Optimal Reservation Policies and Market Segmentation

When operating in a market with heterogeneous customers, a service firm (e.g., a car rental company or a hotel) needs to manage its capacity so as to maximize its revenue. To gauge the potential demand, a service firm often allows each customer to reserve a unit of service in advance. However, to avoid the loss associated with “no-shows”, service firms may require a non-refundable deposit. To determine an optimal reservation policy with a non-refundable deposit, we consider the case in which the market is divided into four segments (high vs. low valuation and high vs. low show-up probability). When customer demand and the firm's capacity are large so that they can be approximated by continuous values, we determine the optimal reservation policy analytically, and we establish analytical conditions under which the firm should discriminate against (i.e., price out) certain customer segments. For the case when customer demand and the firm's capacity are finite so that they take on discrete values, we find that some of the insights obtained from the “continuous” case continue to hold especially when the firm's capacity is large. However, the key difference is that in the former case, the firm discriminates mostly based on customers' valuation, whereas in the latter case it discriminates mostly based on customers' show-up probability.