No-show forecasting: A blended cost-based, PNR-adjusted approach

Overbooking and no–show forecasting remain critical components of airline revenue management, even as the airline business model changes to meet low–cost challenges. In this paper, we examine the performance of a blended cost–based, PNR–adjusted approach to no–show forecasting, as compared to traditional methods of no–show forecasting. Our results show that this method generates up to 10 per cent revenue gains per available seat–mile compared to historical average no–show rates, and consistently outperforms expert no–show forecasts. We also estimate the revenue gains from overbooking to range between 15 percent and 18 percent of the total revenue gains from revenue management for a 33–day period in the peak summer season.