Chapter 6
Analytical Attribute Approaches:
Introduction and Perceptual Mapping
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.
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)
A Determinant Gap Map Figure 6.2
A Data Cube Figure 6.3
.700
s
nt
.
de
.
on
2
sp
1
Re
1 2 3 .... Options .... X Ideal
1
2
Attributes
.
.
.
.
.
.
.
15
Obtaining Customer Perceptions Figure 6.4
Rate each brand you are familiar with on each of the following:
Disagree Agree
1. Attractive design 1..2..3..4..5
2. Stylish 1..2..3..4..5
3. Comfortable to wear 1..2..3..4..5
4. Fashionable 1..2..3..4..5
5. I feel good when I wear it 1..2..3..4..5
6. Is ideal for swimming 1..2..3..4..5
7. Looks like a designer label 1..2..3..4..5
8. Easy to swim in 1..2..3..4..5
9. In style 1..2..3..4..5
10. Great appearance 1..2..3..4..5
11. Comfortable to swim in 1..2..3..4..5
12. This is a desirable label 1..2..3..4..5
13. Gives me the look I like 1..2..3..4..5
14. I like the colors it comes in 1..2..3..4..5
15. Is functional for swimming 1..2..3..4..5
Snake Plot of Perceptions
Figure 6.5
(Three Brands)
Ratings
5
4.5
4
Aqualine
3.5
3 Islands
2.5
Sunflare
2
1.5
1
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 Attributes
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 Number Figure 6.6
of Factors
Factor Eigenvalue Percent Variance
Explained
1 6.04 40.3
2 3.34 22.3
3 0.88 5.9
4 0.74 4.9
45 5 0.62 4.2
40 6 0.54 3.6
Percent Variance
35 The Scree 7 0.52 3.5
30 8 0.44 3.0
25 9 0.40 2.7
Explained
20
15
10
5
0
1 2 3 4 5 6 7 8 9 No. of Factors
Factor Loading Matrix Figure 6.7
Attribute Factor 1 -- Factor 2 --
“Fashion” “Comfort”
1. Attractive design .796 .061
2. Stylish .791 .029
3. Comfortable to wear .108 .782
4. Fashionable .803 .077
5. I feel good when I wear it .039 .729
6. Is ideal for swimming .102 .833
7. Looks like a designer label .754 .059
8. Easy to swim in .093 .793
9. In style .762 .123
10. Great appearance .758 .208
11. Comfortable to swim in .043 .756
12. This is a desirable label .807 .082
13. Gives me the look I like .810 .055
14. I like the colors it comes in .800 .061
15. Is functional for swimming .106 .798
Factor Scores Matrix Figure 6.8
Attribute Factor 1 -- Factor 2 --
“Fashion” “Comfort”
1. Attractive design 0.145 -0.022
2. Stylish 0.146 -0.030
3. Comfortable to wear -0.018 0.213
4. Fashionable 0.146 -0.017
5. I feel good when I wear it -0.028 0.201
6. Is ideal for swimming -0.021 0.227
7. Looks like a designer label 0.138 -0.020
8. Easy to swim in 0.131 0.216
9. In style -0.021 -0.003
10. Great appearance 0.146 0.021
11. Comfortable to swim in -0.029 0.208
12. This is a desirable label 0.146 -0.016
13. Gives me the look I like 0.148 -0.024
14. I like the colors it comes in 0.146 -0.022
15. Is functional for swimming -0.019 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.
The AR Perceptual Map Figure 6.9
Comfort
Aqualine
Gap 1
Islands Molokai
Fashion
Splash
Sunflare
Gap 2
Dissimilarity Matrix Figure 6.10
Aqualine Islands Sunflare Molokai Splash
Aqualine X 3 9 5 7
Islands X 8 3 4
Sunflare X 5 7
Molokai X 6
Splash X
The OS Perceptual Map Figure 6.11
Co
Aqualine
mf
or
t
Islands Molokai
Fashion
Sunflare
Splash
Comparing AR and OS Methods Figure 6.12
AR Methods 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
(factors) must be interpreted by the researcher
Where Used
Situations where attributes are easily Situations where it may be difficult for the
articulated or 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.
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?
