Market Forces January 2008

Vol. 3 No. 4

ANALYSIS OF CONSUMER
PREFERENCE OF MOBILE
PHONES THROUGH THE USE
OF CONJOINT ANALYSIS
FAISAL AFZAL SIDDIQUI, MUHAMMAD SAEED AWAN

E:mail: fasidfas@yahoo.com

Abstract
The aim of this article is to provide an introduction to conjoint analysis as a
research tool, and to indicate its value for analyzing consumer preference based on the
value that the consumers attach to the attributes of the goods that they intend to purchase.
The study involves the use of the tool of conjoint Analysis to evaluate consumer
preference vis.a.vis Brand, Price Level, FM Radio, and Camera. The results of the study
indicate that the most important attribute behind consumer preference for mobile phones
was Brand, followed by Price, then Camera, and finally FM Radio.

Key Words
Conjoint Analysis, Coefficient of Determination, Sample, Orthogonal Design.
Aim of the Article
The aim of this paper is two fold: one, to provide an overview of the nature and
use of conjoint analysis as a research tool, and two, to indicate its value through the
application of the tool to analyze the value that the consumer attaches to the features /
attributes of a mobile telephone when making his/her purchase decision. Although the
mobile telephone industry has reached maturity, high growth and fashion trend cycles
keep accelerating. The presence of more and more brands with new features has created a
competitive environment that is unprecidented (Rutter & Edwards, 1999:31).
Conjoint analysis has been used in research for many years (Green & Sprinivasan,
1978). Hair et al (1998:388) state that the application of conjoint analysis in the United
States has been paralleled in other parts of the world including Europe. In Pakistan,
however, it appears to have only recently attracted the attention of a local research
community.

Past research mostly used survey methods that directly measure consumers
attitudes towards products and their attributes. The drawback of such methods according
to Lang and Crown (1993), is the possibility of interaction effects among attributes being
overlooked. It is possible that the preferences of the consumers may depend on the joint
influence of product attributes such as quality, style, features and price. It is therefore
necessary that the joint effect of several product attributes on the final decision to
purchase a specific item should be taken into consideration when researching consumer
purchase decisions.
The Nature of Conjoint Analysis
The concept of conjoint analysis is described by Hair et al (1998:392) as follows:
Conjoint analysis is a multivariate technique used specifically to understand how
respondents develop preferences for products or services. It is based on the simple
premise that consumers evaluate the value of a product or service by combining the
separate amounts of value provided by each attribute. Sudman and Blair (1998:229-230)
emphasize that it is not a data analysis procedure like factor analysis or cluster analysis. It
must be regarded as a type of thought experiment designed to show how various
elements of products or services (price, brand, style) predict customer preferences for a
product or service. Kotler (2000:339) defines conjoint analysis as a method for deriving
the utility values that consumers attach to varying levels of a products attributes.
Churchill and Iacobucci (2002:748) refer to conjoint analysis as conjoint
measurement, which relies on the ability of respondents to make judgments about
stimuli. These stimuli represent some predetermined combinations of attributes, and
during a laboratory experiment, respondents are asked to make judgments about their
preferences for various attribute combinations. The basic aim, therefore, is to determine
the features they most prefer. The value of conjoint analysis lies in the fact that it
estimates how much each of these attributes is valued, and as Churchill and Iacobucci
(2002:748) state, the word conjoint has to do with the notion that the relative values
of things considered jointly can be measured when they might not be measurable if taken
one at a time.
Key Decisions or Steps when Designing a Conjoint Value Analysis
There are many different conjoint methods. The researcher should weigh each
research situation and pick the right combination of tools for the project. Sudman and
Blair (1998:235) distinguish between an arrangement that uses all possible combinations
of features (full factorial design) and one that uses only some of the combinations
(fractional design). A general rule of thumb, according to these authors, is to limit the
descriptions to no more than 30. Full-profile conjoint value analysis (CVA) is useful for
measuring up to about six attributes (Hair et al, 1998:401).
CVA is designed for paper and pencil study, but can be computerized by means of the
Ci3 System for computer interviewing. CVA calculates a set of utilities for each
individual, using traditional full-profile card-sort (either rating or ranked) or pair-wise

