Airline revenue management (RM) has evolved in three stages and we are now in a fourth. The first, which lasted from the 1960s to the 90s, was all about pricing. Recognising that all the seats on big jets like the 747 could not be sold at the traditional high fares, airlines introduced super savers and other discounted tickets which had to be booked in advance and could not be exchanged or refunded freely. These powerful methods are still used by many airlines today, but have been augmented by more advanced techniques.
In the second wave of RM, airlines used inventory classes to manage the number of seats that could be sold on each plane on a flight-by-flight basis. This is known in the trade as leg and segment (“leg-seg”) RM and many airlines stick with this method. In the third wave, starting in about 2000-05, airlines began to offer fares based on a passenger’s complete itinerary, or their origin-destination (“O&D”) pair.
Under O&D RM way a passenger travelling from London to Bangkok via Dubai could be offered different availability on the Bangkok flight to a passenger originating in Paris. This was a really exciting development – an airline might want to differentiate by O&D to account for different local fare levels and willingness to pay, or as part of a promotion targeting Paris but not London. This method was being introduced at Qatar Airways when I was there between 2007 and 2012 so I have seen how it is done close-up.
At the heart of both the second and third waves of RM was a company based in Houston, Texas called PROS, which I believe stands for “Pricing and Revenue Optimisation Solutions”. Their platforms are known in the industry as being mathematically rigorous but both capex and opex expensive. Unfortunately it looks like as we move into the fourth wave of RM they may also be becoming obsolete.
The key to PROS’ success in RM’s third and fourth waves was what we know in RM as the Littlewood rule, which states in effect that you should not sell a seat for a fixed price today if you can sell it for more tomorrow. PROS were the experts in telling you whether or not this was true.
They did this using historical data about what the airline and the industry had done in the past. For example if people had been buying summer holidays to Disneyland in January for the last ten years, they would buy summer holidays to Disneyland again in January this year. An airline’s analysts augment this idea within a PROS system with specific adjustments, which PROS call “influences”, to reflect specific circumstances like the economic cycle, TV series and local trends.
If it turned out that a PROS forecast was wrong, it would be corrected quickly the next time their algorithms checked bookings received against those expected, which PROS call a “Data Collection Point” (DCP).
What PROS did not do was:
1. Update the influences automatically in real time
2. Use forward looking data from non-airline sources
3. Estimate willingness to pay
4. Show you what change and refund prices or other associated ticket conditions should be.
Each of these caused problems. Manual influences require airlines to maintain a team of analysts managing the PROS system who could otherwise be employed answering other RM questions, leading to inefficient RM organisations. Because the data is historical and not forward looking the algorithms may find it hard to deal with significant shocks like COVID, leading to revenue loss. And since willingness to pay and ticket flexibility is not evaluated the risk of airlines leaving money on the table by under-pricing their fares at peak times or for high value markets could be high.
Until 2019 these costs were acceptable for most airlines in exchange for the analytical rigour underling a PROS system. But now it looks like PROS might be becoming obsolete. Their problems are twofold.
First, the historical data they rely on is unlikely to be valid for decision-making post-COVID. When you listen to people at PROS and read their website you realise that the principle of forecasting based on history is deeply wound in their DNA. I find it hard to imagine them being able to adjust their mindset to a world where history is indeed no longer valid.
Following COVID this is now the world we live in and PROS platforms seem silent on all the big questions in travel these days. Questions like:
· Will people book earlier or later than they did before?
· How will people’s booking behaviour change as borders re-open and in the years afterwards?
· What will happen to corporate travel?
· Will comfy class demand from the premium leisure segment come at a higher or a lower willingness to pay than before?
· How will business trip length of stay and sensitivity to price levels and refunds change?
The second problem that PROS will face is that there are startups and scale-ups ready to take their place in the fourth wave of RM, which is all about answering the questions that PROS do not. Airlines have shown that they will pay large amounts for technology and investors will be paying sales teams to chase these budgets right now. As I said before, when you listen to people at PROS and read their website they seem quite determined to play the same record again and again – that demand is all about history. I worry that they are complacent and not ready to change.
But the next disruptive startup could be just around the corner. If you are a PROS customer today there could be a risk that if their product is indeed obsolete they might not even be in business to service your system a few years from now.
All in, I think that it is risky for airlines to buy a PROS platform right now and would recommend caution before any deal with this vendor. It is not too late for PROS though. They are a highly capitalised business who could transform themselves for the fourth wave of RM if the willingness is there. It will be interesting to see.
oliver AT ransonpricing DOT com