The Impact of Customer Multichannel Choices on Revenues and Retention

There is broad agreement among researchers and practitioners of a positive relationship between multichannel shopping behavior and sales volume. However, this issue needs to be analyzed more thoroughly in two ways: (1) Is this relationship due to improved retention or higher revenues? (2) Which multichannel choice combinations have particularly strong impact, i.e. which hit the “sweet spot” of improving both retention and revenues? We develop a joint model of channel choice, quantity decisions and contract renewal (i.e., retention) to answer these questions. We follow a two-stage approach. Fist we represent channel choice and renewal through a multivariate probit model. Second, we model revenues (dollar spent) through a random effects Tobit model where all possible multichannel combinations (measured by channel-combination purchase probabilities derived from the probit model) are included as covariates. Data were obtained from a major European retailer operating across three channels – Internet, store, and catalog. We study a cohort of new customers whose transactions were longitudinally tracked for four-and-a-half years. We find evidence of a tradeoff between revenues and retention for some multichannel combinations. In particular our results show that certain multichannel choices increase revenues but decrease retention. On the other hand, there are some multichannel combinations that enhance both revenues and retention. In general, our results suggest that a vaguely defined strategy of encouraging customers to be multichannel could actually hurt retention, revenues, and hence overall sales volume.