Data-driven Inventory Management under Customer Substitution

Abstract

Most retailers that sell perishable goods offer multiple products in a product category (e.g., fresh food or fashion). Managing the inventories of these products is especially challenging due to frequent stock-outs and resulting substitution effects within the category. Furthermore, the true demand distributions of products are usually unknown to the decision maker. New digital technologies have enormously expanded the availability of data, storage capacity, and computing power and may thereby help improve inventory decisions. In this paper, we present a novel solution approach for the multi-product newsvendor problem. Our method is based on modern machine learning techniques that leverage large available datasets (e.g., data on historical sales, weather, store location, and special days) and are able to take complex substitution effects into account. We empirically evaluate our approach on two real-world datasets of a large German bakery chain. We find that our data-driven approach outperforms the model-based benchmark on the first dataset and performs competitively on the second dataset.

Keywords: inventory, multi-product, substitution, machine learning