Searching to maximize revenue in unrestricted and less‐restricted competitive fare environments: Combining hybrid forecasting and fare adjustment with various unconstraining methods

In their competitive environment, airlines are always searching for ways to maximize revenues. This article considers both less‐restricted and fully unrestricted fare environments. We use the sophisticated Passenger Origin–Destination Simulator (PODS) to examine the revenue impact of four implementations of hybrid forecasting (HF) and fare adjustment (FA) in combination with three different methods of unconstraining – Expectation Maximization (EM), Projection Detruncation (PD) and Booking Curve (BC). Owing to the competitive nature of PODS (two airlines competing head‐to‐head for customers) and its allowance for customer choice, we are able to assess all the implications of these HF and FA implementations in combination with unconstraining, including the impact of spill, upgrades and recapture. We find that in the less‐restricted fare environment, under either EM or PD, the optimal combination is HF with FA equal to 0.5, and that under BC unconstraining the optimal combination is HF with FA equal to 1.0, independent of optimization technique (leg or network) or demand level (medium or high). For the unrestricted fare environment, we find that under leg optimization, generally it is best to use HF with FA equal to 0.5 combined with willingness‐to‐pay estimate curves of either Frat5c (EM/PD) or Frat5a (BC). Under network optimization, it is best to use HF with FA equal to 1.0, combined with WTP estimate curve Frat5c for all three unconstraining methods. When using realistic booking data from major global airlines to calibrate the PODS D6 network, we show that becoming more aggressive with HF and FA can lead to significant revenue gains of 1.6–5.8 per cent (less‐restricted environment) or 3–15.9 per cent (fully unrestricted environment).