ratings. CVA can attach prices to each attribute level to measure price sensitivities for
individual features. This is realistic for modeling categories in which buyers actually see
the prices for each component of the product, such as with restaurant meals, cellular
phones and packages (Omre, 2000:4). If the full-profile approach is used, it is important
to limit the number of attributes and levels, increase the number of profiles, or use more
arsimonious models (such as the vector or ideal point models) so as to increase the
degrees of freedom for conjoint estimation (Green & Srinivasan, 1990). Hair et al
(1998:400-436) provide a comprehensive description of the design of a conjoint analysis
experiment (see Figure 2). The Hair model consists of seven phases, consisting of 14
steps. Churchill and Iacobucci (2202:754) See (Figure 1) present a simpler 6 stage model
based on the more critical decision points in a conjoint experiment.
1.

Select attributes

The attributes used emerge primarily from the objectives of the study. The
researcher should be guided by the principle that the attributes should be both capable of
being acted on, and are important to consumers. These attributes are those that the
company can do something about, i.e. it has the technology to make changes that might
be indicated by consumer preferences (Churchill & Iacobucci, 2002:754).
2.

Determine attribute levels

The number of levels for each attribute has a direct bearing on the number of
stimuli respondents will be asked to judge. The more there are, the heavier the burden
that is placed on the consumer. Churchill and Iacobucci (2002:754) suggest that the
researcher make the range for the various attributes somewhat larger than the range
normally found but not so large as to make the options different to choose.
3.

Determine attribute combinations

This will determine what the full set of stimuli will look like. One cannot expect a
respondent to provide meaningful judgments if there are five attributes and three levels
(3x3x3x3x3=243) each of rank-order judgments.
4.

Select form of presentation of stimuli and nature of judgments

Basically, three approaches can be used, namely: verbal description, paragraph

description, and pictorial representation. When visual aids are used they are normally
used in combination with verbal descriptions. The nature of the judgments that must be
secured from the respondents is related to the form of presentation. One method is to ask
them to rank the alternatives according to preference or intention to buy, another method
which has gained popularity among the researchers is the use of rating scales.
5.

Decide on aggregation of judgments

This step basically involves the decision whether the responses from consumers
or groups of consumers will be aggregated, and if so, how will this be done? If groups are
formed, operationally this means estimating the utilities for the individual-level models
and then clustering them into homogeneous groups. According to Churchill and
Iacobucci (2002:759), this highlights an attractive feature of conjoint analysis because it
allows market share predictions for selected product alternatives.
6.

Select analysis technique

The final step in the design of a conjoint analysis project is to select the technique
that will be used to analyze the data. The choice depends largely on the method that was
used to secure the input judgments by the respondents. For example, when rank-order
data have been obtained, the assumption of a linear relationship may be dubious, so a
non-metric regression model may be substituted to estimate the utilities (Churchill &
Iacobucci, (2002:759).

APPLICATION OF CONJOINT ANALYSIS TO EXAMINE CONSUMER

PREFERENCE FOR MOBILE TELEPHONES

This study aims to find the consumer preference regarding the purchase of a new
mobile telephone in the Karachi market. The consideration are: one, brand preference,
Nokia, Sony Ericsson, Samsung, LG, and Motorola; two, price, three, the feature of FM
Radio, and Four, Camera. The analysis was based on randomly selected samples. The
numerical values assigned to each factor are given below. The ranking of brands is based
on survey according to which the market share of the four brands under consideration are:
Nokia 50%, Sony Ericsson 25%, Samsung 15%, LG 05%, and Motorola 05%).
Brand
Nokia
Sony Ericsson
Samsung
LG
Motorola

Rank
1
3
4
5

FM Radio
Yes
No

Rank
1
2

Camera
Yes
No

Rank
1
2

Brand

Price without

Price only for

Price only for

Price for having

Nokia
Sony Ericsson
Samsung
LG
Motorola

3000
2800
2800
2700
2500

3300
3000
3000
2700
2800

5500
4200
4000
3600
3300

6300
5300
5500
5400
5400

METHODOLOGY

We use the Conjoint Analysis tool to find the relation of brand preference, price
level, and two of features FM Radio and Camera.
EMPIRICAL STUDY

The study involves 50 respondents filling a questionnaire incorporating 16

choices as shown in the following table. The respondents had to fill column six indicating
their preferences. The figures holed in column 6 of the table below give the overall
preference of the respondents based on the communication frequency of their choices.
Card # Brand
1
Motorola

