Partial reservation policy for cost-effective inventory management incorporating customer preferences

Effective backorder management is essential for balancing costs and customer satisfaction in inventory systems. This study proposes a partial reservation policy that classifies backorders into two types based on customer choices: delayed fulfillment with compensation (Type-1 backorder) or immediate fulfillment without compensation (Type-2 backorder). The policy is applied to a continuous-review production–inventory system with both lost sales and backorders. The system is modeled as a quasi-birth–death (QBD) process to derive long-term performance measures. Numerical experiments demonstrate cost reductions of up to 30% in scenarios without lost sales, with up to 12% achieved even when no compensation is provided. In high-demand settings with lost sales and high lost-sale costs, reductions of 5%–6% are observed. The policy maintains stable performance across varying levels of production time variability, offering a practical solution for reducing operational costs in e-commerce and omni-channel retail environments.