New demand forecasting models

The prediction of demand for air travel is a crucial part of the network management function and a vital prerequisite for every kind of network design or scheduling process. Since aircraft capacities are both expensive and perishable, a high seat load factor is indispensable for profitable operations by a network; a trial and error approach is not an option. A solid demand forecast is, of course, no guarantee, but it is the best conceivable way to achieve the optimum match between aircraft capacity and demand. Traditional forecasting methods rely on a blend of historic booking data and macroeconomic trends. A leading supplier of O&D-based booking data is Sabre’s database solution MIDT. Refined through experience-based calibration, the results are a sound basis for network design in a stable environment. The next generation of forecasting tools is already available, however: models that rely more on current economic flows than on historic data, assuming a true O&D perspective (door to door) instead of an airport-to-airport view and explicitly taking intermodal competition into account.