Product Information Display Optimization with Quality Uncertainty and Returns

Problem definition: Display quality refers to product information such as the images and textual descriptions of product quality presented to consumers shopping online and can directly influence market demand during the sales phase and returns during the postsales phase. In this paper, we investigate the optimization problem on assortment planning, display quality, and pricing, taking into account the actual quality uncertainty and consumer returns. Methodology/results: We employ the widely used multinomial logit (MNL) model to investigate the impact of display quality on consumer choices. In particular, we incorporate display quality into the utility function regarding consumers’ purchase and return behavior, which is significantly affected by display quality. We then conduct a joint optimization analysis on assortment planning, display quality, and pricing; characterize the optimal policy; and propose an efficient method to find the optimal offer set that includes all the products with positive conservative adjusted keep rates. Moreover, we investigate the estimation problem and conduct a comprehensive empirical validation with real data. Our results reveal that model misspecification and optimization on display quality could result in substantial losses. Managerial implications: Because of its high impact on both consumer choices and returns, display quality alters the structure of the optimal pricing and assortment decisions. We demonstrate the robustness of our optimal policy in various scenarios, including continuous optimization, consumer reviews, alternative return modeling, heterogeneous consumer markets, and discrete optimization, under both exogenous and endogenous pricing policies. In addition, we investigate return policy design and the impact with display quality considerations. Funding: X. Yan was supported in part by the National Natural Science Foundation of China [Grant 72571256]. Y. Yu was supported in part by the National Natural Science Foundation of China [Grants 72091215/72091210 and 72188101]. Supplemental Material: The online appendices are available at https://doi.org/10.1287/msom.2024.1147 .