The revenue management processes usually consist of four components: Forecasting, overbooking, seat inventory control and pricing ([McGill and van Ryzin, 1999]). In this test study at the Dutch low cost carrier transavia.com, a full subsidiary of Air France-KLM, a research is done to improve the forecasting component by testing demand forecasting methodologies with actual data some scheduled routes in a Matlab^TM environment. The new revenue management software has been used since August 2008 and has given new opportunities for the revenue controllers to optimize their flights. The forecasting possibilities are however not yet tested.

In a research study, multiple methods are tested with data of the carrier. The additive pickup method turns out to perform the best on several criteria, but a lot of adjustments need to be made in order to give a valuable forecast for the revenue controllers. Important is the dataset that is given as input for the forecast, but also the way fare classes are used within the low cost environment has to be dealt with. This paper suggests three modifications to the basic additive pickup and evaluates them eventually in a simulated environment for the revenue controllers.