Influence of self-pickup incentives on last-mile deliveries: a route-optimization-based simulation approach considering consumer behavior

Official URL: https://dx.doi.org/10.1016/j.retrec.2026.101746

The boom in online shopping over the last decade has put immense pressure on e-commerce companies and logistics service providers, particularly on their last-mile delivery operations. Parcel lockers, which are smart cabinets with compartments of various sizes, serve as alternative shared delivery locations that enable order consolidation while providing customers with the flexibility of self-pickup, thereby reducing many inefficiencies associated with home deliveries. Nevertheless, incentivizing parcel locker usage remains crucial since most consumers still prefer home delivery. The role of such incentives in planning last-mile deliveries considering customer delivery preferences among their designated address and parcel lockers has received limited attention in the literature. In this study, we address this gap by performing Monte Carlo simulation. In each simulation replication, customer delivery preferences are generated using a Multinomial Logit Model, and the realized delivery points are used to determine delivery routes and schedules through a tailored Variable Neighborhood Search algorithm. Our numerical experiments show that incorporating parcel lockers into last-mile delivery operations substantially reduces the travel costs and fleet size, achieving average cost savings of 14.5% in urban settings. These results demonstrate that the advantages of parcel locker utilization reported in the literature persist even when uncertain customer delivery preferences and incentive payments are considered. However, the extent of these benefits may be strongly influenced by the location and deployment of parcel lockers in rural areas, highlighting a challenge for logistics service providers. Finally, our sensitivity analysis indicates that these benefits become considerably greater when customer delivery time windows exist.