Research
2025
- Working PaperSteering Consumers’ Learning: Evidence from Stockout Substitutions in Curbside PickupAndrew Zeyveld2025
Items ordered for curbside pickup sometimes go out of stock. This obliges the store to choose substitutes on the affected consumers’ behalf. Using novel data from a supermarket chain, I show that these “stockout substitutions” influence consumers’ future purchases through the mechanism of learning. This presents the store with the following opportunity to increase its future profits: if the store selects substitutes from profitable brands that the consumers have never tried before, some of them will learn that they like the brands of their substitutes and then purchase these brands’ products in the future. However, I find that consumers are less likely to accept such substitutes than they are to accept substitutes from brands they have previously purchased. To quantify the trade-off between steering consumers’ learning and maximizing the probability that substitutes are accepted, I estimate a learning-based model of differentiated products demand. The gains from steering consumers’ learning depend on their respective purchase histories as well as the extent of learning in the product category.
@article{strategic_stockout_subs, title = {Steering Consumers' Learning: Evidence from Stockout Substitutions in Curbside Pickup}, author = {Zeyveld, Andrew}, year = {2025}, } - Working PaperDemand Estimation When Consumers’ Preferences Vary over TimeAndrew Zeyveld2025
This paper shows that workhorse demand systems fail to reproduce important substitution patterns when individual consumers’ preferences vary over time. This failure is rooted in the independence of preferred alternatives (IPA) properties of conditional and mixed logit, which restrict the relationship between consumers’ purchases and their preferences among unpurchased goods. To assess the empirical relevance of the IPA properties, I employ novel data from stockout substitutions in curbside pickup. For the two product categories that I study, I document substitution patterns that are inconsistent with the IPA property of conditional logit. As for mixed logit, its IPA property proves consistent with the substitution patterns in one of the two product categories. To quantify the benefits of relaxing the IPA property of mixed logit, I compare the model’s goodness of fit with that of mixed probit (which does not display an IPA property). In keeping with the descriptive evidence, the results of this comparison vary by product category.
@article{identify_demand_orders_and_subs, title = {Demand Estimation When Consumers' Preferences Vary over Time}, author = {Zeyveld, Andrew}, year = {2025}, } - Work in ProgressChoice Set–Dependent Preferences in Grocery Purchases2025
(One-Sentence Summary) We employ grocery scanner data to examine how the attributes of unchosen alternatives in a category (e.g., eggs) alter a person’s relative preference among products in that category.
@article{contex_dep_choice, title = {Choice Set--Dependent Preferences in Grocery Purchases}, author = {Conlin, Mike and Harris-Lagoudakis, Katherine and O'Donoghue, Ted and Zeyveld, Andrew}, year = {2025} }