Printer Friendly

Segmentation of kidswear market in India through cluster analysis--an analytical study.


Driven by huge demand from brand conscious children, the Indian kidswear retail market is expected to touch Rs 58,000 crores by 2014, according to retail consultancy firm Technopak Advisors. At present, the size of kidswear market in India is estimated at about Rs 38,000 crores--accounting for 25 per cent of the total Indian apparel category growing at the rate of 17% per annum (Shukla, 2010). As per Technopak Advisors in another report, the kids' clothing market in India is growing up; and in 2011, it was Rs. 37,300 crores (around US$7 billion) which will record a compound annual growth rate (CAGR) of above 10% for the next decade. A study by apex chamber of commerce ASSOCHAM forecasted the market size in 2015 at Rs. 80,000 crores (US$14.8 billion) (Pennsylvania, 2013).

John, 1999, reviewed that there can be no doubt that children are avid consumers and become socialized into this role from an early age. Throughout childhood, children develop the knowledge, skills, and values they will use in making and influencing purchases now and in the future. Modi & Jhulka, 2012, concluded in their study that uncovering the reason of people's purchase is an extremely difficult task. The buyer's psyche is a black box whose working can only be partially deduced. The buyer is subject to many influences which trace a complex course through his psyche and lead eventually to overt purchasing responses. So, it becomes extremely important for the marketers to know about the segmentation of the market which has been defined as a process of dividing the market into groups of customers with different needs, wants or characteristics; who therefore might appreciate products geared especially for them (Grewal & Levy, 2008). Unless and until the marketer knows their segments properly, positioning becomes wrong, which has been explained as the marketing strategy to differentiate the brand or product and place it in the human mind, which will aid the customer in fast recalling (Saxena, 2002). Due to the increase in the number of new players in the market, product differentiation in this competitive market is gaining its importance--which has been defined as a product offering and is perceived by the consumer to differ from its competition on any physical or nonphysical product characteristic (Dickson, 1987). Market segmentation and target marketing explains how the customer wants to identify product attributes and their relative importance in the targeted segment and is instrumental in the creation of superior customer value by segmenting a market and identifying appropriate target segments, along with finding out the relevant product attributes that appeal most to these segments (Lonial, Menezes, & Zaim, 2000).

In the changing scenario of the market, and that too in the kidswear segment where so many categorisations have been done on the basis of gender, age, product etc, it is really tough for the marketers to formulate appropriate marketing policies as per the various segments. In this context, this paper attempted to segregate the kidswear market into various segments on the basis of some market driven factors related to product attributes and consumer preferences or negative biases towards it, through the help of cluster analysis.


According to (Cadogan & Foster, 2000), price is probably the most important consideration for the average consumer; but consumers with high brand loyalty are willing to pay a premium price for their preferred brand, so, their purchase intention is not easily affected by price. In addition, customers have such a strong belief in the price and value of their favorite brands that they would compare and evaluate prices with alternative brands (Keller, 2003). (Michael & Barnett, 2012), stated that the upcoming brands try following the latest fashion because fashion also largely influences the purchase to a particular brand by consumer. (Frings, 2009), concluded that fashion trends are the styling ideas that major collections have in common and is one of the important motive of a consumer's buying of apparel products, as a consumer's selection is frequently influenced by his or her opinion of what is currently fashionable.

Size is a very important aspect in kidswear as it influences the fitting of the garment. Consumers tend to buy clothing for their children which can last for a little longer, because of their dynamic growth rate. (Frings, 2009), stated that the ability to create a good fit is the most important skill needed in the development of the garment. The size of the garment must fit to give the comfort and ease of movement to the user. A comfortable fit was the most commonly mentioned product influence among males and females.

Regardless of brand or price features, a garment still needs to "look right" if a person is going to buy it. A balance needs to be struck by retailers between clothes that will generally fit well and clothes that will fit specific people extremely well. Fit has a dual role: physical comfort and mental comfort (i.e. the way you feel when wearing a garment) (Mullarkey, 2001).

As a result of the psychological effects that color has on human beings, the use of colour has become an important marketing tool because colour affects the moods of the customers; retailers agonise over the colour of the product. Colour is not only the factor that is important when trying to sell a product, but it is color that attracts the customer. In fact, colour is ranked among the top three considerations in the purchase decision (Cooper, 2001). Colour induces moods and emotions, influences consumers' lerperceptions and behavior and helps companies position or differentiate from the competition (Aslam, 2006).

