Can Purchase Behavior Predict Relationship Perceptions and Willingness to Donate

As arts organizations frequently seek purchases and donations from the same consumers, one of their challenges is identifying which consumers are likely to become donors in the future. With all of the data currently available, many organizations are looking to their databases to predict future consumer behavior. In the customer relationship management literature, models have been developed that allow organizations to use consumer purchase data to predict future purchase behavior. This goal of this research is to extend this literature to determine whether it is possible to predict consumer willingness to donate using purchase behavior data. It is not expected that purchase behavior data will be nearly as accurate a predictor of future donation behavior as past donation behavior would be. Rather, the goal is simply to determine whether organizations would be able to use purchase behavior data as a proxy for relationship perceptions to improve their ability to identify prospective donors among their consumers. Three commonly used methods of prediction based on past behavior are tested using a combination of purchase behavior data and consumer surveys: a segmentation approach, a recency/frequency/monetary value model, and a Pareto/NBD customer relationship management model. Across all three methods, purchase behavior is found to be a poor predictor of both customer relationship perceptions and willingness to donate to the organization. © 2014 Wiley Periodicals, Inc. The nonprofit arts industry is one of only a few industries in which organizations are funded through a combination of earned and contributed income. In particular, the unique combination of selling tickets to performances or exhibits and asking for individual donations implies that arts organizations frequently seek both purchases and donations from the same consumers. While many individual donors are likely to be consumers of the organization’s products, most organizations find that only a small percentage of their consumers make donations of time or money to the organization. One of the challenges for arts organizations, therefore, is identifying which consumers are likely to become donors in the future. With all of the data currently available to arts organizations, many organizations are looking to their databases to try to predict future consumer behavior. In the customer relationship management literature, numerous models have been developed that allow organizations to use consumer transaction data to segment consumers and predict their future purchase behavior (Reinartz & Kumar, 2000, 2003; Verhoef, 2003). Particular focus has been placed on using past behavior to predict the duration or value of consumer relationships (Fader, Hardie, & Lee, 2005; Ho, Park, & Zhou, 2006; Reinartz & Kumar, 2000, 2003; Schmittlein, Morrison, & Colombo, 1987). This approach yields metrics such as customer lifetime value and expected lifetime that may be used by the organization to target consumers and allocate resources. These models have also been applied to donation behavior, and researchers have found that past donation behavior can be used to predict future donation behavior (Fader, Hardie, & Shang, 2010; Netzer, Lattin, & Srinivasan, 2008). Fundamental to this literature is the assumption that consumers’ observed behaviors are manifestations of their underlying relationships with the organization (Fader, Hardie, & Lee, 2005; Netzer, Lattin, & Psychology and Marketing, Vol. 31(8): 647–659 (August 2014) View this article online at wileyonlinelibrary.com/journal/mar © 2014 Wiley Periodicals, Inc. DOI: 10.1002/mar.20725