A mixed-integer optimization approach to E-tailing supply chain resilience through substitution and transshipment

Efficient order fulfillment in e-tailing networks is increasingly challenged by inventory shortages, dispersed fulfillment centers, and heterogeneous customer preferences. This study develops a mixed-integer programming model that jointly optimizes order allocation, substitution, and lateral transshipment while explicitly incorporating customer satisfaction through behavioral coefficients. The model is solved using a Genetic Algorithm (GA) and evaluated on a numerical case representing an online electronics retailer. The results show that dissatisfaction costs associated with product substitution dominate the objective function (around 72% of the total cost), whereas lateral transshipment accounts for less than 6%, indicating that well-designed substitution policies can substantially reduce the need for inter-center transfers. Sensitivity analysis on substitution-related parameters confirms that misestimating customer acceptance can noticeably increase total cost and alter the balance between substitution and transshipment. The findings provide actionable insights for e-tailers seeking to design resilient fulfillment strategies, improve inventory allocation, and maintain customer satisfaction under product shortages. • Develop an integrated model combining substitution and transshipment strategies for e-tailing. • Optimize order fulfillment using analytical methods that minimize cost and enhance satisfaction. • Incorporate customer behavior to evaluate substitution preferences in supply chain decisions. • Validate the model through numerical analysis to demonstrate efficiency and resilience gains. • Provide insights that guide e-tailers in balancing cost, satisfaction, and operational flexibility.