Unlocking the value in product return data: Inventory management with sales dependent stochastic product return flows from multiple periods

In the fast fashion retail sector, handling product returns has become a significant challenge due to rapidly changing consumer preferences and high product return rates. These retailers are now inclined to consider product return flows in managing product inventories using detailed product return data. This study investigates an optimal inventory control policy for a retailer facing stochastic product returns from multiple previous sale periods to maximize expected profit during a single selling season. The problem is formulated using dynamic programming, and due to its computational complexity, we propose an Approximate Dynamic Programming value iteration algorithm using basis functions. Our proposed algorithm reduces the solution time drastically without a significant sacrifice from optimality. We quantify the value of leveraging detailed return information and demonstrate that our proposed model increases the retailer's profit by 9% in the base case and up to 31% considering other cases compared to a model ignoring such information, especially under decreasing product prices over time or per period order capacity constraints. Finally, using an extensive computational study, we propose managerial insights on how to best leverage the value in the product return data using advanced analytics for fast-fashion retailers.