Dynamic Tiered Assortment Optimization

Due to the sheer number of available choices, online retailers frequently use tiered assortment to present their products. In this case, groups of products are arranged across multiple pages or sections, and a customer clicks on "next'' or "load more'' to access them sequentially. Despite the prevalence of such assortments in practice, this topic has not received much attention in the existing literature. In this work, we focus on a sequential choice model which characterizes customers' behavior when product recommendations are presented in tiers. We analyze different variants of tiered assortments by imposing "no-duplication'' and/or capacity constraints, and establish the hardness result on the computation of the optimal solution. For the offline version with known customers' preferences, we characterize the properties of the optimal tiered assortment and propose an algorithm that improves the computational efficiency compared to an existing benchmark. To the best of our knowledge, we are the first to study the online version of the tiered assortment optimization problem. In particular, we consider both non-contextual and contextual settings and quantify their respective regret bound. Lastly, we perform numerical experiments to demonstrate the efficacy our proposed algorithms