The Use of Conjoint and Cluster Analysis For Preference-Based Market Segmentation

Abstract

Market segmentation implies dividing the market into smaller segments that are homogenous inside, and heterogenous in relation to one another. Market segmentation is a part of strategic marketing, together with targeting and positioning, preceded by marketing research and followed by creating, implementing and controlling the marketing mix. As well as being an essential field of academic research, market segmentation proved to be a very useful concept for managers. The market segmentation concept dates back almost six decades, but issues raised regarding market segmentation more than three decades ago are still relevant and the number of different approaches in answering these questions is high. Preference-based segmentation can also be viewed in this context. Differences in consumers' preferences are, in fact, one of the vital causes of market segmentation. Preference-based market segmentation can be conducted by using conjoint analysis in combination with cluster analysis. The benefits of such approach are multiple, as they give rise to opportunities for: formulating the questions in a manner similar to real-life purchasing situations; reducing the number of socially desirable answers; introducing several attributes of a given product thus enabling an insight into the relative importance of the product's attributes, and order of preferences for different levels of these attributes, and, based on this, defining different market segments; testing the significance of various other market segmentation criteria (sociodemographic, psychographic and/or behavioural) for profiling the established segments; increasing the comparability of results of different studies; using this approach in planned new product launches and on emerging and developing markets as well as studying the difference between expressed preferences and purchases, in terms of obstacles preventing preferences to transform into purchase phase. Besides theoretical aspects regarding market segmentation and preference-based market segmentation by using conjoint and cluster analysis, this article demonstrates the use of such approach in researching student population in Vojvodina (the northern province of the Republic of Serbia) and this population's preferences to yoghurt (250 questionnaires handed out, 123 returned filled in, 98 included in the final analysis). In addition to results at the total sample level, two segments were identified and differences between them were investigated. Implications to the questions raised in the theoretical section of the article were sought on the obtained results. The article also seeks to point to the benefits of using this approach in domestic conditions, where its use is almost negligible.

Marketing Research on Passenger Satisfaction With Public Transport Service in the City of Belgrade

“…their (dis)satisfaction determine the quality of services [7]. For making appropriate marketing decisions and improving the quality of services, passenger satisfaction has to be permanently studied [8][9].…”

Requirement analysis considering uncertain customer preference for Kansei quality of product

“…Cluster analysis is a common data-grouping method that reveals data trends and characteristics. Ward's method is a hierarchical clustering method that classifies the whole data in a well-balanced manner, and is often employed in marketing analyses (Djokic et al, 2013). As shown in Figure 3 (left panel), the present analysis of customer preference similarity yielded three customer preference clusters; Cluster 1 (with 7 members), Cluster 2 (also with 7 members) and Cluster 3 (with 6 members).…”

Promoting green buildings: Do Chinese consumers care about green building enhancements?

“…Socio‐demographic factors have been also wildly applied in consumer segmentation which has been confirmed a key aspect in marketing (Shim & Bickle, ). Market segmentation can be conducted using conjoint analysis in combination with cluster analysis which is designed to classify groups of entities that share certain familiar characteristics, has been an important and most fundamental tool in academic consumer studies and also applied marketing (Djokic, Salai, Kovac‐Znidersic, Djokic, & Tomic, ; Punj & Stewart, ). Presuppositions are not necessary in cluster analysis, while it supplies ways to apply post hoc, or cluster‐based segments, and the core idea is to let the data speak for themselves for seeking consumer groups who share common trade‐offs, attitude, preference towards products (Green et al, ).…”