| Yield management (YM) has been used successfully in the airline and hotel industries to improve both utilization and average rates paid by customers. Both American and United airlines reported that YM adds several hundred million dollars to the bottom line each year. The extension of these principles to general business situations is explored in this dissertation. Common threads are identified that unify all situations where yield management is currently used (perishability, ability to segment the market, fixed capacity). Although yield management is the commonly-used term, a new term is defined in Chapter Three which is considered more appropriate (perishable asset revenue management or PARM). Chapter Three also establishes consistency in the terminology, definitions, and the structure of yield-management problems by defining a generic yield-management problem, and developing its solution. A taxonomy with 13 elements is defined that encompasses all possible assumptions.;Chapter Four formulates and solves the discrete, generic PARM problem. Several extensions to this basic problem are solved including a PARM problem with uncertain show up, several different optimal pricing problems, and an approach to tackling the difficult PARM problem of greater than two price classes with diversion. The next chapter solves a related problem where the decision is to choose the capacity. The focus is on isolating the forecasting benefit of early discount pricing relative to single-level pricing. A decision rule is derived that determines when it is optimal to offer early discounts strictly from the standpoint of better planning. The effect of changes in the different inputs on the decision to offer early discounts is presented.;Chapter Six implements an optimal decision rule through a model of customer arrivals that derives the needed probabilities. Heuristic approaches are compared to the proper closing out of price classes to see how much of an improvement in expected contribution can be made. The values of input parameters that make the static 1-dimensional rule an excellent approximation to the static 2-dimensional rule are explored. Finally, the estimation of the parameters for the arrival model is discussed relative to the available data. |
