FRs is shown as sector BCDGH. The set of thresholds r0. fi is considered as a one-dimensional FR. 1 January 2009 . To do so. Figure 3 Kano classifier and Kano categories 94 Design Studies Vol 30 No.

To alleviate such a problem. IÞ. such that & zi ¼ 1. zi. At the same time.1) generates a random number between 0 and 1. if fi is an indifferent FR. ai) to indicate the probability that an FR is contained in a production configuration. ¼ 0:8. i. it is deemed to be difficult to define universal thresholds for different products..4. 0. as is discussed in Section 2. zi zi zi zi ¼ 1. if fi is a one-dimensional FR. 1Þ. the configuration index ri is proportional to the importance index ri. This research uses sensitivity analysis to select and testify the appropriate settings of the Kano classifiers. and zi denotes the probability that fi be included in the product configuration. data points located near the borders of two adjacent regions may be classified as different categories. 1Þ. Moreover. pffiffiffi 2 2 ai  1À ri ri ¼ 3 p ð2Þ Given a particular ai. The purpose of this strategy is to provide better decision support to product configuration design. for a specific value An analytical Kano model 95 . if fi is an attractive FR. a proper performance indicator must be defined to justify the significance of the classification with respect to different settings of Kano classifiers. A binary variable.3 Configuration index for product configuration design Decision-making based on the Kano classification inevitably suffers the discontinuity problem..e. The probability zi is dependent on the Kano category that an FR belongs to. .. In practice. ci ¼ 1.e. which agrees with the observation that an FR with greater influence on customer’s satisfaction/dissatisfaction is more likely to be included in the product configuration.Heuristic 1: A product configuration (p) is generated as a combination of FRs ðfi . is used to indicate whether fi is contained in the product or not. ¼ 0:2. if fi is a must-be FR. The configuration index ðri Þ is defined as a function of Kano indices (ri. otherwise: where the Random(0. Determining appropriate values of Kano classifiers is challenging in that these threshold values may be problem-specific and context-aware for different applications. 2. i. 2. ¼ Randomð0. this research proposes a configuration index to indicate the priority of an FR in fulfillment of customer expectations. while their distinction is actually minor. if zi ! Randomð0.

when ri ! Randomð0. IÞ. The following heuristic is used to simulate the effect of configuration index on product design. 1Þ.4. one-dimensional and attractive.4 Kano evaluator for leveraging customer’s satisfaction and producer’s capacity The Kano classifiers and configuration indices provide two alternative mechanisms to determine the FRs to be included in a product. U C E¼ ð3Þ 96 Design Studies Vol 30 No. Variable zi ¼ 1. zi. a product configuration p is represented as a combination of FRs ðfi .e. otherwise: The value of zi is assigned in the product design stage based on a stochastic procedure. ri decreases with an increase of the satisfaction index ai. 1Þ. increasing customer’s satisfaction and enhancing producer’s capacity. and the producer surplus is simulated as the overall cycle time index (C). 2. if fi is contained in product p. 0. 2. product development involves two interrelated perspectives.1 Objective function This research adopts the concept of shared surplus to leverage upon both customer’s and producer’s concerns. Accordingly. In this respect. is used to indicate whether fi is contained in the product or not. . For example. and zi ¼ 0. which reflects the decreasing priorities associated with the Kano categories in the order of mustbe.. The selection of FRs that contribute to a product configuration is contingent on the configuration indices. Let Random(0. It should be noted that product configuration design is subject to other configuration constraints.. the Kano evaluator (E) is defined as a shared surplus-based performance indicator. i.of ri. ci ¼ 1. the producers have to strive to design products at affordable costs. 2. The consumer surplus is modeled as the overall customer’s satisfaction (U) of the product. Heuristic 2: A binary variable. Based on this heuristic. two FRs may be incompatible with each other such that only one of them may be selected in the configuration.e. an FR with a larger configuration index has a higher chance to be included in the product. & zi ¼ 1. In particular.. 1 January 2009 . when ri < Randomð0. While the resulting product configuration reflects the customer’s preferences. Therefore. i.1) be a function that generates a random number between 0 and 1. the A-Kano model explicitly defines a Kano evaluator to estimate the value of the planned products.

