Measuring Multi-Channel Advertising Effectiveness Using Consumer-Level Advertising Response Data∗

Advances in data collection have made it increasingly easy to collect data on consumer-level purchases that are linked to those same customers’ advertising exposures. However, advances in advertising response modeling (i.e., marketing mix modeling) have lagged behind the availability of this granular data. Extending extant models to multi-channel consumer-level data, we develop a Bayesian Tobit model that can be used to measure the effectiveness of advertising exposures, while accommodating the typical sparsity of consumer-level response data. Building on the traditional ad-stock framework, we are able to differentially determine separate decay rates for each advertising medium. This allows us to estimate channel-specific shortand long-term effects of advertising. We demonstrate the applicability of our model on six data sets covering a variety of marketing platforms including television, online display, social media, e-mail and catalog. In each case, we show how the results provide insights into the interplay between different marketing channels in driving consumer response. We also illustrate how the model can be used to score and target individual customers on the basis of their advertising responsiveness, finding, in one case, that targeting the most responsive customers nearly doubles returns versus targeting the heavy purchasers.