(Michael & Barnett, 2012) stated that the upcoming brands try following the latest fashion because fashion also largely influences the purchase to a particular brand by the consumer. (Frings, 2009), concluded that fashion trends are the styling ideas that major collections have in common and is one of the important motives of a consumer when buying apparel products, as a consumer's selection is frequently influenced by his or her opinion of what is currently fashionable.

From the above literature review, certain factors were identified and market tested to confirm whether these variables may be used for segmenting the kidswear market through cluster analysis. The segmentation of customers is a standard application of cluster analysis which allows segments to be formed that are based on data that are less dependent on subjectivity (Mooi & Sarstedt, 2011). While doing the literature survey, which was a major hurdle we have not come across many studies which are of this type, and especially any study in the apparel sector.


There are two types of research problems--those that relate to states of nature, and those which relate to relationships between variables. (Kothari, 2004) to formulate the research objectives, two steps were involved, i.e. understanding the problem thoroughly and rephrasing the same into meaningful terms from an analytical point of view. In the research study, the term research objective and research problem shall be used in the same connotation (Nargundkar, 2004). The study tried to define the objectives unambiguously to help discriminate relevant data from irrelevant one. In the research study, to define the problem, we took into account the purpose of the study, the relevant background information, what information was needed and how it would be used in decision making (Nargundkar, 2004). The primary objective of the study is to find out the various segments in the kidswear market on the basis of some market linked factors. It also tried to define the market characteristics of every segment on the basis of those attributes. It will help the marketers to formulate appropriate marketing policies according to the segments, which can also act as a guideline where the marketer can decide which segment(s) should be catered through their product offerings.


The research design of the study is partly exploratory and partly descriptive in nature. The objective of this exploratory research is to explore or search through a problem or situation to provide insight and understanding (Kothari, 2004). The major objective of the exploratory research is to identify and define the problem and scope by helping to arrive at the best research design, method of data collection and sample; which is characterised by highly flexible, unstructured and at times informal research methods (Easwaran, Singh, & Sharmila, 2010). Descriptive studies attempt to determine the frequency with which something occurs or the relationship between two phenomena--here emphasis would be on obtaining the relative frequency of occurrence of the given phenomenon (Mazumdar, 1991).

Type of Data

In this study, both primary and secondary data was used. Primary Data is originated by the researcher for the specific purpose of addressing the problem at hand and Secondary Data (Kothari, 2004) has already been collected for purposes other than the problem at hand.

Data Collection Method

Survey method was used for data collection on opinions, thoughts and feelings and personal interview was been used as media for the study, where the sample size was restricted to 100 respondents.

Data Collection Tool

As a data collecting tool we have used structured non-disguised questionnaire with both open and close ended questions. Non-disguised approach is a direct approach in which the purpose of the project is disclosed to the respondents or is otherwise obvious to them from the questions asked. The reason for asking structured questions is to improve the consistency of the wording used in doing the study at different places which increases the reliability of the study by ensuring that every respondent is asked the same question (Nargundkar, 2004). The 5 point Likert Scale is used to collect the data where '5' represents strong agreement with the statements and '1' represents strong disagreement.

Sampling Technique

In the proposed research study, we have implemented Probability Sampling Technique (Nargundkar, 2004), where each sampling unit has a known probability of being included in the sample. Systematic sampling technique has been used in the study, where the sample frame is a representation of the elements of the target population, and is the list of loyal customers in the kidswear market provided by a chain store. It has been chosen for the low cost of data collection, no need to enumerate all study objects and operationally, it is easier to control (Mazumdar, 1991). The sample size was bound within 100 respondents.

Data Analysis

Statistical inferences were drawn from the primary data collected by applying statistical tools like SPSS 19 and statistical analysis like Cluster Analysis.


Cluster Analysis

It is a data reduction tool that creates subgroups that are more manageable than individual datum. Cluster analysis (CA) is an exploratory data analysis tool for organizing observed data (e.g. people, things, events, brands, companies) into meaningful taxonomies, groups, or clusters, based on combinations of factors, which maximises the similarity of cases within each cluster, while maximising the dissimilarity between groups that are initially unknown (Banerjee & Agarwal, 2013).

