Multi-echelon distribution network inventory optimization for cross-border omni-channel e-commerce

Purpose Cross-border e-commerce has rapidly grown, with many companies adopting omni-channel strategies to unify online and offline shopping experiences. However, in multi-echelon distribution networks, firms face high inventory pressure due to cost differences between domestic and overseas warehouses, logistics uncertainty and demand fluctuations. Unpredictable replenishment cycles often lead to overstock or stockout. Thus, optimizing inventory allocation to adapt to dynamic demand while balancing cost and uncertainty is a critical challenge for cross-border e-commerce companies. Design/methodology/approach This study presents a cross-border multi-echelon inventory optimization model to address challenges in cross-border e-commerce supply chains. The model is solved using the Adaptive Genetic Basin Hopping (GABH) algorithm, which accounts for uncertainties in logistics, dynamic demand fluctuations and cost differences across distribution nodes. This approach efficiently balances inventory levels, demand and costs, providing an effective inventory management strategy. Findings The comparison with the SAGA algorithm shows that, under the same parameters, the GABH algorithm uses less computation time and achieves a lower optimal cost, demonstrating higher computational efficiency. Sensitivity analysis reveals that demand fluctuations, replenishment cycles and inventory configuration significantly impact costs. Increasing safety stock reduces stockout risks and replenishment costs, although it raises holding costs, still ultimately lowers overall costs. Originality/value The study shows that the GABH algorithm improves inventory optimization by reducing costs and boosting computational efficiency. It identifies how demand volatility, replenishment frequency and safety stock affect outcomes, offering practical guidance for responsive, cost-effective strategies in complex cross-border supply chains and supporting strategic inventory planning under uncertainty.