FM Radio Camera Price (PKR)

No
No
2500

Preference
15

FM Radio

2
3
4
5
6
7
8
9
10
11
12
13
14
15
16

Sony Ericsson
Nokia
Motorola
Nokia
Sony Ericsson
LG
LG
Sony Ericsson
Sony Ericsson
Samsung
Nokia
Nokia
Samsung
Samsung
Samsung

Yes
No
Yes
Yes
Yes
Yes
No
No
No
No
No
Yes
No
Yes
Yes

No
No
Yes
No
Yes
No
Yes
No
Yes
Yes
Yes
Yes
No
Yes
No

3000
3000
5400
3300
5300
2700
3600
2800
4200
4000
5500
6300
2800
5300
3000

6
3
16
2
4
13
14
9
5
10
8
1
11
7
12

ANALYSIS AND DISCUSSION:

The total utility of the consumer can be found by adding utilities of Brand, FM
Radio, Camera, and Price.
Total utility = utility (Brand) + utility (FM Radio) + utility (Camera) + utility
(price) + Constant
The value of the constant was determined as 44.319. Details of the utility for each
factor / attribute is detailed out in the table below:

Total Factors with original values

Brand
FM Radio Camera Price
Constant Nokia
yes
yes
6.387
-6.832 44.319
Sony
Ericsson
Samsung
2800
Motorola
3300
3600
4000
4200
5300
5400
-14.758
-15.031
-17.217

Factors s Utility
Brand FM Radio Camera Price
2500 -3.764
-3.766
no
no
2700
-7.529
-11.293
-18.822

-7.652
-9.019

LG

-9.838
-10.932
-11.478
-14.484
5500
6300

-7.531

-12.7

The range of the utility values (highest to lowest) for each factor / attribute
provides a measure of how important the factor is to overall preference. Factors with
higher utility ranges play a more significant role than those with lower utility. In our
colleted data we found that Brand has the highest importance in a consumers preference
for a mobile phone. In terms of percentage the value of brand is 42.303%, next in line is
price at 29.176%, followed by the feature of Camera which has a weight of 17.942 % and
FM Radio was 10.579 %.
From above finding, it is evident that the Brand Equity plays a key role in the
consumers purchase decision. Conjoint analysis can therefore serve as a powerful tool
for the companies, not only for developing their pricing strategy, but also for deciding
whether to decresae price or to increase their brand equity through the increase in their
advertizing and promotion budget to increase their market share. Conjoint analysis can
also facilitate the company in comparing the cost of adding new features like FM radio
and Camera with the extra revenue on price increase that is possible for product with
those features. For example, the analysis revaled that the maximum total utility was 23.57
when brand was Nokia with FM Radio and Camera and price was Rs 2500, as against a
total utility of 13.185 at the current market price of 6300. The implication being, if
Nokia reduces the price of the new model with camerea and FM Radio to 2500, it would
be able to get 100% market share and all other brand would be forced out from the
market (assuming that the other brands do not follow suit and reduce their prices).If the
price of the Model was 3600 instead of 2500 the total utility will decline from 23.57 to
20.564 and its market share would decline from 100 to 85%, with the remaining 15%
being the share of the competing brands. Nokia mobile phone with a camera but without
FM radio and Sony Ericsson is with FM Radio and Camera have almost the same level of
utility. In other words the difference in consumer utility with regard to the brand
preference of the two is equal to the consumer utility for FM Radio.
REFERENCES
Leeuw, J., 2006 R in Psychometrics and Psychometrics in R. (UCLA Department of
Statistics,
Preprint
#
491)
[Internet].
Available
at:
http://preprints.stat.ucla.edu/index_body
Swait, J., 2005 Comment on Current Issues and a wish list for Conjoint Analysis. Applied
Stochastic Models in Business and Industry, [online] 21: pp 331332. Available at:
www.statisticalinnovations.com/articles/swaitcomments.pdf
Levin, R.I., and Rubin, D.S., 2000, Statistics for Management. 7 ed. New Jersey: Prentice
Hall.
Curry, J., 1996 Understanding Conjoint Analysis in 15 Minutes. (Sawtooth Software
Research Paper Series, Sawtooth Technology, Inc). [Internet]. Available at:
www.sawtoothsoftware.com

