A stochastic on-line model for shipment date quoting with on-time delivery guarantees

The paper introduces a new model for shipment date quoting with potential applications in e-commerce. First, a customer sends to the vendor a request for an item advertised at a certain price on the company's Web site. Upon receiving the request, the vendor immediately quotes the customer a no-Iater-than shipment date for the requested item, taking into account the amount of time to produce the item and any outstanding order previously placed but not yet fulfilled. If the quoted date is deemed acceptable, the customer subsequently places an order for the item; otherwise, the customer rejects the quote and looks for an alternative vendor (the deal is thus lost). The back-end of the quoting system is a single server production system. We propose heuristics that account for the intricate combinatorics of the server scheduling problem, as well as the uncertainty in customer demand and customer behavior.