Simultaneous vs. Sequential: Optimal Assortment Recommendation in Multistore Retailing

Problem definition: We study a multistore assortment planning problem in which a seller centrally decides the assortment of products in each store. Each store is visited by a certain fraction of customers. Upon arrival, customers observe the products offered in the current store. The seller can either (1) show products offered in other stores to customers simultaneously (called simultaneous offering strategy) or (2) show products offered in other stores to customers if they decide not to purchase in the current store (called sequential offering strategy). Customers may incur a disutility if they choose products from other stores. We study the optimal assortment planning problem under each of the two strategies and compare their performance. Methodology/results: We show that under the simultaneous strategy, the optimal assortment for each store is revenue-ordered, and the seller cannot do better than if it operates each store separately. In contrast, adopting the sequential strategy can significantly increase the seller’s revenue. In particular, when customers’ disutilities of purchasing from other stores are universally homogeneous, the optimal assortment under the sequential offering strategy is revenue-ordered for each store, and the store with a lower (higher, respectively) demand should offer a larger (smaller, respectively) assortment to facilitate the sequential selling process. We then compare the two offering strategies in terms of the seller’s revenue and customer surplus, and we analyze the impact of parameters on the revenue comparison. We also consider several extensions (e.g., heterogeneous valuation, capacity constraint, and partial recommendation) and find that our main results still hold qualitatively. Managerial implications: The sequential offering strategy may be a better strategy for sellers in practice. We provide detailed guidelines for sellers to implement the sequential strategy, such as regarding the optimal assortment decisions and when sellers can exploit the most benefits. Funding: Yan Liu was supported by the National Natural Science Foundation of China [Grants 72293564/72293560]. Z. Wang was supported by the National Natural Science Foundation of China [Grants 72394361 and 72425013] and the Guangdong Key Lab of Mathematical Foundations for Artificial Intelligence [Grant 2023B1212010001]. Supplemental Material: The online appendix is available at https://doi.org/10.1287/msom.2024.0761 .