| Customer-choice behavior is observed in almost all revenue management (RM) situations, but is not considered in much of the existing RM literature. In this dissertation, we formulate and analyze several revenue management problems where customer behavior is modeled explicitly.; First, we consider revenue management for a set of flights between a common origin and destination with customer choice among the flights. We consider both pricing and seat allocation as revenue management controls, and formulate the problems in the Markov decision process (MDP) framework. The resulting MDPs are quite complex, owing to their multidimensional state and action spaces. We develop and analyze various bounds and heuristics. In a series of numerical experiments, we find that some of the heuristics have good performance over a wide range of test problems.; Second, we study the effect of strategic customer behavior on pricing, rationing, and capacity decisions of a supplier selling a single product. The product is sold in two periods at two possibly different prices, and the supplier may ration the quantity offered in the second (clearance) period. This situation arises, e.g., when airlines offer last-minute discounts on unsold seats. Some customers are strategic and respond to the pricing and allocation decisions by timing their purchases. We study the impact of customer behavior on supplier revenue under various assumptions on supplier capacity and pricing flexibility.; Third, we study a stochastic service system with multiple customer classes and no backlogging. We consider both dynamic pricing and admission control problems. We model customer diversion by assuming that the arrival rate of one customer class depends on the prices of all customer classes. We derive structural properties of optimal policies for special cases of both problems. In addition, we propose a fluid approximation model and formulate it as an optimal control problem. We build on the analysis of the fluid model to study the interplay between revenue management decisions and capacity choice of the service provider.; Finally, we summarize and describe some future research directions. |
