Influencer Marketing Augmented Personalized Assortment Planning: A Two-Stage Optimization Problem

Jing Yuan, Twumasi Mensah-Boateng, Shaojie Tang

Abstract: Assortment optimization presents a significant challenge for online retail platforms. Its primary objective is to create an optimal selection of products from a vast array of substitutes, which will be displayed to customers with the aim of maximizing expected revenue. The purchase behavior of customers is typically influenced by a choice model that determines the probability of purchasing each product from a given assortment. This paper extends traditional assortment optimization by introducing the integration of influencer marketing, a practice that involves enlisting influencers to promote products and enhance their appeal to customers. While conventional assortment optimization assumes fixed product attractiveness, our model enables platforms to strategically enhance the attractiveness of selected products through influencer marketing, thereby increasing revenue potential. Consequently, we present a novel problem formulation encompassing assortment and influencer marketing planning. Leveraging recent advancements in submodular optimization, we develop effective and efficient solutions for this joint optimization problem.