Check out the on-demand sessions from the Low-Code/No-Code Summit to learn how to successfully innovate and achieve efficiency by upskilling and scaling citizen developers. Watch now.

Airlines have come under severe criticism recently for how they handle overbooked flights. And several high-profile PR disasters have caused many airlines to re-evaluate their approach for ejecting booked passengers.

Involuntary passenger bumping is actually a data problem that can be solved quite easily via improvements in data analytics and data transparency.

Let’s start with data analytics. Currently, how do airlines choose which passengers to bump from an overbooked flight? Judging from what we’ve heard recently in the media, it appears the current algorithm considers factors such as price paid for a ticket and the passenger’s frequent flier status. However, an airline has far more information about a passenger that it can factor into the equation. For example:

  • A passenger who booked the flight a short time before the date of the flight may indicate a business traveler with less flexibility than a holiday traveler.
  • A passenger who frequently travels to a destination probably has a better safety net of places to stay an extra night than a first-time visitor.
  • Someone travelling with small children will likely have a harder time making ad hoc travel and lodging arrangements if they are bumped.

This kind of information is available for any passenger, based on past travel history and the current booking. For some travelers – those who registered for the airline’s frequent flier program – the airline has even more information (e.g., demographics and preferences) that it can use to make more intelligent decisions when deciding who to bump.

But better data transparency could eliminate the need for bumping entirely.

Think of an overbooked airplane as a marketplace, where the airline negotiates with passengers over the cost of deplaning. Perhaps no passengers are willing to be bumped in return for the airline’s standard incentive of, say, $450. Yet for the right price, volunteers will step forward. That price might be a lot higher than $450, but compared to the passenger-bumping PR disaster that drove one airline’s stock price down to a quarter of its previous value, that price is still cheap.

There is currently no way to create a marketplace in which the airline auctions off the right to be bumped to the passengers willing to accept the lowest payment. But it just takes a simple mobile app to solve this problem.

Imagine the following scenario: You are seated on a crowded flight, when the flight attendant announces, “Unfortunately this flight has been overbooked, so we can’t take all of you. But, you have an opportunity to profit from our miscalculation. Please log on to our mobile app and tell us whether you are willing to be bumped and at what price. The bidding ends in five minutes, then we’ll select the lucky winners.”

All around you, passengers are engaged in hurried conversations with their fellow travelers and with colleagues on the phone to determine whether they have the needed schedule leeway. Many of them are lost in thought, trying to come up with the winning strategy that will get them bumped. Too low, and it isn’t worth their while. Too high and someone else will be selected. Some passengers are frantically trying to formulate joint bidding strategies with their neighbors. After five minutes, the flight attendant announces the names of the winners and the process ends on a high note instead of as a video on CNN.

If the airlines created a transparent marketplace, it could establish an environment in which bumped passengers are actually happy to leave the airplane because they participated in the decision. I believe that on almost all flights there are enough people willing to deplane for the right price, and that price declines in the face of competition from other passengers.

For those flights where there are not enough volunteers at any price, the airline can always fall back on the advanced data analytics described above to more intelligently identify the passengers least likely to have a major issue with the change.

Another problem solved thanks to data science. Now, about that carry-on policy …

Moshe Kranc is CTO at Ness Digital Engineering.

VentureBeat's mission is to be a digital town square for technical decision-makers to gain knowledge about transformative enterprise technology and transact. Discover our Briefings.