Distance and Local Competition in Mobile Geofencing

• Published in: Information systems research 2020-12, Vol.31 (4), p.1421-1442
• Main Authors: Ho, Yi-Jen (Ian), Dewan, Sanjeev, Ho, Yi-Chun (Chad)
• Format: Article
• Language: English
• Subjects: Advertising; Advertising agencies; Analysis; Bayesian analysis; Boundaries; Competition; Conversion; Datasets; Email marketing; Geofences; geofencing; hierarchical Bayes; location-based advertising; Marketing research; mobile; Product differentiation; purchase funnel; Working hours
• ISSN: 1047-7047; 1526-5536
• DOI: 10.1287/isre.2020.0953

The growing ubiquity of GPS-enabled smartphones has ushered in a new era of location-based services and online-to-offline commerce. Geofencing is one instance of this broader phenomenon, and it is being widely adopted in the context of retail, restaurant, entertainment, and other local services. By targeting users on mobile apps while they are in the vicinity of physical establishments, there is potential for higher levels of engagement and consumption of the products or services on offer. However, the level of consumer interest is likely to depend on distance from the establishment and local competition in the surrounding areas. This study examines the impact of these two factors on consumer response to geofence advertising at different points in the purchase funnel, namely the click stage and conversion phase. Analyzing a rich data set from one of the leading location-based marketing agencies and using a sophisticated Bayesian empirical methodology, we find that having one more competitor in the consumer’s vicinity reduces click-through rate by about 1%, and a 1-mile increase in distance is associated with a 17.6% reduction in conversion rate. These and other results suggest that accounting for distance and local competition in data-analytic mobile targeting would increase both the return on advertising spend and consumer welfare. This research studies the performance of geofencing, a practice where mobile users are targeted within a predefined virtual geographic boundary around an advertiser’s establishment. We argue the significance of distance (i.e., the mileage from a consumer to a focal establishment) and local competition (i.e., the number of alternatives in consumer vicinity) in ad responses. Drawing on the notion of the purchase funnel, we develop a two-stage hierarchical Bayesian model to examine consumer click and conversion choices. A unique data set of geofencing ad impressions is collected from one of the largest location-based marketing agencies in the United States. The results suggest that local competition matters in the click stage, whereas distance influences the propensity of conversion. Quantitatively, one additional competitor in the consumer vicinity zone lowers the click-through rate by 1.03%, whereas a 1-mile increase in distance results in a 17.64% decrease in the conversion rate. We also find a significant interactive effect, whereby a higher degree of local competition amplifies the negative impact of distance on the likelihoo