CusToMer MigraTioN PaTTerNs : eVideNCe froM a NorTh aMeriCaN reTaiLer

© 2014 M.E. Sharpe, Inc. All rights reserved. Permissions: www.copyright.com ISSN 1069–6679 (print) / ISSN 1944–7175 (online) DOI: 10.2753/MTP1069-6679220302 The customer management literature has made significant progress toward evaluating the effectiveness of marketing activities using customer profitability as a basis. Customer profitability is an individual-level metric based on the discounted expected future cash flows minus the cost to serve a customer (Mulhern 1997). This metric has been successfully used as a proxy to estimate the market value of a firm (Gupta, Lehmann, and Stuart 2004), calculate the value of a company’s customer base (Johnson and Selnes 2004), and evaluate the return on marketing investments (Rust, Lemon, and Zeithaml 2004). More recently, research has demonstrated that improvements in customer choice models can be achieved by incorporating customer dynamics. Specifically, Netzer, Lattin, and Srinivasan (2008) extend a choice model by incorporating a Hidden Markov Model to account for changes in customer–firm relationships over time. This research was extended by Montoya, Netzer, and Jedidi (2010) to incorporate marketing activities in the pharmaceutical industry. Similarly, Mark et al. (2013) extend a hurdle model to incorporate both marketing and customer dynamics in a retail context. A consistent finding in this stream of research is that ignoring customer and marketing dynamics in models of customer behavior results in inconsistent estimates of the impact of marketing activities on buying behavior (Mark et al. 2013). Despite the significant progress in the customer management literature, very little research has been conducted to understand why relationships are formed and how they evolve over time. Fournier and Avery argue, “because relationships come in different shapes and sizes, companies need to be cognizant of the requirements of diverse types of relationships beyond the loyalty ideal. Different relationships require different ways of relating, and managers must adapt their CRM programs accordingly” (2011, p. 64). In their study, they conducted several interviews and collected surveys from 900 customers to develop a perceptual map of the various types of customer–brand relationships. Indeed, their map suggests customers build relationships with brands for a variety of reasons (i.e., weak and superficial or strong and deep). However, we argue that more research is needed to understand how these different types of relationships evolve over time. Our main objective is to address this gap in the literature by developing a customer dynamics framework that offers empirical insight into how different types of customer–brand relationships evolve over time using readily available transactional data. Furthermore, we assess how marketing activities influence the trajectories (e.g., increasing, decreasing, or stable) of these relationships over time. Our customer dynamics framework consists of four steps. First, we determine the number of latent behavioral Tanya Mark (Ph.D., University of Western Ontario), Assistant Professor of Marketing at the University of Guelph, Guelph, ON, Canada, [email protected]. Katherine N. Lemon (Ph.D., University of California, Berkeley), Accenture Professor, Boston College, Chestnut Hill, MA, lemonka@ bc.edu. Mark Vandenbosch (Ph.D., University of British Columbia), Kraft Professor in Marketing, University of Western Ontario, London, ON, Canada, [email protected]. CusToMer MigraTioN PaTTerNs: eVideNCe froM a NorTh aMeriCaN reTaiLer