Geo-Tracking Consumers and its Privacy Trade-offs
65 Pages Posted: 5 Oct 2023 Last revised: 28 Mar 2024
Date Written: March 27, 2024
Abstract
Can geo-tracking data allow firms to better predict consumers' future behaviors? If so, how might potential privacy regulations limit the usefulness of geo-tracking data for prediction? Using data with over 120 million driving instances for 38,980 app users, and their visits to 422 restaurants in Texas, the authors quantify the extent to which geo-tracking data allow restaurants to better predict the number of visits one week ahead. They show that geo-tracking data increase the performance of prediction models by 14.77% relative to models that use demographic, behavioral, and static home location information. Simulation exercises that limit what data are tracked and in what form, where, and how frequently these data are tracked show a decrease in the predictive performance of models that use geo-tracking data. However, the decrease varies by the type of restriction; regulations that restrict what data are geo-tracked (i.e., summaries of driving behaviors) and in what form (i.e., synthetic data generated with nearby users’ data) result in the largest decreases in predictive performance (16.24% and 8.09%), while regulations that restrict where (i.e., within a few miles of a business location) and how frequently (i.e., at longer intervals) data are geo-tracked result in smaller decreases (3.56% and .77-2.46%, depending on the frequency). Importantly, models with restricted geo-tracking generally outperform models that do not use any geo-tracking information. These findings can assist managers and policymakers in assessing the risks and benefits associated with the use of geo-tracking data.
Keywords: Mobile app, location tracking, privacy, prediction, machine learning, transformers
JEL Classification: C23, C45, M1, M3, Z23, Z28
Suggested Citation: Suggested Citation