zi ¼ 0.2 Cycle time index The cycle time index (C) is based on the concept of process capability index which provides an economic latitude of process variations due to product customization (Jiao and Tseng. producer’s capacity is related to the technical feasibility of a design and the engineering cost. it is necessary to find the mapping relationships between different attribute levels and their expected consumptions of standard times within legacy process capabilities. Any product configured from available attribute levels is justified based on its expected cycle time.Based on the customer’s responses to the functional/dysfunctional form of Kano questions. I X r i zi ð4Þ U¼ i¼1 where & 1. 1999). in which standard routings can be reconfigured to accommodate various products derived from the corresponding product platform (Jiao et al.e. and sT. The USLT can be determined ex ante based on the worst case analysis of a given process platform.4. and the standard deviation of the estimated cycle time. which demonstrates the distinctions between variables that differ as a result of random error and are often well described by a normal distribution (Tielemans. reflecting the compound effect of multiple products on production in terms of process variations. These part-worth standard time accounting relationships are built into the product and process platforms (Jiao et al. the customer’s satisfaction (U) is defined as the summation of the configuration indices ðri Þ of FRs contained in the product configuration.. 2003b). if fi is contained in product p. otherwise On the other hand. and sT are the upper specification limit. the configuration index reflects the customer’s satisfaction level for individual FRs. An analytical Kano model 97 . The process capability index gives an indication of how expensive of a customization is to be if implemented in production... 2003b). 2. 1995). which refers to the producer surplus. The idea is to allocate costs to those established time standards from previously reported work and time studies and thereby circumvent the tedious tasks of identifying various cost drivers and cost related activities. Hence. The one-side specification limit process capability index (PCI) can be formulated as: PCI ¼ USLT À mT 3sT ð5Þ where USLT. the mean. This research proposes the cycle time index (C) to estimate the producer’s capacity as discussed next. Variations in the cycle time are characterized by mT. i. respectively. mT. To do so.

mT ¼ I XÀ i¼1 zi mT zi þ u i Á ð7aÞ vffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi u I uX 2 sT ¼ t ðsT zi Þ i i¼1 ð7bÞ where zi and u are regression coefficients. sT).. In this research. The A-Kano model can assist decisionmaking in the process by prioritizing the FRs according to their impacts on the customers and producers. and a general roadmap is shown in Figure 4. the cycle time index C corresponding to a product p can be formulated based on the respective process capability index: 1 C ¼ l exp CI P   3sT ¼ l exp USLT À mT   ð6Þ where l is a constant indicating the average dollar cost per variation of process capabilities. respectively. PCI indicates how expensive a product is expected to be if produced within the existing process capabilities. For example. It can be determined ex ante based on the analysis of existing product and process platforms. and is manifested by the size and lighting conditions of the gauges. i. for evaluating the extent of overall process variations among multiple products. 98 Design Studies Vol 30 No.Owing to the close correlation between costs and the cycle time. The meaning of l is consistent with that of the dollar loss per deviation constant widely used in Taguchi’s loss functions. and fulfillment of the CNs. 3 A-Kano decision-making The product planning stage features a series of processes including elicitation.e. Such a cost function produces a relative measure. Thus. F hffi ji ¼ i. A coherent A-Kano process model is developed to combine the various analytical techniques presented in Section 2. In general. visibility of the gauges in a car dashboard is a CN. To allow for unambiguous understanding. 3. analysis. and zi is defined the same as in Eq. instead of actual dollar figures. 1 January 2009 .. is assumed to be a linear function of the part-worth standard times for the FRs assumed by product p. the questionnaire is designed according to a set of CNs. . (4). The estimated cycle time for product p. namely (mT. Ig.1 Identification of functional requirements The A-Kano model requires the survey results of customer’s satisfaction using the Kano questionnaire. the FRs are typically related to customizable product attributes from which the customers can choose according to his/her preferences. 2. namely the FRs. mT and sT are the mean and the i i standard deviation of the part-worth standard time associated with fi. the CNs are translated into FRs.

