Cross-docking or traditional warehousing: what is the right distribution strategy for your product?

Abstract

PurposeThe objective of this paper is to investigate the impact of products' characteristics on the performance of three distribution strategies: traditional warehousing (TW), cross-docking pick by line (XDPL) and cross-docking pick by store (XDPS).Design/methodology/approachBased on a case study of an FMCG “Fast Moving Consumer Goods” company and a major French retailer, we empirically analyse the impact of the products' characteristics on the performance of the three distribution strategies. We consider a three-echelon supply chain composed of one supplier DC, one retailer DC and multiple retailer stores. The inventory at each echelon is controlled according to an order-up-to (OUT) level policy. The demand is forecasted by means of a single exponential smoothing method. A sensitivity analysis is also conducted to analyse the impact of the supply chain parameters on the comparative performance of the strategies when the parameters' values deviate from the empirical base case.FindingsThe empirical investigation shows that the use of XDPL results leads to an increase in the supply chain total cost, whereas XDPS reduces the cost. Moreover, we show that for a service-level target, cross-docking strategies should be selected for products with low variability, high shelf space, low value and short lead-time. For an inventory reduction target, these strategies should be selected for products with high demand volume. We also propose a managerial framework for choosing the right strategy for each product.Originality/valueThis paper fills a gap in the literature by presenting empirical results based on a real business case of a multi-echelon supply chain. Both cost and service are used to evaluate the performance of the strategies.Research limitations/implicationsOur work has the limitation to ignore the transportation cost implications when selecting the right distribution strategy. Hence, including such cost in the analysis would constitute an interesting extension of this work. Moreover, our empirical analysis represents a practical rich context that makes the scope for transferability of findings learned from this article substantial. However, for the generalisability of the findings, larger datasets in the retail supply chain would be interesting to consider