Learning Dynamic Selection and Pricing of Out-of-Home Deliveries

Home delivery failures, traffic congestion, and relatively large handling timeshave a negative impact on the profitability of last-mile logistics. A potential solution is thedelivery to parcel lockers or parcel shops, denoted by out-of-home (OOH) delivery. In theacademic literature, models for OOH delivery are so far limited to static settings, contrast-ing with the sequential nature of the problem. We model the sequential decision-makingproblem of which OOH location to offer against what incentive for each incoming cus-tomer, taking into account future customer arrivals and choices. We propose dynamicselection and pricing of OOH (DSPO), an algorithmic pipeline that uses a novel spatial-temporal state encoding as input to a convolutional neural network. We demonstrate theperformance of our method by benchmarking it against two state-of-the-art approaches.Our extensive numerical study, guided by real-world data, reveals that DSPO can save19.9 percentage points (%pt) in costs compared with a situation without OOH locations,7%pt compared with a static selection and pricing policy, and 3.8%pt compared with astate-of-the-art demand management benchmark. We provide comprehensive insights intothe complex interplay between OOH delivery dynamics and customer behavior influencedby pricing strategies. The implications of our findings suggest that practitioners adoptdynamic selection and pricing policies.