Using cluster analysis, a customer 'type' can represent a homogeneous market segment. Identifying their particular needs in that market allows products to be designed with greater precision and direct appeal within the segment. Targeting specific segments is cheaper and more accurate than broad-scale marketing. Customers respond better to segment marketing which addresses their specific needs, leading to increased market share and customer retention.

Cluster analysis, like factor analysis, makes no distinction between dependent and independent variables. The entire set of interdependent relationships is examined. Whereas factor analysis reduces the number of variables by grouping them into a smaller set of factors, cluster analysis reduces the number of observations or cases by grouping them into a smaller set of clusters. Data were collected on four variables to segment the market, namely price/price range, fashion/trend, size/fit and colour/design which were been identified from market scanning through pilot survey and literature review also.

A five point Likert Scale (1-5) was formulated on four statements (factors), where 1 denotes strongly disagreement, 2 denotes disagreement, 3 denotes neither agreement nor disagreement with the statement, 4 denotes agreement and 5 denotes strong agreement with the statement. Survey was done on the 100 sample size and responses based on opinions were collected.
State your agreement or disagreement with the
statements between 1-5 (1 = Strongly Disagree,
2 = Disagree, 3 = Neither Agree Nor Disagree, 4
= Agree, 5 = Strongly Agree)

Price is not an important factor
while choosing any kidswear

Fashion & Trend is not an aspect
which I follow for purchase of my

Size & Fit is not an important
factor for my kidswear purchase.

Color & Design is not a very
important factor which affects my

Technique Adapted

As we don't know the number of groups or clusters that will emerge in our sample and because we want an optimum solution, a two-stage sequence of analysis occurs as follows:

1. An hierarchical cluster analysis using Ward's method applying squared Euclidean Distance as the distance or similarity measure was carried out. This helped to determine the optimum number of clusters we should work with.

2. In the next stage, the hierarchical cluster analysis was rerun with the selected number of clusters, which enabled us to allocate every case in our sample to a particular cluster.

Hierarchical Cluster Analysis

This is the major statistical method for finding relatively homogeneous clusters of cases based on measured characteristics. It starts with each case as a separate cluster, i.e. there are as many clusters as cases, and then combines the clusters sequentially, reducing the number of clusters at each step until only one cluster is left. The clustering method uses the dissimilarities or distances between objects when forming the clusters. The SPSS programme calculates 'distances' between data points in terms of the specified variables.

Ward's Method

This method is distinct from other methods because it uses an analysis of variance approach to evaluate the distances between clusters. In general, this method is very efficient. Cluster membership is assessed by calculating the total sum of squared deviations from the mean of a cluster. The criterion for fusion is that it should produce the smallest possible increase in the error sum of squares.

The results start with an agglomeration schedule which provides a solution for every possible number of clusters from 1 to 100 (the number of our cases). The column to focus on is the central one which has the heading 'coefficients'. Reading from the bottom upwards, it shows the agglomeration coefficient for one cluster to another.

Ward Linkage

If we rewrite the coefficients as in Table 2 (since it is not provided on SPSS), it is easier to see the changes in the coefficients as the number of clusters increase. The final column, headed 'Change', enables us to determine the optimum number of clusters. In this case it is 4 clusters, as succeeding clustering add much lesser to distinguishing between cases. A clear demarcation point seems to be there after 4th Row.

K-means Clustering

This method of clustering is very different from the hierarchical clustering and Ward method, which had been applied previously when there was no prior knowledge of how many clusters there may be or what they are characterised by. K-means clustering is used when you already have hypotheses concerning the number of clusters in your cases or variables. This is the type of research question that can be addressed by the k-means clustering algorithm. In our study, we have used both the hierarchical and the k-means techniques successively. The former (Ward's method) is used to get some sense of the possible number of clusters and the way they merge, as seen from the dendrogram. Then, the clustering is rerun with only a chosen optimum number in which to place all the cases (k means clustering). One of the biggest problems with cluster analysis is identifying the optimum number of clusters. As the fusion process continues, increasingly dissimilar clusters must be fused.

