Frontiers in Consumer Search in Retailing
(A2025-124319)

AUTHORS

Siham El Kihal, WU Vienna; Raluca Ursu, NYU Stern; Eitan Muller, NYU

ABSTRACT

This special session centers around the theme of consumer search in online retailing. Even though consumer search is a mature field both theoretically and empirically, modern technologies are reshaping how consumers search for products. This session aims to expand the boundaries of consumer search research by introducing innovative methodologies, exploring emerging phenomena, and leveraging AI technologies to uncover insights that were previously unattainable. In the paper “Visual Browsing: How Consumers Search over Visual Content in Online Retailing” by Siham El Kihal, , Daria Dzyabura, Raluca Ursu, and Sumon Chaudhuri, the authors examine the relationship between customer search behavior and product visual information through individual-level customer search journeys on a fashion website. They show that the product image ranks as nearly as high as the price and product category. Additionally, they show that different visual features hold varying levels of importance at different stages—search versus purchase. Finally, to make these insights into actionable strategies, the authors develop a reinforcement learning model that optimizes product design by balancing visual feature importance at the click and purchase stages. In the paper “Product Return as a Sequence of Search Processes: Optimality and Search Duration” by Xiaoyu Fan, Srikanth Jagabathula, Zijian Liu, and Eitan Muller, the authors model the consumer’s return behavior as a search process using dynamic programming: The consumer first searches among a given set of products and decides which one to purchase. Then, after purchasing the product and examining it offline, the consumer decides whether to return the on hand product and continue on searching. The authors show that the optimal action of the consumer follows a sequence of Weitzman processes, each can be viewed as a generalization to the standard Weitzman process that is widely studied in operations, economics, and marketing. They prove the optimality of the entire sequence and study the parameters that will affect the consumer’s return behavior. They then introduce the notion of the degree of disappointment, which characterizes the online-offline discrepancy of a product. In the paper “Using GenAI to Study Search” by Raluca Ursu, Pranav Jindal, and Anocha Aribarg, the authors propose a new approach to analyzing clickstream data. Specifically, the propose to use GenAI to uncover even richer insights, by asking it questions such as what is the goal of the consumer when searching (e.g., search to buy now, search to learn about a future purchase), what search method is the consumer using (sequential or simultaneous search), is the consumer displaying a maximizing or a satisficing mindset, and is search compensatory or non-compensatory. Answering these questions can inform models of consumer search and can demonstrate an additional value of GenAI in helping managers develop marketing strategies.