If the division of market segments is not evident. the configuration index for each FR is computed. 3. 1. Next. In addition. In each method.Figure 4 A design process model of analytical Kano 3. sT Þ is identified from the standard routing that constitutes the i i An analytical Kano model 99 . the Kano questionnaire is fabricated and surveys are conducted to acquire the customers’ assessment of the FRs according to the functional and dysfunctional forms of questions. 1978). Many methods and tools are available to assist the process. such as conjoint analysis (Green and DeSarbo. the candidate product configuration is evaluated against the Kano evaluator. and (2) the configuration index-based method. 1993). the customers’ immediate perceptions of the importance of the FRs are extracted according to the self-stated importance (Table 4). income. it becomes necessary to carry out market investigations to differentiate the customer groups. etc. part-worth standard time ðmT . the customer’s average satisfaction/dissatisfaction with respect to the functional/dysfunctional form question is computed according to Eq. such as age. 3.2 Division of market segments Customers are grouped into different market segments based on their demographic information. and data mining (Jiao and Zhang.4 Computation of Kano indices For each market segment. perceptual mapping (Moore and Pessemier. which indicates the priority of the FR in product configuration design. 2005).3 Kano survey With respect to the FRs. Accordingly. gender. 3. the Kano indices are computed and plotted in the A-Kano diagram. As a prerequisite. namely (1) the Kano classification-based method.5 Product configuration design and evaluation Two alternative methods are adopted for product configuration design.

process platform.9 with an increment of 0. it is necessary to gain insight of the FRs with respect to the customer’s satisfaction and the producer’s capacity. the customer satisfaction (U) and cycle time index (C) of the set of N products are evaluated according to Eqs. 2005) is adopted to solve such a combinatorial optimization problem. 1 January 2009 . and a set of N product configurations are generated based on Heuristic 1. 3. (4) and (6). The dashboard is an important subassembly in the automobile with a number of customizable components and features (Figure 5).1. the relationship between the selection of FRs and the producer’s production capability is not considered in product design. leading to an expected value of the Kano evaluator (E). A full-factorial experiment requires 9 Â 9 Â 9 ¼ 729 runs of experiment. aL.1 to 0. aL changes from 5 to 45 with an increment of 5 . Sensitivity analysis entails an iterative process of evaluating the Kano evaluator based on specific FR classification schemes. and ultimately determines his/her satisfaction.1 Kano classification-based configuration design The threshold values of the Kano classifiers. aL. Genetic algorithms (GAs) have been proven to excel in solving combinatorial optimization problems in comparison with traditional calculus-based or approximation optimization techniques.5. leading to an expected value of the Kano evaluator (E). aH) are tested such that the FRs are classified as four categories. which in turn influence the design of product configuration and the expected value of the product. The optimal product configuration is generated as the one that produces the largest Kano evaluator. The optimal values of Kano classifiers are selected as those that produce the largest Kano evaluator. This is a typical combinatorial optimization problem where a large search space needs to be explored. However. These components can be represented as FRs in the design stage. 3. Sensitivity analysis is carried out to examine different settings of Kano classifiers. aH). Subsequently. 4 Case study The proposed method is applied to car dashboard design in an automobile company. the exact correspondence between FRs and customer’s satisfaction is not evident. Therefore. This research adopts a strategy where r0 changes from 0. a heuristic GA (Jiao and Zhang. The actual selection of the FRs collectively constitutes the car interior environment. Moreover.5. the FRs are classified according to each set of Kano classifier k ¼ (r0. namely k ¼ (r0. which influences the driver’s perceptions. In each run. In this regard.2 Configuration index-based configuration design Product configuration is generated using Heuristic 2. 100 Design Studies Vol 30 No. the customer’s satisfaction (U) and cycle time index (C) of the products are evaluated. using the Kano evaluator as the performance indicator. Different threshold values lead to different classification schemes. and aH changes from 50 to 90 with an increment of 5 . Subsequently.

