Evaluating warehouse strategies for two-product class distribution planning

Distribution networks often manage products with varying life cycles, where demand for some products is relatively stable throughout the year (basic) and the demand for others is short-lived (fashion). Beyond the coordination of inventory and transportation decisions, decisions at the warehouse must be considered as its resources are frequently shared by both product classes simultaneously. For this two-product class distribution planning problem, we focus on characterising three real-world distribution strategies observed in industry and evaluating them based on total distribution cost and warehouse measures (e.g. workforce plan and workload variation) against a benchmark ILS-based heuristic. Experimental results suggest that there are in fact strategies in industry that under specific system configurations may provide competitive solutions compared to the benchmark heuristic on large problem instances (e.g. 200 stores, 1000 products, and 28 days). Several managerial insights are derived to compare such distinct warehouse strategies and the corresponding impact on the network.