Ask or infer? Strategic implications of alternative learning approaches in customization

Learning about a customer's preferences is a critical first step in the customization process. Broadly, firms adopt two alternative learning approaches: (1) ask, i.e., solicit preference information directly from the customer (S-Learning), or (2) infer, i.e., deduce preference information based on past observations of the customer as well as those of other customers (O-Learning). Most existing research on customization strategy focuses on a firm's marketing mix decisions, implicitly assuming away how the firm learns about customers. We contribute to this literature by examining how a firm's use of a specific learning approach impacts competition, particularly its rival's choice of learning approach. We find that O-Learning provides a credible signal for relaxing price competition, while S-Learning does not. Further, S-Learning by a firm creates a disincentive for rivals to also invest in S-Learning. We survey business customers and find significant evidence supporting our theory. We conclude with several managerial implications of our theory including how a firm can optimally select its learning strategy in order to impact its competitive environment.