It is at this point that clear distinguishing characteristics of the clusters are visible and the Cluster 2 is the most attractive cluster, as the market size is the highest.

Table No 4 states that Cluster 2 has 62% of the respondent profile, which has a large population of prospective customers of the kidswear market, while Cluster 3 has very exclusive strata, which may be an unattractive segment of the kidswear market with 2% of the cases.

The ANOVA Table indicates which variables contribute the most to our cluster solution. Variables with large mean square errors and lowest F statistics provide the least help in differentiating between clusters. Thus, price as per the Table 5, is not a very helpful factor as the other variables in forming and differentiating clusters. So, most of the variables are very significantly different among the clusters. This can be perfectly defined for customers in the kidswear market, because although the parents check the price, it is not the defining factor while choosing the brands for their kids. They are ready to pay the price, if they like the product because of an increasing disposable income, double income in the family and a single kid.

The differences between Final Cluster Centres Table, shows the Euclidean distances between the final cluster centres. Greater distances between clusters mean there are greater dissimilarities. The dissimilar cluster groups have been ranked as per the Table 7:
Table 7: Cluster Groups and Distances

Rank Cluster Groups Distances

1 Cluster 1 & 3 6.170
2 Cluster 2 & 3 5.265
3 Cluster 3 & 4 4.435
4 Cluster 4 & 2 2.911
5 Cluster 1 & 4 2.856
6 Cluster 1 & 2 2.478

When cluster memberships are significantly different, they can be used as a new grouping variable in other analyses. The significant differences between variables for the clusters suggest the features in which the clusters differ or on which they are based; the more the difference, the more the uniqueness in the segment. This helps the marketers know if they want to enter into multiple segments with their product lines or can target the next segment in their growth strategy. Cluster 3 is quite different from any other cluster. These differentiations do not indicate any positive or negatives aspects of a cluster; it depends on the subjective evaluation of the retailers.

Table 8, explains the market's characteristics, as formed from Table 3. Cluster 1 may be defined as a utility driven customer who focuses on basic product attributes (economy brands that offer utility value to the customers may target this segment), Cluster 2 represents more value for money customers, who want the most at a defined resource (economy brands with value additions but are ready to extend their budget for a "suitable" product), Cluster 3 is more focused on fashion and trends in mass market products, where they are not very accustomed to product standardisation (manufacturers making popular products /knockoffs/ duplications may target this segment) and Cluster 4 may be defined as a more rational segment in their buying--decision pattern (the premium and the high end brands may be more suitable in this segment) . Looking into the subjective aspects of retail market today, these segments perfectly suits the kidswear market, although cross segment buying is becoming a common phenomenon in the market.


So, overall cluster analysis by Hierarchical methods and K-Means reveals the various characteristics of the different 4 clusters on the basis of four market--oriented factors chosen namely; price/price range, fashion/trend, size/fit and colour/design. This will help the retailers to think of the various segments in the kidswear market on the basis of four important variables and can develop their marketing policies accordingly. Most of the clusters are significantly different from each other and significant differences also lie between the various groups of clusters also. Now, it depends on the marketer as to which segment(s) they would prefer to cater through their product offerings, so that they can create distinct brand positioning in the market without eating away the market share of their own brands. This study creates further scope of research on marketing strategies for each identified segment. (Kotler & Keller, 2006) suggested that a company would be wise to enter one segment at a time. Competitors must not know to what segment(s) the firm will move next and can create segment-by-segment invasion plan. Cluster 1 and 2, 1 and 4 and 4 and 2 may be targeted at the same time with little stretching in the product depth by introducing multiple brands, as they have a close range of preferences; but at the same time the marketers have to be very careful, so that the brands or product offerings do not eat away each other market share. A clear distinction among the offerings must be maintained in segmentation. For Cluster 1 and 3, 2 and 3 and 3 and 4, it is advisable not to target multiple segments as the market requirements and the preferences are quite distinct from each other and it will be tough for the marketers to position their brands in an extended stretch.

High growth categories in kids boys market include T-Shirts/
Shirts, Bottomwear and Uniforms

 Market Size CAGR (2009-2020)
 2009 (INRCr)

Uniforms 4,980
Tee/Shirts 3,890
Bottom wear 3,340
Ethnic 1,300
Winter wear 1,280
Others 680
Denim 295

Boys Male < 14yrs

Note: Table made from bar graph.

