Tal Shoshani

We examine how brands' sociopolitical involvement affects customer store visits. We assemble a unique, large-scale dataset that combines granular weekly foot traffic for approximately 150,000 U.S. stores operated by 47 major brands from 2018-2022 with detailed local demographic and presidential election data to characterize store neighborhoods and visitors. We then use Google News to identify instances of sociopolitical involvement associated with these brands, spanning issues such as firearm regulation, race, and sexual orientation, and arising from either corporate actions or employee behavior. To estimate the impact of each event, we apply a Synthetic Difference-in-Differences model combined with Coarsened Exact Matching, comparing foot traffic at affected versus unaffected brands' stores. On average, sociopolitical involvement leads to a 0.8% decline in store visits, with substantial heterogeneity across events. Corporate-initiated actions tend to generate more variable and less negative responses than employee-triggered incidents. Moreover, brands' subsequent responses—such as public apologies or employee training initiatives—appear to attenuate negative consumer reactions. Three case studies further show that these effects vary systematically with store neighborhood characteristics, including political preferences and racial composition. To our knowledge, this study is among the first large-scale empirical research to examine the effects of brands' sociopolitical involvement on consumer store visits.

https://papers.ssrn.com/sol3/papers.cfm?abstract_id=6149614

We examine the effects of short-term rental (STR) regulations on urban service markets, focusing on the restaurant industry. Our setting is New York City's 2023 STR regulation (Local Law 18), which took effect in September 2023 and triggered an immediate 80% contraction in active Airbnb listings citywide. Using restaurant-month–level credit and debit card transaction data across 19 major US metropolitan areas, combined with a matched difference-in-differences design, we find that the policy led to a 7.5% decline in restaurant total spending, a 5.0% reduction in transaction volume, and a 2.3% decrease in average spending per transaction. These effects are consistent with multiple reinforcing mechanisms: the sharp reduction in STR supply reduced tourist inflows and raised accommodation costs, while simultaneously depressing local income through lost host revenues and reduced employment in tourism-dependent sectors. We provide empirical support for this interpretation through three complementary analyses. First, the declines in total spending and transaction volume are concentrated among restaurants catering to non-local customers and higher-priced establishments, whereas average spending per transaction falls similarly across restaurant types—a pattern consistent with reduced tourist visits and tourists' discretionary spending as the primary channel and a modest contraction in local discretionary spending as a secondary one. Second, NYC airports experienced a 3.5% to 4.3% decline in passenger arrivals relative to other major US airports, confirming reduced tourist inflows. Third, neighboring New Jersey markets experienced a 10.1% increase in restaurant spending, suggesting partial spatial displacement of economic activity rather than outright destruction of demand. Taken together, our findings highlight that STR regulations can have economically meaningful consequences well beyond the housing market, underscoring the importance of accounting for downstream impacts on local service industries in urban policy design.

Mobile location data provide granular insight into consumer mobility, making them a valuable resource for targeted marketing. Yet this granularity can expose sensitive information, such as visits to healthcare facilities or places of worship, to unauthorized parties. We ask whether firms' need for actionable insights and consumers' desire for privacy can be simultaneously accommodated. We evaluate privacy-preserving data aggregation along two business dimensions: predicting future brand visits, the core audience-profiling task of location data providers, and identifying potential brand switchers, category-active consumers who allocate limited visit share to the focal brand and represent prime geo-conquesting targets. We also evaluate two privacy threats: re-identification of individual devices via a membership inference attack (MIA) and inference of sensitive personal attributes from non-sensitive location patterns via an attribute inference attack (AIA). We show that aggregating devices into homogeneous behavioral clusters preserves performance on both business objectives while substantially reducing re-identification risk. Models trained on compromised device data achieve above-chance accuracy when inferring sensitive attributes for the devices on which they were trained, but this accuracy does not generalize to the broader population. Our findings provide practical guidance to data providers and regulators navigating the privacy–utility trade-off in mobile location data.

https://papers.ssrn.com/sol3/papers.cfm?abstract_id=5168538

• Finalist, Wharton Innovation Doctoral Symposium Best Paper Award
• Funded by the Marketing Science Institute (MSI) and the USC Marshall Institute for Outlier Research (iORB)

Retailers’ efforts to monetize consumer location data remain dominated by inefficient protocols (e.g., geofencing) that customize marketing interactions based solely on app users’ current location. Although extant trajectory mining techniques can remedy these shortcomings, they require high-frequency location data, which poses severe risks to consumers’ privacy. The authors present a novel method to extract marketing value from low-granularity urban mobility data and demonstrate its use in analyzing gas station choice to value customers. The data, also used to infer gas station visits, contain 1.06 million location records on nearly 27,000 devices observed near selected retailers including gas stations during a six-month period in Staten Island, New York. The authors pool consumers’ mobility trajectories from several days to dynamically calculate the distance of stores from consumers’ anticipated trajectories. They then supplement the data with station-level daily fuel prices and estimate a conditional logit model to assess how consumers trade off gas prices versus store distance. In addition to a generally high station loyalty, the authors find that consumers strongly prefer not to deviate far from their common trajectories for fueling trips. Applying their methods in a predictive context, the authors infer the value of newly acquired customers to the studied gas stations to be between $3.00 and $7.59.

https://doi.org/10.1177/10949968231179148

• Lead article
• Finalist, 2024 Journal of Interactive Marketing Best Paper Award

In this paper, we show how retailers can use consumer mobility data to assess the relative performance of each store within their network. We use mobile location data from over 5M devices in Manhattan, NY to construct a weighted network of Starbucks stores as nodes, with the edge weights between any two stores reflecting both the overlap between the customers of as well as the distance between the stores. We then compute network centrality measures to capture consumption dynamics in the network. Finally, we employ these variables to train machine learning models predicting whether or not each store closed down during the 20 months following our observation period. Our findings indicate that including network centrality measures derived from urban mobility data using our methods can lead to a better identification of underperforming stores in a retailer’s network, revealed by subsequent store closure decisions.

Link to draft