Revenue Management with Customers’ Reference Price: Are the Existing Methods Effective?

Abstract

Existing revenue management methods and heuristics rely on specific demand-side assumptions such as customers’ independent decisions over time. We relax the assumption that purchasing decisions depend only on the current price and are independent of previous prices of the same or similar products. On the contrary, we assume that customers’ decisions depend on the product’s past prices through a reference price. With this new dimension, a firm needs not only to manage its remaining capacity but also to control the reference price to maximize its expected future profit. In this situation, we show that some of the main analytical properties such as monotonicity or modularity of the firm’s value function no longer hold. Consequently, the effectiveness of existing heuristics that have been developed based on these properties become limited and might generate poor results. Examples of such heuristics are the bid price and the booking limit, which are popular methods of revenue management in practice. For the new setting, incorporating a reference price, we update the existing revenue management heuristics and measure their updated effectiveness. We use a comprehensive simulation study to illustrate our results.