The respondents were divided into three groups based on their age. as shown in Table 5. representing three market segments as shown in Table 6. including both the functional and dysfunctional forms. and give the perception of the importance of an FR using the self-stated importance (Table 4). 4. gender and income levels.Upper instrument panel Gauge display Air bag (inside) Storage compartment Vent Figure 5 Main components and features of car dashboard subassembly Steering wheel column Radio/CD control Lower instrument panel The company has conducted extensive market studies to collect the CNs which are represented according to well-known FR descriptors. The form of Kano questions used in this survey is shown in Table 1. This case study focuses on a set of 10 CNs. Each respondent was required to answer the Kano questions with respect to each and every FR. Table 5 Functional requirements of car dashboard Customer need Gauge display type Gauge display size Gauge display lighting Instrument panel material Functional requirement Analog display Digital display Large (15 inch) Small (10 inch) Back light External light Plasticewooden Plastic Plasticemetal Cream Gray Black Plasticewooden Plastic Plasticeleather High (24 inch) Low (20 inch) Integrated in steering wheel On control panel Top Side Front Passenger side Front Side Code f1e1 f1e2 f2e1 f2e2 f3e1 f3e2 f4e1 f4e2 f4e3 f5e1 f5e2 f5e3 f6e1 f6e2 f6e3 f7e1 f7e2 f8e1 f8e2 f9e1 f9e2 f9e3 f9e4 f10e1 f10e2 Instrument panel color Steering wheel material Steering wheel height Position of CD/radio control Storage configuration Air bag An analytical Kano model 101 .1 Kano survey A total of 50 car drivers constituted the Kano survey respondent set. which are refined as 25 FRs.

the average level of satisfaction/dissatisfaction for the functional/dysfunctional form question within each market segment is computed. 4. the Kano indices . the FRs are Figure 6 Kano model of functional requirement classification 102 Design Studies Vol 30 No.2 Comparative study A comparative study is carried out to validate the effectiveness of the A-Kano method.hðri .Table 6 Customer groups in Kano survey Market segment Segment 1 (s1) Segment 2 (s2) Segment 3 (s3) Age 46þ 31e45 21e30 Gender M/F M/F M Income (103S$/year) 120þ 60e119 30e59 Next. Table 7 also shows the part-worth standard times of the FRs. Accordingly. Yi Þ as shown in Figure 6. In the first procedure. 1 January 2009 . the company has identified and established standard routings as basic constructs of its process platform. With assembly-to-order production. which is further used to compute the ri configuration index. for each FR. The configuration index with respect to three market segments is summarized in Table 7. The company fulfills customer orders through assembly-to-order production while importing all components and parts via global sourcing. ai Þ are derived. The study involves three independent procedures to design products based on the same set of survey data. resulting in the data point ðXi . The part-worth standard time of each FR is established based on time and motion studies of the related assembly and testing operations.

