Multi-Stage Assortment Optimization for an Omnichannel Retailer
DescriptionOnline shopping is popular and many retailers operate both online and offline in order to stay competitive. Operating two retail channels in the form of an online store and a physical store lets a retailer experiment with new strategies to improve revenue. Instead of treating the two retail channels as competitors, a retailer can make joint operating decisions that utilize the unique strengths of each channel. Such a retailer is known as an omnichannel retailer. In this proposal, I study assortment optimization in the context of two strategies that can help an omnichannel retailer improve his total revenue across retail channels. The first strategy recognizes that customers are more likely to make impulse purchases in a physical store. A retailer can redirect demand from his online store to his physical store by offering a Buy-Online-Pickup-In-Store option with discounts, so that customers can choose to wait for the product to ship or pick up the product immediately at the store. When customers choose the latter option, the retailer may earn supplementary revenue if customers purchase additional products at the store. Apart from deciding on assortments for each retail channel, the retailer also selects a subset assortment that is available across both retail channels to offer under BOPS, in order to encourage customers to visit the physical store. The second strategy recognizes that the retail space to operate a physical store is expensive and limited. Suppose products consume some combinations of resources when they are sold. A physical store can only display a small assortment and carries limited inventory of resources to assemble products. Although an online store is not subject to these limitations, research shows that customers prefer visiting a physical store when the products can be experienced with the five senses. Hence, a retailer can use the smaller assortment in his physical store to advertise his online store. I consider the assortment optimization problem of an omnichannel retailer faced with customers who arrive over time to his physical store. The retailer offers each customer an in-store assortment, as well as a discounted, larger assortment if she switches to the online store. In essence, the retailer uses his assortments and promotions to redirect demand to the unconstrained online store. For each of these problems, I will present approximation algorithms with provable performance guarantees. I will extract managerial insights to understand the structure of the optimal assortments in the two retail channels.
|Effective start/end date||1/01/21 → …|