ASSESSMENT OF DECISION FACTORS THAT AFFECT FORECASTING ACCURACY IN REVENUE MANAGEMENT

摘要

This paper investigates decision factors that affect the forecasting accuracy of airline seat demand. Of particular interest are the aggregation levels of demand across fare classes and time horizons. Real life data is used to asses various decision scenarios. Several traditional forecasting techniques are evaluated in addition to a proposed Markov Chains based Transition Matrices (MCTM) approach. Findings suggest that differences in demand are not pertinent to forecasting accuracy and that a finely disaggregated fare class system enhances accuracy. Future research directions based on these findings include enhanced focus on lost demand and cancelled reservations.