Revenue Management - On a New Shipment Information Record-based Method to Predict Show-up Rates in Air Cargo

摘要

Car rentals, hotels, cruise lines and airlines are important service industries offering perishable services with a high share of fixed costs and a capital intensive production. Unused capacity is irretrievably lost and causes significant losses. Therefore it is crucial to optimize the capacity utilization regarding load factors and skimming off consumer surplus with sophisticated acceptance pricing. These are the major tasks of revenue management (RM). As one of the major disciplines ofRM, overbooking aims to offset no shows by virtually increasing the aircrafts capacity. The most sophisticated overbooking applications have been developed for passenger airlines. Without major changes, these procedures based on Show Up Rate (SUR) forecasts have been adopted to air cargo. The forecasts work with historical data consolidated on flight level. Due to some methodical flaws, research and business applications face a situation of limited potentials for further improvement of these flight event based methods for air cargo. We suggest a new approach based on the Shipment Information Record (SIR) to address these limitations and benchmark this method against currently applied methods that work on flight event level using real world two year data from a major European cargo airline. We prove that this new approach is superior to flight event based ones, regarding the forecast method itself as well as the related overbooking process. These results therefore are the basis for the next generation of air cargo overbooking models and will find their way into actual business implementations.