Predicting Mobile Advertising Response Using Consumer Colocation Networks

Abstract

Building on results from economics and consumer behavior, the authors theorize that consumers’ movement patterns are informative of their product preferences, and this study proposes that marketers monetize this information using dynamic networks that capture colocation events (when consumers appear at the same place at approximately the same time). To support this theory, the authors study mobile advertising response in a panel of 217 subscribers. The data set spans three months during which participants were sent mobile coupons from retailers in various product categories through a smartphone application. The data contain coupon conversions, demographic and psychographic information, and information on the hourly GPS location of participants and on their social ties in the form of referrals. The authors find a significant positive relationship between colocated consumers’ response to coupons in the same product category. In addition, they show that incorporating consumers’ location information can increase the accuracy of predicting the most likely conversions by 19%. These findings have important practical implications for marketers engaging in the fast-growing location-based mobile advertising industry.