Data-driven ordering and transshipment decisions for online retailers and logistics service providers

Emerging advanced technologies enable logistics service providers (LSPs) to offer supply chain services, such as inventory management, rather than only shipping services. In the traditional retailer-led system, the retailer makes the ordering and transshipment decisions and outsources shipping to the LSP, but this paper studies an LSP-led preventive transshipment system where the LSP is the decision-maker, providing the optimal ordering and transshipment policy solution to the online retailer. An optimal two-threshold transshipment policy is proposed for the LSP-led system. The LSP has lower operational cost for transshipment than the retailer, so one may expect that the LSP can bear more risk of mismatch between the transshipment and uncertain demand, leading the LSP to be more aggressive in transshipment. Interestingly, the analysis reveals that the transshipment quantity chosen by the LSP is not necessarily high. The LSP will transfer less inventory between distribution centers than the retailer if the delivery cost or transshipment cost of the LSP, or the lost sale cost, holding fee, or selling price of the retailer are high. Based on real industrial data, the evaluation result shows that preventive transshipment in the LSP-led system can bring an average increase of 40.33% (even 93.31% for products with high demand uncertainty in the first selling period) in the total profit of the retailer and the LSP over the non-transshipment scenario. Importantly, even if the transshipment cost is the same for both the retailer and LSP, using the optimal policy decided by the LSP rather than the retailer will generate a profit improvement of 6.37%.