1 6-1

McGraw-Hill/Irwin © 2003 The McGraw-Hill Companies, Inc., All Rights Reserved.

2 6-2

CHAPTER SIX

ANALYTICAL ATTRIBUTE APPROACHES:

INTRODUCTION AND PERCEPTUAL MAPPING

3 6-3

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.

4 6-4

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)

5 6-5

A Determinant Gap MapFigure 6.2

6 6-6

1 2 3 .... Options .... X Ideal

12 . . . . . . .15

Att

ribut

esR

espo

nden

ts

12 . .

700.

A Data Cube Figure 6.3

7 6-7

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..57. Looks like a designer label 1..2..3..4..58. Easy to swim in 1..2..3..4..59. 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

Obtaining Customer PerceptionsFigure 6.4

8 6-8

1

1.5

2

2.5

3

3.5

4

4.5

5

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15

Snake Plot of Perceptions (Three Brands)

Aqualine

Islands

Sunflare

Attributes

Ratings

Figure 6.5

9 6-9 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”

10 6-10

Factor Eigenvalue Percent VarianceExplained

1 6.04 40.32 3.34 22.33 0.88 5.94 0.74 4.95 0.62 4.26 0.54 3.67 0.52 3.58 0.44 3.09 0.40 2.7

0

5

10

15

20

25

30

35

40

45

1 2 3 4 5 6 7 8 9 No. of Factors

Pe

rce

nt

Va

ria

nc

eE

xp

lain

ed

The Scree

Selecting the Appropriate Number of Factors Figure 6.6

11 6-11

Attribute Factor 1 --“Fashion”

Factor 2 --“Comfort”

1. Attractive design .796 .0612. Stylish .791 .0293. Comfortable to wear .108 .7824. Fashionable .803 .0775. I feel good when I wear it .039 .7296. Is ideal for swimming .102 .8337. Looks like a designer label .754 .0598. Easy to swim in .093 .7939. In style .762 .12310. Great appearance .758 .20811. Comfortable to swim in .043 .75612. This is a desirable label .807 .08213. Gives me the look I like .810 .05514. I like the colors it comes in .800 .06115. Is functional for swimming .106 .798

Factor Loading MatrixFigure 6.7

12 6-12

Attribute Factor 1 --“Fashion”

Factor 2 --“Comfort”

1. Attractive design 0.145 -0.0222. Stylish 0.146 -0.0303. Comfortable to wear -0.018 0.2134. Fashionable 0.146 -0.0175. I feel good when I wear it -0.028 0.2016. Is ideal for swimming -0.021 0.2277. Looks like a designer label 0.138 -0.0208. Easy to swim in 0.131 0.2169. In style -0.021 -0.00310. Great appearance 0.146 0.02111. Comfortable to swim in -0.029 0.20812. This is a desirable label 0.146 -0.01613. Gives me the look I like 0.148 -0.02414. I like the colors it comes in 0.146 -0.02215. Is functional for swimming -0.019 0.217

Sample calculation of factor scores: From the snake plot, the mean ratings of Aqualine on Attributes1 through 15 are 2.15, 2.40, 3.48, …, 3.77. Multiply each of these mean ratings by the correspondingcoefficient 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.

Factor Scores MatrixFigure 6.8

13 6-13

Aqualine

Islands

Splash

Molokai

Sunflare

Gap 1

Gap 2

Fashion

Co

mfo

rt

The AR Perceptual MapFigure 6.9

-2 2

-2

2

14 6-14

Aqualine Islands Sunflare Molokai SplashAqualine X 3 9 5 7Islands X 8 3 4Sunflare X 5 7Molokai X 6Splash X

Dissimilarity MatrixFigure 6.10

15 6-15

Aqualine

Islands

Splash

Molokai

SunflareC

omfort

Fashion

The OS Perceptual Map Figure 6.11

16 6-16

AR Methods OS MethodsInput Required

Brand ratings on specific attributes Overall similarity ratingsAttributes must be pre-specified Respondent uses own judgment of similarity

Analytic Procedures Commonly UsedFactor analysis; multiple discriminant analysis Multidimensional scaling (MDS)

Graphical OutputShows product positions on axesAxes interpretable as underlying dimensions(factors)

Shows product positions relative to each otherAxes obtained through follow-up analysis or mustbe interpreted by the researcher

Where UsedSituations where attributes are easily articulated orvisualized

Situations where it may be difficult for therespondent 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.

Comparing AR and OS Methods Figure 6.12

17 6-17

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?