54 0.6 216 247 sT (s) i 21 35 22 20 6 4 15. the FR classification is applied to assist the generation of product configurations.48 0. The merits of the three methods are compared according to the product performance in terms of the Kano evaluator.3 0.43 0.61 0.38 0.4 32 30.60 0. a twodimensional Kano diagram is established according to customers’ satisfaction and dissatisfaction levels.53 0.3 0.48 0. Heuristic 1 is adopted to generate product configurations based on the Kano classification.56 0. called Product I.3 40.53 0.74 0.63 0.54 0. as shown in Figure 6.33 0.4 9.35 0.44 0.57 0.19 0. the results of only one market segment (s1) are presented for purpose of brevity.59 s3 0.4 8 5.35 0.49 0.2.38 0.45 0.55 0.6 5. The analysis of the other market segments can be carried out following the same procedures.1 Kano-based product configuration design Based on the method proposed by DuMouchel (Berger et al.5 45 442 365 483 365 483 99 99 69 54 80.56 0.46 0. the respective FRs are classified using the Kano classifiers proposed in the A-Kano method.95 Part-worth standard time mT (s) i 498 557 505 493 90 50.74 0.30 0. According to this model.6 0.09 0.3 0.5 6 28 30.52 0. The product configuration.55 0.47 0.39 0.32 0.4 0.56 0. and accordingly the FRs are classified into four categories.2 0.5 5.51 0.7 0.08 0..53 0.4 32 9.38 0.42 0.49 0.42 0.29 0.74 0.26 0. 4.5 18 13 5. In both methods.55 0.67 classified using the Kano method proposed by DuMouchel (Berger et al. where a number ‘1’ indicates An analytical Kano model 103 . is generated manually and is shown in Table 8. the configuration indices of the FRs are used to guide the design of product configurations following the decision framework presented in Section 3.1 0..33 0.46 0.66 0. In the second procedure.Table 7 Configuration index and part-worth standard time FR Configuration index (market segment) s1 s2 0.35 0.08 0.69 0.64 0.14 0.55 0. 1993).55 0. In the third procedure.5 462 465 411 50 47.23 0.69 0.36 0. the diagram is divided into four quadrants.22 0. In this case study.31 0.74 0. 1993).2 0.85 0.6 4 33 40 f1e1 f1e2 f2e1 f2e2 f3e1 f3e2 f4e1 f4e2 f4e3 f5e1 f5e2 f5e3 f6e1 f6e2 f6e3 f7e1 f7e2 f8e1 f8e2 f9e1 f9e2 f9e3 f9e4 f10e1 f10e2 0.66 0.

aL and aH. the Kano evaluator is computed. i.63 1.03 14. an iterative process of sensitivity analysis is carried out. To select the appropriate values of Kano classifiers. For example.63 0. The Kano evaluator is computed using Eqs. it is left blank. otherwise. A-Kano-based product configuration design e Kano classifiers The Kano classification is implemented based on the Kano indices and Kano classifiers.2 104 Design Studies Vol 30 No. k ¼ (r0.2. (3)e(7). 1 January 2009 . Each set of Kano classifiers. aH) results in a scheme of FR classification.33 7.. namely r0.e. Figure 7 4. Hence. The sensitivity analysis involves three variables. Once the product configurations are created.48.Table 8 Product configurations generated based on three methods FR f1e1 f1e2 f2e1 f2e2 f3e1 f3e2 f4e1 f4e2 f4e3 f5e1 f5e2 f5e3 f6e1 f6e2 f6e3 f7e1 f7e2 f8e1 f8e2 f9e1 f9e2 f9e3 f9e4 f10e1 f10e2 Customer satisfaction Cycle time index Kano evaluator Product I (Kano-based) 1 1 Product II (A-Kano: classifier) 1 1 1 1 Product III (A-Kano: configuration index) 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 7.71 6.23 1.01 1 1 1 1 1 7.48 1 1 1 1 1 1 1 7. which in turn guides the generation of the product configurations following Heuristic 1. aL. such that Product I results in a Kano evaluator of 0. it is difficult to visualize the response surface using a three-dimensional graph. A graphical illustration can be made by setting one variable at a fixed value while changing the values of the other two variables.16 that the respective FR is included in the product.