High growth categories in kids girls market include
Ethnicwear, Dresses, Bottomwear Uniforms

 Market Size CAGR (2009-2020)
 2009 (INRCr)

Uniform 4,230
Ethnic 3,295
Dresses 2,225
Bottom wear 2,180
Winter wear 1,220
Others 540
Tee/Shirts 456
Denim 45

Girls: Female <14yrs

Note: Table made from bar graph.


(1.) Aslam, M. M. (2006). Are You Selling the Right Colour? A Cross-cultural Review of Colour as a Marketing Cue. Journal of Marketing Communications, 12(1), 15-30.

(2.) Banerjee, S., & Agarwal, N. (2013). Cluster and Conjoint Analysis of the Entry Level Brands in Casual Wear Market in India w.r.t. Product Attributes. Journal of Marketing & Communication, 8(4), 44-51.

(3.) Cadogan, J. W., & Foster, B. D. (2000). Relationship Selling and Customer Loyalty: An Empirical Investigation. Marketing Intelligence and Planning, 18(4), 185-199.

(4.) Cooper, M. (2001). The colour of money may actually be fuchsia. New York : New York Public Library Online .

(5.) Dickson, P. R. (1987). Market segmentation, product differentiation, and marketing strategy. Journal of Marketing, 51(2), 1-10.

(6.) Easwaran, S., Singh, & Sharmila, J. (2010). Marketing Research-Concepts, Practices and Cases (11th ed.). Oxford University Press.

(7.) Frings, G. S. (2009). Fashion from Concept to Consumer. (7th ed.). New Delhi, India: Pearson Education.

(8.) Grewal, D., & Levy, M. (2008). Marketing. New Delhi, India: McGraw Hill Companies.

(9.) John, D. R. (1999). Consumer Socialization of Children: A Retrospective Look at Twenty-Five Years of Research. Journal of Consumer Research, 26(3), 183-213.

(10.) Keller, K. L. (2003). Strategic Brand Management: Building, Measuring and Managing Brand Equity. Ney Jersey: Prentice Hall.

(11.) Kothari, C. R. (2004). Research Methodology-Methods & Techniques, (2nd ed.). New Delhi, India: New Age International.

(12.) Kotler, P., & Keller, K. L. (2006). Marketing Management (12th ed.). New Delhi, India: Pearson Education.

(13.) Lonial, S., Menezes, D., & Zaim, S. (2000). Identifying Purchase Driving Attributes and Market Segments For PCs Using Conjoint and Cluster Analysis. Journal of Economic and Social Research, 2(2), 19-37.

(14.) Mazumdar, R. (1991). Marketing Research; Text, Applications and Case Studies. New Delhi, India: New Age International (P) Ltd. Publishers.

(15.) Michael, & Barnett. (2012). Bringing up the baby chains. Centaur Communications, 35 (44), 20-21.

(16.) Modi, S., & Jhulka, T. (2012). Consumer Buying Behaviour: Changing Shopping Patterns. IJBMER, 3(3), 527-530.

(17.) Mooi, E., & Sarstedt, M. (2011). A Concise Guide to Market Research. Springer-Verlag Berlin Heidelberg.

(18.) Mullarkey, G. K. (2001). The influence of brands in fashion purchasing process. University of Auckland; Business Review, 3(1), 56-71.

(19.) Nargundkar, R. (2004). Marketing Research-Text and Cases (2nd ed.). New Delhi, India: Tata McGraw Hill.

(20.) Pennsylvania, W. U. (2013, February 7). India's Kid's Clothing Market Grows Up. Retrieved October 2, 2013, from http://knowledge.wharton. article/indias-kids-clothing-market-grows-up/.

(21.) Saxena, R. (2002). Marketing Management (2nd ed.). New Delhi, India: Tata Mc Graw Hill Publishing Ltd.

