Language
English (en)
Date of Award
8-16-2024
Degree Type
Dissertation
Degree Name
Doctor of Philosophy (PhD)
Chair and Committee
Jacob Feldman
Committee Members
Deniz Aktürk, Naveed Chehrazi, Lingxiu Dong, M. Ali Ülkü
Abstract
This dissertation incorporates two models for handling product returns—a pressing issue for retailers—into assortment optimization strategies: the established Sequential Returns model (SRM) and the newly proposed Bulk Returns model (BRM). In the SRM, customers are assumed to purchase products one at a time, engaging in a short trial period after each purchase to familiarize themselves with the product. They can then choose to keep the product, exchange it to continue their search, or return it and abandon the search process. Conversely, the BRM represents a new perspective tailored to convenience-driven consumers who lack the time or patience for sequentially ordering the products that interests them. Under the BRM customers simultaneously order multiple products, try them each on, and then keep the one they most prefer, while returning the rest. This dissertation primarily focuses on the assortment optimization problem under these two returns model. The goal here is to select a subset of products to make available for purchase aiming to maximize expected profits. A key distinction between assortment problem formulations under the BRM and that of the classical assortment problem is the inclusion of return costs incurred for each returned product. Our findings reveal that the assortment problem under the BRM is NP-Hard, even in the absence of return costs. Moreover, we show that there exists an optimal polynomial-time algorithm for the assortment problem under the SRM when the return costs are homogeneous.
Through a detailed comparative analysis, we uncover the strategic implications of each model for online retail management. We first show that when the respective assortments problems are instantiated with the same set of primitives, the expected profit generated by the optimal assortment under the SRM is at least that of its counterpart under the BRM. This result suggests that retailers should encourage (or enforce through policy) their customers to make sequential purchases and returns. Next, to better understand the how the choice probabilities compare across the two models for a fixed assortment, we first show that the overall market share is larger under the SRM, implying a reduced likelihood of a no-purchase event. We then show under the SRM, customers are generally more likely to purchase highly attractive products, and less likely to purchase less attractive products, as compared to its bulk returns counterpart. Generally, these results suggest that customers benefit more from the sequential framework. However, the underlying assumption of common return disutilities across the two models is challenged both on the spot and in our computational experiments.
To complement our theoretical findings, we conduct numerical experiments using real-world data to measure the profit improvements afforded by the SRM. We also analyze whether these profit gains offset the potentially increased number of return trips required by customers, which is limited to one in the BRM. Considering the increased return disutility and costs, especially for shipping in the SRM, we explore whether the profit boost justifies the potential rise in return costs. By comparing the expected number of return trips and their impact on profits between the models, we aim to determine if the SRM's profit advantages outweigh the additional costs of more frequent returns.
In conclusion, this dissertation offers a comprehensive analysis of two returns models addressing the product returns. By introducing the novel BRM and juxtaposing it with the SRM, we offer key insights into how to make the optimal product recommendation decisions and which types of return behaviors should be encouraged or discouraged for increased profitability and customer satisfaction amidst the challenges of return costs and return inconveniences. Our research advances the theoretical framework of return management strategies and provides actionable advice for retailers aiming to achieve a balance between customer convenience and operational efficiency. Through theoretical analysis and empirical investigation, our findings significantly contribute to the field of revenue management, emphasizing the crucial role of return policies in shaping the future of online shopping and their importance in strategic retail planning.
Recommended Citation
Sahan, Sahika, "Incorporating Returns into the Assortment Optimization Landscape" (2024). Olin Business School Electronic Theses and Dissertations. 54.
https://openscholarship.wustl.edu/olin_etds/54