Welcome to “Distribution Week” here on The Cranky Flier. Today I’m looking at a brief history of how airlines sell their tickets, aka distribution. Tomorrow I’ll look at how the airlines want to sell tickets in the future. Then on Thursday I’ll wrap it up with a look at the hurdles the airlines face in making this a reality.
A couple weeks ago I wrote about the new Farelogix interface and promised I’d follow up soon with why you can’t use that today. But as I started digging in, I found that this issue was far more complex than even I thought it would be and I’ve been looking into this for quite some time. I’ve talked to airlines, Global Distribution Systems (GDSes), corporate travel managers, and more, and now I’m wrestling with the tough task of trying to boil this down into something coherent. As Cory Garner at American Airlines said, this is the kind of stuff a grad student will write a thesis on.
The problem is that there are a lot of different players in airline distribution and there are some very old, goofy business models that have created incentives to stifle innovation. Let’s start with this basic snapshot of how airlines sell tickets today.
The funny thing is that this model doesn’t have to actually change much for the airlines to do what they want to do. It’s the responsibility of each player that needs some tweaking to support the kind of innovation airlines would like to see. One thing is clear: you can’t stop innovation if it’s good for the end user. That’s the beauty of disruption. But when you’re a big player with a lot of clout, you can slow it down. In the end, you lose if you play that game, and I think that all players involved are going to realize that sooner or later. But we aren’t there yet, so it’s quite the slog.
Naturally, it’s hard to predict the future, but there are airlines out there that have a strong vision of how they’d like to sell tickets someday. And that’s where I’m going to take this tale tomorrow. But first, we need to look at how we got to where we are today.
A Brief History of Airline Distribution
A long, long time ago in a galaxy far, far away, airlines did everything by hand. You would call the airline or a travel agent for a reservation. They might not have been able to confirm immediately because they had to go to find the file where that flight availability was written down and then check off a seat for you. Pricing was fixed by the government, so it was never an issue of what the price would be – just if there were seats available.
As things got more complex, the airlines began working on a computerized reservation system (CRS). The first was American’s Sabre system and that was a huge leap forward because the airline could instantly look up availability from throughout its network. But soon American realized that if it could roll out its computers in a network to travel agents, then it could make millions. American knew that it would have to display all airline flights or it wouldn’t be compelling enough for agents to bother, but it could put its flights at the top of the list (using what today they call “bias”) and sell more seats, stealing share from other airlines. United did the same with its Apollo system and the race was on. Eventually bias was outlawed, though today it is legal in the US even though it’s banned in many other places around the world.
When deregulation occurred in 1978, airfares became much more complex and the CRS became necessary not just for availability but for performing these complicated calculations. The system that was in place back then is effectively the same that’s in place today. Fares were filed with a third party group called ATPCO while schedules were filed elsewhere. These fares were filed in specific “buckets” of availability. If there was a seat made available by the airline in the required bucket, that fare could be sold.
So the CRS (which morphed into a Global Distribution System – GDS), took the schedules and fare information from those relatively static sources and then sent a real-time request to the airline for bucket availability. Bringing those three pieces together, the GDS could calculate and push out all offers. The travel agent could make a decision and then book a flight with the GDS. When that happened, the GDS would send the booking back to the airline electronically. It’s quite an impressive system considering it was built before the internet.
Originally, regular travel agents were the only GDS users outside of the airlines. Travelers either had to call airlines or travel agents to book a flight. But then the internet magically appeared (or something like that). And online travel agents (OTAs) were set up to show the results from the GDS directly to travelers online. People could book directly, or so they thought. In reality, the same GDS processing was still happening behind the scenes – it just removed the human intermediary. For the GDSes, this was like a cash register. Every time anything was booked, the GDSes collected several dollars per flight segment from the airlines. It plowed some of that back into the OTAs to make sure they didn’t think about jumping ship to another system. They were able to mint money.
Of course, the airlines themselves built their own websites early, and have now succeeded in pushing a big chunk of their sales to those sites. These allow the airlines to bypass the GDSes and sell direct, and that has given the airlines some leverage at getting fees reduced. Throughout the 2000s, as airlines started changing the way they sold tickets from just “price and schedule” to include ancillary options, the websites became the only vehicles that could adequately support that. But it’s not just about having ancillary options displayed. It’s now about being able to know more about the traveler so airlines can tailor the offers they provide. If this sounds scary, it’s not a surprise. But it really is a good thing, and I say this as someone who certainly has concerns about privacy.
Tomorrow, I’m going to get into the details of how airlines would like to sell tickets and why it’s good.