Integrating neighborhood delivery services into parcel delivery networks

Problem definition: Leveraging developments in the sharing economy, several innovative delivery models have been adopted in the parcel delivery industry. One such innovation is the neighborhood delivery service model, where local residents receive parcels and deliver them within their neighborhood. We study the integration of neighborhood delivery services into a parcel delivery network. Methodology and results: We combine distributionally robust optimization with continuous approximation to build a model that captures the interaction between demand volatility, neighborhood delivery service capacities, and the delivery performance of a parcel delivery company. Efficient lower bounding procedures determine fleet size requirements that guarantee meeting service level targets for individual neighborhoods. An algorithm based on column generation solves the problem for a network of multiple neighborhoods. Analytical results show that average demand profiles, demand volatility levels, and service level targets determine the ability of neighborhood delivery services to reduce the fleet size and save total network cost. Numerical analyses with empirical data from a case study emphasize the important role of neighborhood delivery service capacity. Even with a modest capacity of 3.3% of the peak demand, the fleet size can be reduced by 4.0%. Larger fleet size reductions, of up to 24.9%, can be achieved when capacity of a neighborhood delivery service is 33.3% of the peak demand in the neighborhood. Interestingly, these fleet size reductions also translate into total network cost savings of 1.3% and 8.6%, respectively. Managerial implications: Our study reveals two key levers for managers to recruit and retain neighborhood delivery services and negotiate for higher capacities. Of those, making minimum compensation agreements is new to the literature and can be used more liberally than the other lever: increasing unit outsourcing cost. Furthermore, we show how managers should consider the average demand, demand volatility, and service level targets in recruiting neighborhood delivery services.