16-1
McGraw-Hill/Irwin
© 2003 The McGraw-Hill Companies, Inc., All Rights Reserved.
26-2
CHAPTER SIX
ANALYTICAL ATTRIBUTE
APPROACHES:
INTRODUCTION AND
PERCEPTUAL MAPPING
36-3
What are Analytical Attribute Techniques?
• Basic idea: products are made up of attributes - a future product change must involve one or
more of these attributes.
• Three types of attributes: features, functions,
benefits.
• Theoretical sequence: feature permits a
function which provides a benefit.
46-4
Gap Analysis
• Determinant gap map (produced from
managerial input/judgment on products)
• AR perceptual gap map (based on attribute
ratings by customers)
• OS perceptual map (based on overall
similarities ratings by customers)
56-5
A Determinant Gap Map
Figure 6.2
66-6
A Data Cube
2.
1
Attributes
1
2
.
.
.
.
.
.
.
15
Figure 6.3
700
.
.
1
2
3
.... Options .... X
Ideal
76-7
Obtaining Customer Perceptions
Figure 6.4
Rate each brand you are familiar with on each of the following:
Disagree
1. Attractive design
2. Stylish
3. Comfortable to wear
4. Fashionable
5. I feel good when I wear it
6. Is ideal for swimming
7. Looks like a designer label
8. Easy to swim in
9. In style
10. Great appearance
11. Comfortable to swim in
12. This is a desirable label
13. Gives me the look I like
14. I like the colors it comes in
15. Is functional for swimming
Agree
1..2..3..4..5
1..2..3..4..5
1..2..3..4..5
1..2..3..4..5
1..2..3..4..5
1..2..3..4..5
1..2..3..4..5
1..2..3..4..5
1..2..3..4..5
1..2..3..4..5
1..2..3..4..5
1..2..3..4..5
1..2..3..4..5
1..2..3..4..5
1..2..3..4..5
86-8
Snake Plot of Perceptions (Three Brands)
Figure 6.5
Ratings
5
4.5
4
Aqualine
3.5
Islands
3
2.5
Sunflare
2
1.5
1
1
2
3
4
5
6
7
8
9
10 11 12 13 14 15 Attributes
96-9
Data Reduction Using Multivariate
Analysis
• Factor Analysis
– Reduces the original number of attributes to a smaller
number of factors, each containing a set of attributes that
“hang together”
• Cluster Analysis
– Reduces the original number of respondents to a smaller
number of clusters based on their benefits sought, as
revealed by their “ideal brand”
Selecting the Appropriate Number
of Factors
Percent Variance
Explained
6-10
10
45
40
35
30
25
20
15
10
5
0
The Scree
1
2
3
4
5
Factor
Eigenvalue
1
2
3
4
5
6
7
8
9
6.04
3.34
0.88
0.74
0.62
0.54
0.52
0.44
0.40
6
7
8
9
Percent Variance
Explained
40.3
22.3
5.9
4.9
4.2
3.6
3.5
3.0
2.7
No. of Factors
Figure 6.6
6-11
11
Factor Loading Matrix
Attribute
1. Attractive design
2. Stylish
3. Comfortable to wear
4. Fashionable
5. I feel good when I wear it
6. Is ideal for swimming
7. Looks like a designer label
8. Easy to swim in
9. In style
10. Great appearance
11. Comfortable to swim in
12. This is a desirable label
13. Gives me the look I like
14. I like the colors it comes in
15. Is functional for swimming
Factor 1 -“Fashion”
.796
.791
.108
.803
.039
.102
.754
.093
.762
.758
.043
.807
.810
.800
.106
Figure 6.7
Factor 2 -“Comfort”
.061
.029
.782
.077
.729
.833
.059
.793
.123
.208
.756
.082
.055
.061
.798
6-12
12
Factor Scores Matrix
Attribute
1. Attractive design
2. Stylish
3. Comfortable to wear
4. Fashionable
5. I feel good when I wear it
6. Is ideal for swimming
7. Looks like a designer label
8. Easy to swim in
9. In style
10. Great appearance
11. Comfortable to swim in
12. This is a desirable label
13. Gives me the look I like
14. I like the colors it comes in
15. Is functional for swimming
Factor 1 -“Fashion”
0.145
0.146
-0.018
0.146
-0.028
-0.021
0.138
0.131
-0.021
0.146
-0.029
0.146
0.148
0.146
-0.019
Figure 6.8
Factor 2 -“Comfort”
-0.022
-0.030
0.213
-0.017
0.201
0.227
-0.020
0.216
-0.003
0.021
0.208
-0.016
-0.024
-0.022
0.217
Sample calculation of factor scores: From the snake plot, the mean ratings of Aqualine on Attributes
1 through 15 are 2.15, 2.40, 3.48, …, 3.77. Multiply each of these mean ratings by the corresponding
coefficient in the factor score coefficient matrix to get Aqualine’s factor scores. For example, on
Factor 1, Aqualine’s score is (2.15 x 0.145) + (2.40 x 0.146) + (3.48 x -0.018) + … + (3.77 x -0.019)
= 2.48. Similarly, its score on Factor 2 can be calculated as 4.36. All other brands’ factor scores are
calculated the same way.
6-13
13
The AR Perceptual Map
Comfort
2
Figure 6.9
Aqualine
Gap 1
Islands
Molokai
Fashion
-2
Splash
Sunflare
Gap 2
-2
2
6-14
14
Dissimilarity Matrix
Aqualine
Islands
Sunflare
Molokai
Splash
Aqualine
X
Islands
3
X
Sunflare
9
8
X
Molokai
5
3
5
X
Figure 6.10
Splash
7
4
7
6
X
6-15
15
The OS Perceptual Map
Aqualine
Islands
Splash
Molokai
Sunflare
Figure 6.11
6-16
16
Comparing AR and OS Methods
AR Methods
Figure 6.12
OS Methods
Input Required
Brand ratings on specific attributes
Overall similarity ratings
Attributes must be pre-specified
Respondent uses own judgment of similarity
Analytic Procedures Commonly Used
Factor analysis; multiple discriminant analysis
Multidimensional scaling (MDS)
Graphical Output
Shows product positions on axes
Shows product positions relative to each other
Axes interpretable as underlying dimensions
Axes obtained through follow-up analysis or must
(factors)
be interpreted by the researcher
Where Used
Situations where attributes are easily articulated or
Situations where it may be difficult for the
visualized
respondent to articulate or visualize attributes
Source: Adapted from Robert J. Dolan, Managing the New Product Development Process: Cases and Notes
(Reading, MA: Addison-Wesley, 1993), p. 102.
6-17
17
Failures of Gap Analysis
• Input comes from questions on how brands differ
(nuances ignored)
• Brands considered as sets of attributes; totalities,
interrelationships overlooked; also creations requiring
a conceptual leap
• Analysis and mapping may be history by the time data
are gathered and analyzed
• Acceptance of findings by persons turned off by
mathematical calculations?