Analytical Attributes Approaches

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.
 Practical sequence: I bought this pair of shoes
because they are made with high quality
craftsmanship and of high quality leather (feature),
because they are excellent walking shoes
(function), because they are very comfortable
(benefit).
 Theoretical sequence: feature permits a function
which provides a benefit.
Quantitative Analytical Attribute
Technique: Gap Analysis

What
about
here?
Collecting Data for Gap Analysis

Product Options
1 2 3 . . . . (or Brands) . . . . X Ideal
1
2
.
.
Attributes

. Brand 1 = Aqualine
. Brand 2 = Islands
. Brand 3 = Sunflare
. Brand 4 = Molokai
. Brand 5 = Splash
.
.
15
Obtaining Customer Perceptions
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
 Aqualine Islands Sunflare
1 2.15 2.27 3.83
2 2.4 2.77 3.57

Snake Plot of Perceptions 3

4
3.48
2.2
2.98
2.48
2.31
3.65
5 3.68 3.2 2.29
6 3.81 3.15 2.57
7 2.29 2.29 3.51
Ratings 8 3.57 3.09 2.21
9 2.52 2.72 3.67
10 2.51 2.72 3.57
11 3.47 3.09 2.32
5
12 2.6 2.33 3.8
13 1.97 2.39 3.68
4.5
14 2.28 2.25 3.65
15 3.77 3.04 2.4
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 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
7 0.52 3.5
35 8 0.44 3.0
30 9 0.40 2.7
Percent Variance

The Scree
25
Explained

20
15
10
5
0
1 2 3 4 5 6 7 8 9 No. of Factors
Factor Loading Matrix
Attribute Factor 1 -- “Fashion” Factor 2 --
“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

Important Note: Factor loadings are different from factor scores. Factor scores
are coefficients of cases (or brands) on the factors, whereas factor loadings are
coefficients of variables (or attributes) on the factors.
Factor Score Coefficient Matrix
Attribute Factor 1 -- “Fashion” Factor 2 --
“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

Comfort
Aqualine
Gap 1

Molokai
Islands

Fashion

Splash
Sunflare

Gap 2
 Shortcomings of Gap Analysis
1. Input comes from responses to questions on how brands
differ (so nuances are ignored)
2. Brands considered as sets of attributes (so totalities and
interrelationships are overlooked)
3. Breakthrough creations often require a conceptual leap
4. No demand or current technology to support gaps
5. Analysis and mapping may be history by the time data
are gathered and analyzed
Quantitative Analytical Attribute Technique:
Trade-Off (Conjoint) Analysis

1. Put the determinant attributes together in

combinations or sets.
2. Respondents rank these sets in order of
preference.
3. Conjoint analysis finds the optimal levels of each
attribute.
Conjoint Analysis Input: Salsa Example
Thickness Spiciness Color Actual Ranking as
Ranking* Estimated by
Model
Regular Mild Red 4 4
Regular Mild Green 3 3
Regular Medium-Hot Red 10 10
Regular Medium-Hot Green 6 8
Regular Extra-Hot Red 15 16
Regular Extra-Hot Green 16 15
Thick Mild Red 2 2
Thick Mild Green 1 1
Thick Medium-Hot Red 8 6
Thick Medium-Hot Green 5 5
Thick Extra-Hot Red 13 13
Thick Extra-Hot Green 11 11
Extra-Thick Mild Red 7 7
Extra-Thick Mild Green 9 9
Extra-Thick Medium-Hot Red 14 14
Extra-Thick Medium-Hot Green 12 12
Extra-Thick Extra-Hot Red 17 18
Extra-Thick Extra-Hot Green 18 17

* 1 = most preferred, 18 = least preferred.

Conjoint Analysis: Graphical Output

Thickness Spiciness Color

2

-1

-2
Regular Thick Ex-Thick Mild Medium-Hot Ex-Hot Red Green
0.161 0.913 -1.074 1.667 0.105 -1.774 -0.161 0.161
Conjoint Analysis:
Relative Importance of Attributes

0 20 40 60 80 100 %

Spiciness 59.8%
Thickness 34.6%
Color 5.6%
Three Attribute Example
Movie Theatre Design
 Price
 $8
 $10
 Seat
 Average-size seats with cup holder
 Large-size seats without cup holder
 Large-size lounge without cup holder
 Concession
 Standard hot dogs and popcorn
 Gourmet deli and snacks
 Rank Across Options Average Factor
Rank Utility Range Importance
Respondent 1
Price $8 1 2 5 6 9 10 5.5 1.0 2 18.18%
$10 3 4 7 8 11 12 7.5 -1.0

Seat Average 1 2 3 4 2.5 4.0

Large 5 6 7 8 6.5 0.0 8 72.73%
Lounge 9 10 11 12 10.5 -4.0

Concession Standard 1 3 5 7 9 11 6.0 0.5 1 9.09%

Gourmet 2 4 6 8 10 12 7.0 -0.5
11
Respondent 2
Price $8 1 2 3 7 8 9
$10 4 5 6 10 11 12

Seat Average 1 2 4 5
Large 3 7 6 10
Lounge 8 9 11 12

Concession Standard 1 4 3 6 8 11
Gourmet 2 5 7 10 9 12

Respondent 1 & 2
Price $8
$10

Seat Average
Large
Lounge

Concession Standard
Gourmet
Conjoint Analysis: Key Assumptions

1. Products can be specified as a bundle of

attributes.

2. Determinant attributes must be known before

the analysis.

3. Respondents should be familiar with the

product category and the attributes to be able to
rank preference and purchase likelihood.

4. Products can be made with the combinations of

desired attributes.
Additional Qualitative Attribute Analysis
Techniques
 Dimensional Analysis
 Stimulator Checklists

And, there are many others.

A Dimensional Attribute List
 Weight  Explosiveness
 Rust resistance  Flammability
 Length  Aroma
 Color  Translucence
 Water resistance  Buoyancy
 Materials  Hangability
 Style  Rechargeability
 Durability  Flexibility
 Shock resistance  Malleability
 Heat tolerance  Compressibility
An Idea Stimulator Checklist for

Consumer Products
 Can it be adapted?
 Can it be modified?
 Can it be reversed?
 Can it be combined with anything?
 Can something be substituted?
 Can it be magnified?
 Can it be minified?
 Can it be rearranged in some way?