B1294
Title: A large-scale mixed joint modeling approach for efficient digital coupon marketing
Authors: Gourab Mukherjee - University of Southern California (United States) [presenting]
Abstract: A large-scale joint modeling framework is developed for analyzing customer responses to different types of email coupons. We use finite mixture generalized linear models that capture latent heterogeneous subpopulations in the customer pool and subsequently estimate heterogeneous effects of different coupons after adjusting for customer types, preferences, attributes, and historical behaviors. Analyzing consumer responses to coupons from the apparel industry, we demonstrate the applicability of our method in formulating optimal coupon marketing strategies.