Effect of demand uncertainty on omnichannel distribution network design strategies

Abstract

To investigate the effect of uncertain demand on various omnichannel distribution network strategies and analyze the impacts of various online order fulfillment policies, a two-staged network optimization model is built. The two stages are a deterministic model to define the optimal distribution network and a simulation model to determine the optimal network flow to maximize the total profit for a given distribution network and online order fulfillment policy. This study shows the importance of taking demand uncertainty, requested delivery time, and customer service level (CSL) into account in designing an omnichannel distribution network. The results of the simulation model indicate that shipments from warehouse (SFW) and hybrid strategies more successfully cope with variation because there is a greater profit decline in shipments from store (SFS) strategies. The results also show that applying a dynamic assignment policy plays an important role in handling variation in demand. The study shows that, as demand uncertainty and CSL increase, having products available in central repositories becomes even more important to cope with demand variation. The study also indicates SFS strategies that seem more profitable in the assumed-to-be deterministic model might be misleading because SFS strategies become less profitable (compared to SFW strategies) when demand uncertainty is taken into account. © 2023 The Authors. International Transactions in Operational Research published by John Wiley & Sons Ltd on behalf of International Federation of Operational Research Societies.