As we all know, airline revenue management is generally not a long-term proposition. Fleet planning looks out 5-20 years; route planning spans 1-5 years; but revenue management typically focuses on the short-term. Although flights can be booked up to 365 days in advance, most bookings still occur 0-60 days before departure, and for revenue managers, often 90+ days constitutes “long-term”.
Although this observation applies to most airlines, hotels, and car rental companies, there are some notable exceptions. Gaylord Hotels, for example, caters to large conventions or group business (groups of 500 and up) that often plan years in advance. Thus, unlike most travel companies, Gaylord Hotels may well strive to be booked at 70-80% of capacity more than a year out.
In addition to the fact that most airline demand is “close in”, airline revenue managers also recognize that many of the factors that influence demand, including schedule, other airlines’ schedules, fares and fare structures, and market conditions, change all the time and are more stable, or at least, better understood 2-3 months out rather than more than six months out.
All airline revenue management systems, nevertheless, mechanically calculate demand by O&D or by “bucket” whether the flight is tomorrow or 12 months from now. They then dutifully allocate seats based on optimizing revenue on the flight or across the network.
Thus, the historically high-demand markets see extreme tightening of inventory even nine months out. If the flight is projected to be full, the system will allocate the scarce seats to the highest value passengers and close out the lowest fares and least profitable O&Ds. If the booking curve is truncated to higher fare passengers booking 30 days in advance, very few bookings will be allowed prior to that. However, this all assumes that the schedules, fares, and market conditions, as currently envisioned, will remain.
Does this make any sense?
Given that markets are dynamic – meaning fares, flights, and markets are all likely to change in a 9 month period - don’t we need a more robust airline revenue management solution than a point forecast based on inputs that assume so many characteristics of the status quo?
Here are four guidelines to achieve such long-term revenue management:
- Give the best markets a demand haircut
- Target a base load factor 90 days out (going into the prime selling period)
- Apply more macro forecast processes to validate the revenue management projection
- Accelerate system responsiveness to key changes (schedule, fares)
The Demand Haircut
With the best markets more likely to deteriorate (as more capacity is added to the market over time or as demand shifts) than to further improve, it is appropriate to lower projected demand for far-out periods. A 10% “haircut” will often be enough to allow some bookings to occur before the 90 day mark and anticipate likely market changes.
It takes courage – or recklessness – to come to the 90 day mark with less than a 10% booked load factor. For the strongest markets, you may not want to book above 25% very far in advance but a small base is prudent given the dynamics of the airline industry.
Airline revenue management systems use a very micro approach – O&D forecast by time of day/day of week, all fare levels, and so on. This may be terrific for extrapolating current trends to flights next month but it is less suitable for a longer term forecast starting from zero bookings.
Gaylord Hotels, for example, doesn’t use a typical revenue management system for managing its hotel inventory. It uses an internally built model that recognizes longer term industry trends (new hotels, general economic indicators, etc.)
Airline planning departments, consistent with their longer term time horizon for routes and aircraft, produce forecasts by route/flight that can be useful in validating any long-term projections by the revenue management system.
Response Time Acceleration
Given that you anticipate change, don’t let change occur without an immediate reaction. Many airline revenue management systems take up to three months to adjust to a change – they are calibrated to sort out “noise” or demand blips that don’t portend a structural change in demand. When such a structural change actually occurs, it is important to insure the system incorporates that change immediately.
Airline revenue management departments and systems properly focus their efforts on the short-term – where most of the revenue lies and where the biggest sell-up opportunities lie. However, airline revenue management systems still produce forecasts, calculate optimal allocations, and close out low fares up to 360 days in advance of a flight departure. Airlines need to ensure these forecasts are robust and that prudence and common sense are applied to automated results longer-term, just as they do closer in.