Profit Maximization Through Customer Relationship Marketing

Measurement, Prediction, and Implementation

Lerzan Aksoy, Timothy Keiningham, David Bejou, Lerzan Aksoy, Timothy Keiningham, David Bejou

• Publisher: Routledge
• Year: 2014
• ISBN:
  9781317993322
• Edition: 1
• Topic: Commerce
• Subtopic: Commerce Général

Discover approaches to make customer relationship marketing more effective Profit Maximization Through Customer Relationship Marketing: Measurement, Prediction, and Implementation takes the various elements of customer centric marketing and brings them together using the latest research and case studies from various industries. Respected top researchers review and discuss research and concepts to provide practitioners, educators, and students with a deeper understanding of the wide range of issues relevant to customer centric marketing. This informative resource focuses on effective strategies and approaches to explain how companies can ensure that their marketing dollar achieves the highest return on investment (ROI). Customer centric approaches such as customer relationship marketing (CRM) aim to increase customer retention, acquisition, satisfaction, loyalty, differentiate customer value, develop customers via up-sell and cross-sell opportunities, and decrease costs. Profit Maximization Through Customer Relationship Marketing: Measurement, Prediction, and Implementation comprehensively explains how to make best use of customer information to better manage customer value and firm profitability. This valuable text also explains the importance of, as well as how to establish a reliable customer segmentation strategy. The book is extensively referenced and includes helpful figures, tables, and photographs to clearly illustrate concepts. Topics discussed in Profit Maximization Through Customer Relationship Marketing: Measurement, Prediction, and Implementation include:

• the goals of customer centric approaches
• various customer segmentation approaches
• cross-selling as a strategy for customer relationship management
• strategies to effectively use customer loyalty
• the value and cultivation of customer satisfaction and customer retention
• and more!

Profit Maximization Through Customer Relationship Marketing: Measurement, Prediction, and Implementation is an invaluable resource for practitioners, educators, and graduate students.

Approaches to Customer Segmentation

Bruce Cooil

Vanderbilt University

Lerzan Aksoy

Koç University, Istanbul, Turkey

Timothy L. Keiningham

IPSOS Loyalty

Summary. Customer segmentation has virtually unlimited potential as a tool that can guide firms toward more effective ways to market products and develop new ones. As a conceptual introduction to this topic, we study how an innovative multi-national firm (Migros Turk) has developed an effective set of segmentation strategies. This illustrates how firms can construct novel and inventive approaches that provide great value. A-priori, and custom designed post-hoc methods are among the most important approaches that a firm should consider.

We then review general approaches to customer segmentation, with an emphasis on the most powerful and flexible analytical approaches and statistical models. This begins with a discussion of logistic regression for supervised classification, and general types of cluster analysis, both descriptive and predictive. Predictive clustering methods include cluster regression and CHAID (Chi-squared automatic interaction detection, which is also viewed as a tree classifier). Finally, we consider general latent class models that can handle multiple dependent measures of mixed type. These models can also accommodate samples that are drawn from a pre-defined group structure (e.g., multiple observations per household). To illustrate an application of these models, we study a large data set provided by an international specialty-goods retail chain. doi:10.1300/J366v06n03_02 [Article copies available for a fee from The Haworth Document Delivery Service: 1-800-HAWORTH. E-mail address: <[email protected]> Website: http://www.HaworthPress.com> © 2007 by The Haworth Press, Inc. All rights reserved.]

Keywords. Latent class model, clustering, cluster regression, logistic regression, classification, conjoint analysis, random effect, multilevel model, inactive covariate, satisfaction