each of which is evaluated against the Kano evaluator.14. For example. The AKano diagram for FRs classification is shown in Figure 8. these two FRs represent different customer perceptions considering their importance and relative customer satisfaction. respectively.3 4. Base on the sensitivity analysis.3. the polar form A-Kano classification assigns two FRs into one-dimensional and indifferent categories. which leads to a Kano evaluator of 1. and aH varies from 50 to 90 . Figure 7 Sensitivity analysis of Kano evaluator w. Although the process is straightforward. and aH ¼ 65 . The Kano evaluator is normalized for purpose of illustration.16.5.36 and 0. In the traditional Kano method. A heuristic GA is adopted to generate a set of solutions. Such a difference is effectively captured by the configuration index in the A-Kano method. A-Kano-based product configuration design e configuration index The A-Kano method proposes using the category index to guide the generation of product configurations. 4.2. respectively. the largest Kano evaluator is 1. aL ¼ 25 . The product configuration (Product III) derived from this method is shown in Table 8. which is obtained when the Kano classifiers are r0 ¼ 0. However. Interestingly. aL and aH An analytical Kano model 105 . respectively. FRs are selected manually based on the Kano category. both f2e2 and f8e2 are classified as indifferent FRs. which is shown in Table 8. which reflects the difference between them. where the configuration indices are 0.3 Results and analysis The major difference between the three methods lies in the criteria to select the FRs that constitute the product configuration.t.shows the result of sensitivity analysis when the Kano classifier r0 is fixed as r0 ¼ 0.01. it does not reflect the differences among FRs within the same category. and the product configuration derived from this classification is called Product II. aL varies from 5 to 45 .r.

both the traditional Kano method and A-Kano method based on Kano classifiers suffer the discontinuity problem. thus indicating identical customer preferences. which are almost identical. Hence. i. 7. would make both f2e2 and f5e2 fall into the indifferent category.. in the A-Kano method. and in turn the Kano evaluator.g. FRs f2e2 and f5e2 which are located closely to each other are classified as indifferent and attractive FRs in the traditional Kano.38..16. they make clear-cut distinctions between FRs that fall into different categories. The overall performance of the Product I is 0. such a distinction is not evident based on the Kano indices. e. respectively. The merit of the product configurations is evaluated against the Kano evaluator. which indicate better strategies to leverage upon customer’s satisfaction and the producer’s capacity.Figure 8 A-Kano model of functional classification requirement On the other hand.e. On the other hand. the classification using A-Kano classifiers reflects such similarities so that the two FRs both fall in the one-dimensional region. The latter two products offer larger expected shared surplus values. it is subjective to assign the threshold values.03. The major difference in the 106 Design Studies Vol 30 No.48. while those of Products II and III are 1. However. Interestingly. the product configurations lead to different customer’s satisfaction and cycle time index. A slight change of the classification threshold value. As can be seen from the results. moving the horizontal dashed line higher. even when these FRs are actually located close to each other in the Kano diagram.71 for the three products.33 and 7. For example. the configuration indices of these two FRs are 0.36 and 0. the overall customer satisfaction values are 7. A further comparison can be made between the product configurations resulted from the traditional Kano model and the A-Kano (Table 8). 1 January 2009 .01 and 1. As can be seen from the results.

