Evaluating choice-based demand management strategies for day-to-day home delivery planning

As demand for online shopping and home delivery increases rapidly, courier companies often offer customer-oriented services, placing severe constraints on delivery planning and making routes inefficient. This study presents a framework for evaluating strategies to balance customer satisfaction and delivery efficiency in day-to-day home delivery planning. We first model the online shopping frequency choices by an ordered logit model and delivery timing choices by a nested logit model and estimate them using stated choice data. Based on the day-to-day delivery demand simulated by these models, we optimize a quasi-dynamic multi-period vehicle routing problem with a variable routing period ( d -MPVRP) and evaluate the delivery efficiency under various demand management strategies. Our results show that different demand management strategies, such as surcharges for highly-demanded time slots and discounts for frequent users, have different effectiveness across urban and rural contexts. The discount on the flexible delivery option for customers who frequently shop online allows consolidation of multiple delivery orders, and was particularly effective in rural areas. Our results suggest that effective demand management in day-to-day home delivery requires strategies considering geographic and demand characteristics to achieve delivery efficiency while maintaining customer satisfaction. • Choice-based demand management for day-to-day home delivery is studied. • E-commerce customers’ choices of order frequency and delivery timing are modeled. • Simulation-based evaluation framework integrates the choice models with a multi-period vehicle routing problem. • Normal delivery option gives couriers flexibility in delivery timing and improves efficiency. • Encouraging frequent customers to choose normal delivery allows for consolidation of multiple orders into a single delivery.