Hèléne Millet and Ann Cederhall have moderated the "Growing Revenue in Travel" Club for nearly a year now on Clubhouse and will soon be transitioning to LinkedIn Live. Recently the club discussed how RM is reinventing itself during difficult times. Hèléne shares her insights from the New Perspective for Revenue Management discussion.
Let’s make it simple.
Decades ago, when Revenue Management was invented, the whole point was to fill planes with the “best” passengers, meaning the ones bringing the most revenue to the airline.
That’s not easy because the passengers that pay less usually book first, so you must keep seats available for the most valuable ones. So, you need to forecast how many of those will show up.
A story of figures, probabilities, forecast and optimization.
And it has been like that since the beginning. All improvements were made in getting better forecasts (down to PNR, working at the airports to improve boarding figures), richer data (including competitors’) or more accurate models (from segments to O&D, etc.).
The fuller the flight, the more useful this “choice of the best passenger” is.
And the base unit for optimization is the seat.
So, what happens when a crisis such as COVID happens?
Schedules change all the time, borders are opened and closed and less people travel, especially on long haul.
Business travel crashes.
Booking curves change dramatically.
In a nutshell: historical RM is questioned.
From there two approaches are possible.
The Two Approaches to RM
One is to carry on with the current positioning, and find new sources of data (shopping data, “intentions” as captured through customers messages on social media, etc.) and new models relying on the present rather than only the past (AI and Machine Learning are buzzwords but those technics have proved excellent results for forecasting accuracy).
The other approach includes new data and models, but also questions traditional positioning.
For reference, read Ricardo Pilon’s article AI and Ticket Personalization
We did not discuss sources of data, nor models: our attendees were not data scientists; nor the fact that we might have called the session “new perspectives for pricing and revenue management.” We dealt with dynamic and continuous pricing in previous sessions.
"For most airlines, the RM department remains unbalanced, with most RM staff managing seats, while a couple of agents deal with ancillaries, whereas the expected revenue is more 50/50."
Our discussion focused more on the second approach to take.
The first issue with current RM usage is straightforward: optimization should be done based on the wallet (what the customer buys: seats and services) versus only on seats.
For most airlines, the RM department remains unbalanced, with most RM staff managing seats, while a couple of agents deal with ancillaries, whereas the expected revenue is more 50/50.
To balance the teams better, we could consider a seat as nothing more than a component (among many others) of the wallet.
To push the model, the goal of an airline could be to optimize the commission it gets from its retailing activities (percentage on hotels, cars, translation services, duty free, etc.) and set the flight as a commodity only, that could even be sold at zero.
For reference, read Oliver Ranson’s article Airfare Arbitrage
This changes perspectives….
The second issue is almost philosophical: with airlines being more and more customer centric, can a function as strategic as RM remain in its ivory tower?
The answer to this quite provocative (and not entirely fair, to be honest) question is obviously negative. Traditional revenue management should open up, if not done already, not limiting its investigations to figures (considering the customer) and sharing its findings with other departments internally.
At the B2C Level
The value of the customer should be known and used as a premium input. We already considered that topic when working on subscriptions and considering customer lifetime value: the cost generated by a subscribed passenger on a very full flight is compensated by the additional revenue her/his subscription generates on empty flights, or when she/he naturally chooses the airline over the competitor no matter what.
The airline needs to take into account any knowledge it may have of the requestor, the FFQ tier being the most obvious. What if you are a Platinum member of an Alliance X and travel on an Alliance Y flight: shouldn’t Y try to make you switch?
Of course! But it usually does not. It is accurate to question in a very practical way, how could Y know your status and even more, your preferences?
"NDC facilitates personalized offers distribution. The question of dynamically building an offer remains though, and the industry moves to that goal with baby steps."
GDPR is not helping but this is worth investigating for valuable passengers. Note that those are often corporate passengers, and the switch of alliances (at least airlines) is often done at the corporate level rather than individual.
In the same way, an airline should be capable of offering packages adjusted to what the customer is looking for, without forcing segmentation (still reflected in branded fares) and behind the scenes, buckets.
NDC facilitates personalized offers distribution. The question of dynamically building an offer remains though, and the industry moves to that goal with baby steps.
Breaking internal silos: It is not trendy anymore to compare an airline with a series of little shops. With the crisis and even long before, the revenue management department started to work with the rest of the departments within an airline: to increase flexibility, to help with planning (where to put planes, when only a few are operating) and operations (where to cancel flights [when on strike or on a snowy day]) and to ensure that decisions are consistent across the company.
Any Vision to Go Further?
Thinking of Google or Amazon ten years from now: no longer sell a seat, but the facility to fly from A to B, whatever the airline (no matter the brand), with a contract.
Considering revenue management as a function that could be outsourced: could an airline outsource everything but the operations of the aircraft? See for instance what Skytra (from Airbus initially) is proposing in forecasting airlines' future revenues, as a way to secure their market value. Could it be a model to reinvent RM?
Let’s be honest. This is difficult to stomach for the aviation lovers we are.
Revenue management is considered part of the core for airlines, and breaking silos is harder to do with external partners.
The businesses of airlines look like a tangled web where a big bang is exceedingly difficult, if possible, at all: partners (other airlines, travel agents, tour operators, etc.) are numerous, and guarantee too high a level of revenue to be ignored. It is all at once or nothing. And that’s probably why the whole system is so stuck.
But it is worth opening Pandora’s Box. It is a good way to find ways out. Let’s continue this fascinating brainstorming.