Inventory Management and Demand Fulfilment in Omni-channel Retail: The Role of the Store

Summary

Consumers increasingly buy at online retailers. Traditional retailers are challenged to also open online sales channels and thus become omni-channel retailers. This PhD thesis studies how the store inventory can be used to fulfil online orders. Although previous research mentions that using physical stores to also fulfil online orders could negatively impact the retailer’s costs and product availability, this thesis shows that with good inventory management, it could actually reduce costs and increase product availability. Using Markov Decisions Processes and Reinforcement Learning, this thesis identifies optimal order policies and inventory allocation rules in this omni-channel setting. The results show that the demand pooling effect that is created at the store level can be used to better match inventory and demand. Furthermore, when using multiple stores the performance improves further, as excessive stock at certain locations can be reduced while protecting lower inventories at other locations in the demand fulfilment process.