Introduction

Market segmentation can be defined as dividing a market into distinct groups of customers, with different needs, characteristics or behavior, who might require separate products or who may respond differently to various combinations of marketing efforts (Kotler & Armstrong, 1999). Some bases of segmentation that may be used include geographic, demographic, psychographic and behavioral. Other variables that may be used for segmentation include situational (e.g., purchase/use occasions), and customer preferences for products or specific product attribute levels. Effective segmentation usually requires that each segment be evaluated on certain criteria such as stability, growth potential, size, accessible, responsiveness, and whether the customers in that segment, and the marketing efforts directed toward them, are consistent with company objectives and resources (i.e., whether the segment is “actionable”). Segmentation is critical because a company has limited resources, and must focus on how to best identify and serve its customers. Individual customer segments are characterized by a certain degree of within-group homogeneity that helps ensure that the members of a segment will respond in similar ways to marketing efforts. This allows firms to more efficiently apply marketing resources to each segment. Of course, companies are motivated to undertake segmentation strategies only as long as these efforts provide a positive expected net payoff. In summary, effective segmentation allows a company to determine which customer groups they should try to serve and how to best position their products and services for each group. Consequently, segmentation is an integral part of the development of marketing objectives and strategies, where defining those objectives will generally include either (Ansoff, 1957; McDonald & Dunbar, 2004): (a) an analysis of how products should be sold or developed, based on an analysis of current customer segments, or (b) the identification of new segments as targets for existing products or for the development of new products.

Wedel and Kamakura (1998, Chapter 3) provide an extensive review of the literature on market segmentation, and carefully review each of several approaches, along with a discussion of the supporting statistical methodology. General approaches to segmentation include both a-priori and post-hoc methods.

1. A-priori segmentation methods require that segments be defined before data are collected. The segments may be determined using customer characteristics or product-specific information. Segments are then studied empirically using data that may provide additional customer information. In some cases, several alternative or overlapping segment bases, that were all defined a-priori, are compared and contrasted. The goal of such an analysis may be primarily descriptive (e.g., cross-tabulation, logistic regression), or it could include the development of models that use the predefined segments to predict one or more dependent variables.
2. Post-Hoc methods identify segments empirically through data analysis. Again the ultimate goal may be primarily to study the groups themselves, or it may be to develop a predictive model for a set of dependent variables.
3. There are also hybrid approaches that combine a-priori and post-hoc analyses (e.g., Green, 1977).

Objectives and Organization of the Following Sections

We will consider analytic approaches that can be used in each of these categories, but our emphasis will be on latent class models and other promising approaches for effective post-hoc descriptive and predictive analyses. We begin in the next section with conceptual examples of how one innovative firm has used customer segmentation. Then we begin our discussion of the analytic approaches to segmentation with a very brief summary of the most effective procedures for a-priori analyses. We refer to this as “segmentation based on supervised classification.” In this framework, the a-priori definition of the segments provides a data set that is “supervised” in the sense that each customer is already classified into a segment, and the goal is to develop a model that allows one to classify new customers. This is followed by a brief review of how various types of cluster analysis have been used in post-hoc frameworks. We consider general clustering procedures that are not based on an explicit statistical model, which are among the primary methods used in post-hoc descriptive studies, but we also briefly summarize important predictive clustering approaches. The final section on methods will consider general latent class models that are appropriate for either descriptive or predictive post-hoc analyses, but which are especially flexible and powerful in post-hoc predictive studies. We conclude with two brief sections: a section summarizing how conjoint analysis provides a framework for segmentation analyses, and a summary discussion.

Conceptual Examples of How an Innovative Firm Uses Customer Segmentation: Migros Turk T.A.Ş. in Turkey

A Brief History

Migros, currently the largest grocery chain in Turkey, was set up in 1954 via the joint initiatives of the Swiss Migros Cooperatives Union and Istanbul Municipality. Migros was founded for the mission of obtaining food supplies and consumables from producers under the supervision of the municipal authorities and to sell these products to inhabitants of Istanbul under hygienic conditions and at reasonable prices. In 1975, all of Migros shares were transferred to the Koç Group, one of the largest holding companies in Turkey.

Following this development, Migros engaged in a rapid expansionary strategy by increasing the number of stores in Istanbul, opening stores in other regions in Turkey and introducing a number of different store formats based on size and product variety. In addition, Migros introduced a number of stores under different brand names to cater to segments with a variety of needs. Şok discount stores were introduced in 1995 to expand the market and broaden the appeal to include the lower tier price-sensitive segment. In 1997, Migros also became one of the pioneers in cyber-shopping and introduced its virtual store, utilizing sophisticated infrastructure and technology.

In 2005, Migros merged with Tansaş, a successful local grocery chain, on the grounds that the combined company would be able to offer a value propos...