Optimizing online grocery service: From customer understanding to multichannel profitability
基于一家大型杂货零售商的交易数据，研究如何通过优化在线商品种类和配送费来最大化多渠道盈利，发现配送费可高于成本、减少配送成本影响有限、在线商品种类并非越多越好。
Optimizing Online Grocery Service: from Customer Understanding to Multichannel Profitability • Empirically grounded store choice model guides online grocery strategy design • Retailers with low physical presence gain most from online service expansion • Delivery fees may exceed costs, especially for retailers with strong offline presence • Reducing fulfillment costs has limited impact on multichannel profitability • Large online assortments may not always be optimal due to fulfillment costs We study the optimal online service for grocery retailers operating both physical and online stores. The challenge lies in optimizing the size of the online assortment and the delivery fees to maximize profitability across channels, while considering customer, operational, and market dynamics. Using transaction data from a major grocery retailer, we employ an alternative-specific conditional logit model to investigate how delivery fees, assortment size, network characteristics, and customer needs influence store choice and spending across physical and online channels. We develop a profitability model that incorporates online service variables, customer behavior, and operational costs, enabling us to explore optimal strategies under various conditions. By identifying favorable conditions for the online store and analyzing optimal service variables, we provide actionable insights for retailers. Our findings challenge common practices in omnichannel retail. We show that delivery fees should not merely cover costs but can be strategically set higher, particularly for retailers with strong offline presence. Additionally, while reducing fulfillment costs improves profitability, its impact is smaller than expected. Multichannel retailers can offset these costs by passing them on to customers, with minimal overall demand loss, as some customers opt to shop in physical stores rather than abandoning the retailer entirely. Lastly, maximizing the online assortment may not always be optimal, particularly if the operational inefficiencies and costs outweigh the value customers place on variety. Our methodological framework provides retailers the opportunity to align their online services with customer preferences and operational constraints and to leverage customer data in shaping their omnichannel strategies.