application of the kano model to sensory ...

Customer 

satisfied 


neutral 


dissatisfied 

Needs not fulfilled 

Attractive 

Attribute 

Needs fulfilled 

Performance 

Attribute 

Must-be 

Attribute 

Application of the Kano model to Sensory 

Science - a new approach to develop the 

“ideal” fruit smoothie 

Presentation at the 8th Pangborn Symposium 

July, 29 th 2009 

8 th Pangborn Symposium Florence

Content 

Background / Research questions 

1. Illustration of the product evaluation by consumers 

2. Illustration of the product perception by the descriptive panel 

3. Data linkage, classification into Kano’s typology 

Management Summary 

8 th Pangborn Symposium Florence 

page 2


Kano´s Model uses three categories of attributes: “Attractive Attributes”, 

“Performance Attributes” and “Must-be Attributes”. 

All three attributes are following the 

rules: 

The more intense the more satisfying 

Or 

The less intense the more satisfying 

Needs not 

fulfilled 

Two possible 

preferred 

directions of 

Intensity 


satisfied 


neutral 


dissatisfied 


Attractive 

Attribute 

Must-be 

Attribute 

Performance 

Attribute 

Needs 

fulfilled 

+ - 

page 3


Two Reasons for modifying the application of Kano´s Model 

1. Consumers are not able to split the complex sensory perception food and 

beverage products into single product attributes in terms of answering the 

functional and dysfunctional question for every attribute 

2. Untrained consumers cannot measure intensities for single attributes 

=> Solution to solve these problems: 

1 

2 

3 

Consumer Test in order to measure the level of satisfaction for 

each sample of the yellow smoothies. 

Descriptive Analysis with the same sample in order to gain 

objective perception measures. 

Linkage of consumer data and perception; adaptation of the 

Kano Model to identify the three attribute categories 


page 4

1 

2 

3 


Structure of the research project 

Consumer Test in order to measure the level of 

satisfaction for each sample of the yellow 

smoothies. 

Descriptive Analysis with the same sample in 

order to gain objective perception measures. 

Linkage of consumer data and perception via PLS1 

- adaptation of the Kano Model to identify the three 

attribute categories 

Three sessions 

with 120 

consumers 

sensory 

profiles of the 

16 smoothies 

1. Linkage of the 16 

sensory profiles to the 

5 satisfying & 5 

dissatisfying products 

(for each consumer) 

2. Clustering the 

consumers in three 

clusters depending on 

both linkages 

3. Identifying the 3 Kano 

attribute categories for 

each cluster 


1. Setting a satisfaction threshold for 

each consumer for the 16 smoothies 

2. Acceptance of 5 satisfying products 

3. Acceptance of 5 dissatisfying 

products 


satisfied 


neutral 


dissatisfied 



16 smoothies 



16 smoothies 

…Acceptance 

of 5 satisfying 

products 

…Acceptance 

of 5 

dissatisfying 

products 

Cluster 1 Cluster 2 Cluster 3 


Attractive 

Attribute 


Performance 

Attribute 

Must-be 

Attribute 


satisfied 


neutral 


dissatisfied 


Attractive 

Attribute 


Performance 

Attribute 

Must-be 

Attribute 


satisfied 


neutral 


dissatisfied 


Attractive 

Attribute 


Performance 

Attribute 

Must-be 

Attribute 

page 5

With apple & 

mango 

bits (Market 

standard) 

With lots of 

apple & 

mango bits 


16 samples were provided for this research, all modifications of an existing 

market product. 

Only 

mango 

bits 

Only 

apple 

bits 

Dominant 

banana 

impression 

Dominant 

passion fruit 

impression 

Brown 

colour 

Standard without bits 

Orange 

Colour 

Coconut 

taste 

Vanilla 

Taste 


Reduced 

viscosity 

Increased 

Viscosity 

Sweet 

alternative 

Sour 

Alternative 

Sweetness 

& sourness 

reduced 

page 6

1 

Illustration of the product evaluation by consumers 

Setting a satisfaction threshold based on the model of limited purchase 

decisions (Foscht/ Swoboda, 2004, S. 154) 

I will certainly not consume this product again 

I will probably not consume this product again 

I will probably consume this product again 

I will certainly consume this product again 

Evoked Set 

☺ 

Total Set 

Awareness Set Unawareness Set 

Inert Set 

 


Dissatisfying products 

Satisfying products 

Inept Set 

 

page 7

1 


1. Session: Each consumer tried all 16 smoothies and decided 

individually if he/she would consume them again or not. 

Step 1: Unforced decision 

Step 2: Forced decision 


page 8

1 



page 9

1 


2. and 3. Session of the Consumer Test: Each consumer rated 5 out of his 

satisfying and 5 out of his dissatisfying smoothies on a 9-point hedonic scale 


satisfied 


neutral 


dissatisfied 


Attractive 

Attribute 


Performance 

Attribute 

Must-be 

Attribute 


2. Session 

5 out of the satisfying 

smoothies 

3. Session 

5 out of the dissatisfying 

smoothies 

page 10

2. Illustration of the product perception by 

the descriptive panel 

Approach 


page 11

2 

Illustration of product perception by the descriptive panel 

The smoothie samples have been evaluated by the Descriptive Panel 

Assessment of objective perception data: 

• The Descriptive Panel (10 panelists // two measurements; n=20) assessed all 16 smoothies that 

have been evaluated by the consumers. 

• All samples have been served with a temperature of 7 to 8°C. 

• For the assessment of the samples the panel uses 44 “descriptors” to describe all sensory 

dimensions (appearance, odor, taste, texture, mouth feeling). 