Such a distinction reflects that Products II and III effectively account for the producer’s capacity in the configuration design.63.63. An outdated product may be very unsatisfying. The polar form of Kano indices . For example. it denotes the relative superiority of a classification in terms of the shared surplus.. it proposes a performance indicator that gives dual considerations of customer’s satisfaction and producer’s capacity to fulfill customer needs. The quantification of customer’s satisfaction based on Kano indices distinguishes the A-Kano method from other research techniques.23. It should be noted the absolute value of the Kano evaluator is not significant. In comparison with traditional Kano model that relies on qualitative Kano categories.. But rather. thus representing different aspects of customer perceptions. 1978). the meanings of satisfaction and dissatisfaction in the A-Kano model are different from those used in the utility theory. Furthermore. 5 Discussions The A-Kano method extends traditional Kano method by adopting quantitative measures of customer’s satisfaction. 2005). while the dysfunction of a product An analytical Kano model 107 . in the sense that it facilitates intuitive understanding of the FRs. The Kano classifiers and configuration indices enhance decision support in product design.Kano evaluator lies in the varied cycle time indices. the Kano model implies two different perspectives e proper functioning of a product feature arouses customer satisfaction. but very useful because one may rely on it to carry out important work. customer preference/demand is modeled as a utility function of various product attributes. In such a respect. represents the importance of an FR and the angle ai represents the relative level of satisfaction and dissatisfaction.. the A-Kano method establishes a coherent decision framework that deals with the interactions between the customer domain and the producer domain. In contrast. ai Þ excels the Cartesian form ðXi .. Wassenaar et al. ri. the magnitude of the vector. Yi Þ proposed by ri DuMouchel (Berger et al. stated choice methods (Louviere et al. hðri . the AKano method models customer preference based on customer’s satisfaction and dissatisfaction. 2003. prices and socioeconomic and demographic background of the market population (Wassenaar and Chen. Apparently. such as conjoint analysis (Green and DeSarbo. In particular. the economic benefits to the producer are modeled as a utility function of customer attributes. 7. 1993) as a quantification measurement of customer preferences. 2000). The main contribution of this research is to extract useful customer need information from Kano survey for decision support in product design. which are 14. In discrete choice analysis. and 6. 2003. Wassenaar et al. 2005). the questions ‘How satisfied are you with this product?’ and ‘How useful/valuable is this product to you?’ are in fact quite different. The major difference lies in the different denotations of customer preference. In classical conjoint analysis as well as the stated choice methods. and discrete choice analysis (Wassenaar and Chen. respectively.

this research does not assert that the Kano’s perspective is more reasonable than that used in the utility theory. such that the customizable product features can be better managed at the product planning stage. A few major challenges remain unaddressed for customer need analysis. decision support. it provides an alternative measurement of customer preferences. such as quantitative measure of customer satisfaction. while consolidating the theoretical foundation. In accordance with the Kano indices. and capacity assessment of the producers. following the basic principles of traditional Kano model. By adopting a polar form representation scheme. The configuration index in combination with the Kano evaluator is deemed to be useful extensions to traditional Kano where the classification criteria are subjective and unjustifiable. the Kano indices are conducive to exploring the nature of customer satisfaction. However. The merit of the designed product is evaluated against the Kano evaluator. 1 January 2009 . the Kano classifiers are deemed to outperform traditional classification criteria based on the logical polar form.Figure 9 Comparison of the utility theory and Kano’s theory feature generates dissatisfaction. and to conduct surveys for customer need analysis. This is a useful extension to the one-dimensional utility measurement. As an alternative decision process. Nevertheless. the configuration index is defined to prioritize the functional requirements so as to provide decision support to product configuration design. as shown in Figure 9. The A-Kano method defines systematic procedures to elicit the customer needs. This paper presents an analytical Kano model for customer need analysis. 6 Conclusions Product development for customer’s satisfaction with considerations of producer’s capacity necessitates logical prioritization of customer needs. The A-Kano model adopts the Kano indices for measuring customer’s satisfaction and dissatisfaction. 108 Design Studies Vol 30 No. The overall A-Kano framework address the customer need analysis from a broader scenario. leveraging upon both customer’s satisfaction and the producer’s capacity.

