Profit-Maximizing Assortment Planning under Uncertainty with Customer-Driven Product Substitution

Abstract

The evolving retail landscape, marked by rising production costs and resource constraints, necessitates the integration of supply chain considerations into assortment planning decisions. Traditionally driven by marketing, these decisions are intricately linked with operational aspects such as procurement, production, and inventory planning. To address the challenges posed by the unpredictable dynamics of today's market, we propose an integrative approach for profit-maximizing assortment planning in a make-to-stock environment. Our adaptive framework revises assortments over multiple periods within the planning horizon while accounting for multi-level uncertainties in supply, demand, and logistics. We employ a rank-based choice model to capture customer preferences and model dynamic substitution behavior, approximating stockout substitution probabilities based on small consideration sets. By endogenizing customer behavior, we effectively align customer preferences with supply chain decisions. This approach enables informed, data-driven assortment and supply chain strategies that maximize expected profits, improve operational efficiency, and adapt to volatile market conditions. We formulate the problem as a multi-stage stochastic program with stochastic and time-varying demand, represented by multi-period look-ahead forecasts commonly observed in practice. The proposed model is validated using data from a North American e-commerce furniture manufacturer and retailer. Results demonstrate the effectiveness of our approach in maximizing expected profit and enhancing resilience against supply and demand uncertainties, offering valuable insights for optimizing assortment decisions in volatile markets.