• For 31 out of the 44 descriptors significant differences can be detectet (p-value is ≤ 0.05) 


page 12

3. Data linkage, classification into Kano’s 

typology 

Segment-specific classification 


page 13

3 Data linkage, classification into Kano’s typology 

101 

Consumers 

x 

For each consumer the influence coefficient on acceptance is estimated by using 

satisfying & dissatisfying acceptance scores and perception data for PLS 1 

sensory profiles 

of the 16 

smoothies 

Consumer 

1 

2 

3 

… 

…Acceptance of 

5 individually 

satisfying 

products 

31 statistically significant 

Descriptors 

yellow 

orange 

colour 

grey 

brown 

tone 

… 


sensory profiles 

of the 16 

smoothies 

Consumer 

1 

2 

3 

… 

yellow 

orange 

colour 

Acceptance of 5 

individually 

dissatisfying 

products 


Descriptors 

grey 

brown 


… 

page 14


In terms of consumers heterogeneity 3 segments have been detected by 

an innovative segmentation approach using Wards-Hierarchical-Clustering 

Consumer 

1 

2 

3 

… 

yellow 

orange 

colour 

Segment 1 

Containing 40 consumers 


descriptors of satisfying products 

grey 

brown 


… 

Consumer 

1 

2 

3 

… 



Descriptors of dissatisfying products 

yellow 

orange 

colour 

Analysing Segments of Consumers by comparing the 62 

acceptance influencing coefficients of every consumer 

Segment 2 


grey 

brown 


… 

Segment 3 


page 15


In the following, all “influential” descriptors were classified to the three Kano 

categories according to the following classification rules Riviére et al., 2006, S. 576) 

• Descriptors found “influential” only in the 3rd 

session modelling (dissatisfying products) 

are must-be attributes 

• Descriptors found influential only in the 2nd 

session modelling (satisfying products) are 

attractive attributes 

• Key drivers for both sessions modelling are 

performance attributes. 


satisfied 


neutral 


dissatisfied 



Session 3 

(dissatisfying 

products) 

Session 2 (satisfying 

products) 

Attractive 

Attribute 


Performance 

Attribute 

Must-be 

Attribute 

page 16


Consequences for optimizing the ideal smoothie 

1. Optimize all must-be attributes to 

avoid dissatisfaction 

2. Optimize all performance attributes to 

ensure satisfaction 

3. Optimize all attractive attributes to 

arouse high satisfaction level and to 

distinguish from competitive products 


satisfied 


neutral 


dissatisfied 



Session 3 

(dissatisfying 

products) 

Session 2 (satisfying 

products) 

Attractive 

Attribute 


Performance 

Attribute 

Must-be 

Attribute 

page 17


Having classified the “influential” descriptors, the direction of influence is 

of critical importance. 

Must-be 

attribute 

attractive 

attribute 

performance 

attribute 

Recommendations for 

product development 

positive influence negative influence 

high perception of this sensory attribute 

leads to absence of consumer 

dissatisfaction and 

low perception of this sensory attribute 

leads to high consumer dissatisfaction 


leads to high consumer satisfaction and 



satisfaction 

the higher the perception of this sensory 

attribute the higher the consumers 

satisfaction and vice versa 

Increase the perception of this 

attribute 



leads to high consumer dissatisfaction 

and 



dissatisfaction 



satisfaction and 


leads to high consumer satisfaction 

the lower the perception of this sensory 

attribute the higher the consumer 

satisfaction and vice versa 

Decrease the perception of this 

attribute 

page 18


Segment 1 


sour 

"influential" descriptor 

A_yellow/ orange 

O_sourly 

O_overall intensity 

O_apricot/ peach 

sweet 

banana 

lemon 

M_slimy 

O_banana 

T_viscous 

A_viscous 

overall intensity 

apricot/ peach 

Kano´s attribute classification 

must-be attribute - 



must-be attribute + 


performance attribute + 


performance attribute - 

attractive attribute + 







page 19


Segment 2 


"influential" descriptor 

coconut 

T_viscous 

A_viscous 

M_slimy 

overall intensity 

sweet 

A_grey-brown-tone 

A_yellow/ orange 

sour 

T_homogenous 

vanilla 

A_visible fruit bits 

O_vanilla 

T_puree level 

Kano´s attribute 

classification 









attractive attribute - 



attractive attribute - 




page 20

Management Summary 

Innovativeness and advantages of this approach: 

• Classifying the products into satisfying and dissatisfying products for every consumer 

individually based on well accepted approach of limited purchase decisions (Foscht/ 

Swoboda, 2004, S. 154) 

⇒Identifying the influence of each descriptor for satisfying and dissatisfying products via 

PLS separately 

• Hierarchical Clustering using the 62 regression coefficients of the satisfying and 

dissatisfying products for depending variables 

⇒homogenous segments of consumers having comparable preferences in purpose of 

making conclusions for a group of consumers instead of identifying the Kano attributes 

for every consumer individually 

• Using Classification Rules (Rivieré et al., 2006) to adapt Kano´s Theory of Satisfaction 

⇒Identifying the three Kano attributes must-be, performance and attractive attributes which 

provide more detailed information about the influence on satisfaction than only 

identifying the preference driving attributes without differentiation between satisfying 

and dissatisfying products 

⇒ interesting for food industries working on innovative food products not knowing which 

product attributes have to be fulfilled (must-be) and which product attributes are able to 

arouse high satisfaction levels (attractive attributes) 


page 21

Thank you for your attention! 

Any questions? 


page 22

application of the kano model to sensory ...

You also want an ePaper? Increase the reach of your titles

Delete template?

Save as template?