D. Concurrent Engineering: Research and Applications Vol 14 No 3 pp 173e185 Jiao. Timko. A. M. New York ` ` Riviere. de Pennington. R. 18e21 Oct. S (1984) Attractive quality and must-be quality. M M (2003a) Towards high value-added products and services: mass customization and beyond. Munich CD-ROM Kano. K and Hinterhuber. D and Swait. Zhang.org). Xie. J and Chen. The Journal of the Japanese Society for Quality Control 14 No 2 pp 39e48 Lai. UK McKay. E A (1993) Product planning management: designing and delivering value McGraw-Hill. J. C. Takahashi. R. Journal of Consumer Research Vol 5 No 1 pp 58e65 Hazelrigg. W L and Pessemier. Technovation Vol 23 pp 809e821 Jiao. Technology and Research (A*STAR) Science and Engineering Research Council (SERC) Thematic Strategic Research Programme grant on human factors engineering (#062 131 0066). Seraku.Acknowledgment This research is partially sponsored by the European Commission (DG Information Society and Media) in the framework of the 6th Research Program (FP6) under grant IST-5-035030-STREP (www. Technovation Vol 8 No 1 pp 25e38 Moore. Journal of Mechanical Design Vol 120 pp 653e658 Jiao. M M (1999) A pragmatic approach to product costing based on standard time estimation. Boger. Rogeaux. C H (2006) Customer requirement management in product development: a review of research issues. J and Tseng. N. J and Zhang. in 2004 IEEE International Engineering Management Conference. M and Walden. S (2003b) Process platform planning for mass customization in Proceedings of the Second Interdisciplinary World Congress on Mass Customization and Personalization. G (2006) Adaptive preference target: contribution of Kano’s model of satisfaction for an optimized An analytical Kano model 109 . J. N. Q and Tseng. L and Pokharel. P. Center for Quality of Management Journal Vol 2 No 4 pp 3e35 Green. A and Baxter. X. W S (1978) Additive decomposition of perceptions data via conjoint analysis. D (1993) Kano’s method for understanding customer-defined quality. Bolster.. K C (2004). Cambridge. Burchill. J. Also acknowledged is support from Singapore Agency for Science. D. F and Tsuji. G A (1998) A framework for decision based engineering design. A. IIE Transactions Vol 37 No 9 pp 801e814 Jiao. Ma. Pouliot. Technical University. J and Saporta. Monrozier. C. Richter. M and Tan. J (2000) Stated choice methods. W. Shen. References Berger. pp 1085e1089 Louviere. F.cater-ist. Computer Aided Design Vol 33 No 7 pp 511e520 Matzler. P E and DeSarbo. Hensher. Y (2005) Product portfolio planning with customereengineering interaction. Blauth. Rubinoff. J (2001) Requirements management: a representation scheme for product. Optimizing product design using the Kano model and QFD. R. G. analysis and application Cambridge University Press. International Journal of Operations & Production Management Vol 19 No 7 pp 738e755 Jiao. DuMouchel. H H (1998) How to make product development projects more successful by integrating Kano’s model of customer satisfaction into quality function deployment. Pages.

ASME Journal of Mechanical Design Vol 127 pp 514e523 110 Design Studies Vol 30 No. Cheng. Wassenaar. P and Ozan. Y. K C and Shen. ASME Journal of Mechanical Design Vol 125 pp 490e497 Wassenaar. X X (2000) Integrating Kano’s model in the planning matrix of quality function deployment. Kauffmann. PhD thesis. A (2005) Enhancing discrete choice demand modeling for decision-based design. H J and Chen. Chen. J and Sudjianto. 1 January 2009 . E (2007) Integration of Kano’s model into QFD for multiple product design. N P (2001) Axiomatic design-advances and applications Oxford University Press. H J. Delft. Total Quality Management Vol 11 No 8 pp 1141e1151 Tielemans.preference analysis using a sequential consumer test. New York Tan. IEEE Transactions on Engineering Management Vol 54 No 2 pp 380e390 Suh. Eburon. P F J (1995) Lead time performance in manufacturing systems. Food Quality and Preference Vol 17 pp 572e581 Sireli. W. W (2003) An approach to decision-based design with discrete choice analysis for demand modeling.

Sign up to vote on this title
UsefulNot useful