(22.) Shukla, A. (2010, December 3). Kidswear market to touch Rs 58K cr by 2014. Retrieved October 2, 2013, from http:/ /

Dr. Sougata Banerjee, Assistant Professor, Dept. of Fashion Management Studies, National Institute of Fashion Technology, Ministry of Textiles (Govt. of India), Block-LA, Plot No: 3B, SectorIII, Salt Lake City, Kolkata--700098; e-mail:

Ms. Sarwat Pawar, Student, Masters in Fashion Management, Dept. of Fashion Management Studies, National Institute of Fashion Technology, Ministry of Textiles (Govt. of India), BlockLA, Plot No: 3B, Sector-III, Salt Lake City, Kolkata--700098; e-mail:
Table 1: Agglomeration Schedule

Agglomeration Schedule

Stage Cluster Coefficients Stage Cluster Next
 Combined First Appears Stage

 Cluster Cluster Cluster Cluster
 1 2 1 2

1 96 100 .000 0 0 3
2 94 99 .000 0 0 64
3 6 96 .000 0 1 54
4 14 95 .000 0 0 67
... ... ... ... ... ... ...
92 1 8 96.846 88 78 96
93 7 69 112.005 85 11 96
94 4 5 128.213 86 89 97
95 36 40 157.150 90 30 99
96 1 7 187.513 92 93 98
97 3 4 259.503 91 94 98
98 1 3 347.547 96 97 99
99 1 36 457.540 98 95 0

Table 2: Reformed Agglomeration Table

Re-formed Agglomeration Table

No. of Agglomeration Coefficients Change
clusters last step this step

2 457.540 347.547 109.993
3 347.547 259.503 88.04
4 259.503 187.513 71.99
5 187.513 157.150 30.363
6 157.150 128.213 28.937
7 128.213 112.005 16.208

Table 3: Final Cluster

Final Cluster Centres


 1 2 3 4

Price 3.29 2.18 1.00 3.93
Fashion 3.86 1.66 2.00 2.53
Size 1.19 1.37 5.00 2.13
Colour 1.14 1.39 5.00 3.40

Table 4: Number of Cases in each Cluster

Number of Cases in each Cluster

Cluster 1 21.000
 2 62.000
 3 2.000
 4 15.000
Valid 100.000
Missing .000

Table 5: ANOVA

 Cluster Error F Sig.

 Mean Df Mean Df
 Square Square

Price 17.494 3 1.044 96 16.750 .000
Fashion 25.683 3 .460 96 55.792 .000
Size 11.174 3 .286 96 39.093 .000
Colour 25.903 3 .468 96 55.406 .000

The F tests should be used only for descriptive
purposes because the clusters have been chosen to
maximize the differences among cases in different
clusters. The observed significance levels are not
corrected for this and thus cannot be interpreted
as tests of the hypothesis that the cluster means
are equal.

Table 6: Distances between Final Cluster Distances
between Final Cluster Centres

Cluster 1 2 3 4

1 2.478 6.170 2.856
2 2.478 5.265 2.911
3 6.170 5.265 4.435
4 2.856 2.911 4.435

Table 8: Final Clusters Attributes

Cluster Cluster Attributes

 Price/Range Fashion/Trend

Cluster 1 Moderately Less fashion-
 priced conscious
 sensitive customer

Cluster 2 High priced Very high
 sensitive fashion-
 customer conscious

Cluster 3 Very high High fashion-
 priced conscious
 sensitive customer

Cluster 4 Less priced Moderately high
 Sensitive fashion-
 customer conscious

Cluster Cluster Attributes

 Size/Fit Colour/design

Cluster 1 Very high size Very high color
 & fitting or design-
 conscious conscious
 customer customer

Cluster 2 High size and High color or
 fitting design
 conscious conscious
 customer customer

Cluster 3 Least size and Least color or
 fitting design
 conscious conscious
 customer customer

Cluster 4 Moderately size Moderately less
 and fitting color or design
 conscious conscious
 customer customer
COPYRIGHT 2013 Institute of Management Technology
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 2013 Gale, Cengage Learning. All rights reserved.

Article Details
Printer friendly Cite/link Email Feedback
Author:Banerjee, Sougata; Pawar, Sarwat
Article Type:Author abstract
Geographic Code:9INDI
Date:Jan 1, 2013
Previous Article:Editorial perspective.
Next Article:Does the investor's age influence their investment behaviour?

Terms of use | Privacy policy | Copyright © 2018 Farlex, Inc. | Feedback | For webmasters