SPSS manual. 2007 [Internet]. Available at: www.spss.com

Brayn, O., 2005 Getting Started with Conjoint Analysis. Research Publisher LCC.
Lehman, D.R., Gupta, S., and Steckel, J.H., 1998 Marketing Research. New York:
Addison Wesley.
Green, P.E., & Srinivasan, V., 1978. Conjoint Analysis in Consumer Research: Issues
and Outlook. Journal of Marketing 55 (September):103-123.
Hair, J.F., Anderson, R.E., Tatham, R.L., & Black, W.C., 1998. Multivariate Data
Analysis. 5th Ed., New Jersey: Prentice-Hall International.
Rutter, N. & Edwards, O., 1999. Ready to Ware. Forbes 163(7): 30-33.
Lang, J. & Crown, E., 1993. Country-of-origin effect in apparel choices: A conjoint
analysis. Journal of Consumer Studies and Home Economics. 17(6):87- 98.
Sudman, S. & Blair, E., 1998. Marketing Research. Boston: McGraw Hill.
Kotler, P. 2000. Marketing Management. The Millennium Edition. London: PrenticeHall.
Churchill, G. & Iacobucci, D., 2002. Marketing Research, Methodological Foundations.
8th Ed. London: Harcourt Publishing.
Omre, B., 2000. Which Conjoint Method should I use? Sawtooths Software.
Green, P.E. & Srinivasan, V., 1990. Conjoint Analysis in Marketing: New Developments
with Implications for Research and Practice. Journal of Marketing. 54 (October):3- 19.

COMPUTER SOFTWARE USED

SPSS 15.0

SPSS CODES
CONJOINT PLAN='D:\Data\Mobiledesign.sav'
/DATA='D:\Data\Mobiledata.sav'
/SEQUENCE=PREF1 TO PREF16
/FACTORS=BRAND (LINEAR LESS)
FMRADIO (LINEAR LESS)
CAMERA (LINEAR LESS)
PRICE (LINEAR LESS)
/PRINT=SUMMARYONLY

RESULT OF SPSS ANALYSIS

Conjoint Analysis
[DataSet2] D:\Data\Mobiledata.sav
Warnings
No reversals occurred.

Model Description

N of Levels
Brand
FMradio
Camera
Price

5
2
2
12

Relation to
Ranks or
Scores
Linear (less)
Linear (less)
Linear (less)
Linear (less)

Dependence of Factor Levels

1
2
3
4
5

Level for Factor A

Price = 2800
Price = 2800
Price = 3000
Price = 5300
Price = 5300

Level for Factor B

FMradio = 2
Camera = 2
Camera = 2
FMradio = 1
Camera = 1

Factor level on left side always occurs with factor level

on right side.

Cramer's V Statistics
Brand
Brand
FMradio
Camera
Price

1
.000
.000
.866

FMradio
.000
1
.000
.913

Camera
.000
.000
1
1.000

Price
.866
.913
1.000
1

Factors are not all orthogonal.

Utilities

Brand

FMradio
Camera
Price

Nokia
Sony Erricson
Samsung
Motrolla
LG
yes
no
yes
no
2500.00
2700.00
2800.00
3000.00
3300.00
3600.00
4000.00
4200.00
5300.00
5400.00
5500.00
6300.00

(Constant)

Utility
Estimate
-3.764
-7.529
-11.293
-15.058
-18.822
-3.766
-7.531
-6.387
-12.773
-6.832
-7.379
-7.652
-8.199
-9.019
-9.838
-10.932
-11.478
-14.484
-14.758
-15.031
-17.217
44.319

Std. Error
.601
1.201
1.802
2.402
3.003
1.611
3.223
3.416
6.832
3.909
4.222
4.378
4.691
5.160
5.629
6.255
6.567
8.287
8.444
8.600
9.851
14.063

Importance Values
Brand
FMradio
Camera
Price

Averaged Importance Score

Coefficients
B
Brand
FMradio
Camera
Price

Estimate
-3.764
-3.766
-6.387
-.003

Correlationsa
Pearson's R
Kendall's tau

Value
.914
.750

42.303
10.579
17.942
29.176

Sig.
.000
.000

a. Correlations between observed

and estimated preferences