Evaluation of delivery cost approximation for attended home deliveries

Abstract

Abstract In the competitive market of online retail, customers and retailers mutually agree on delivery time windows for order delivery. Since overall information on customers and their desired time windows becomes only available incrementally during the booking process, decisions on the acceptance of customers are challenging: customers arrive step-by-step on the retailer’s website and expect immediate feedback on the availability of desired time windows, which requires effective solutions in short run times. The retailer’s aim is to accept as many customers as possible to stay profitable, and fixed delivery capacities have to be considered when customers request for delivery. Customer acceptance mechanisms provide guidance for retailers to estimate the delivery costs and hence the number of customers that can be accepted without exceeding available capacities. Within this paper, we present a new customer acceptance mechanism, which is based on a cost approximation approach from the literature. We perform a computational study to evaluate the presented approximation method and compare it to customer acceptance mechanisms from the literature. We consider various demand scenarios, partially obtained from real order data from a German online supermarket.