Why Changes in Airline Distribution Take So Long – A Look at What the Players Stand to Gain and Lose

Welcome back to the last post of Distribution Week here on The Cranky Flier. I know I’m getting really wonky here, and I apologize for that to my casual traveler readers. But this is something that really interests me, and I wanted to dig in deep. This post was the most difficult to write because trying to untangle the web of relationships is not easy. It got even more difficult when American and Travelport announced that some of these things will be implemented. If you missed the first two, see Part 1 which talks about the history of airline distribution followed by Part 2 which focuses on how some airlines want to sell tickets.

You’ve now seen the vision for how some airlines would like to sell tickets, but why hasn’t it happened yet? That’s the million dollar question. Ultimately, the relationships between all the different players are incredibly complex thanks to goofy business models and years of entwined history. It’s not easy for those things to change, but the shift is happening. Just yesterday, American and Travelport announced they had come to an agreement that will start them on the path to making this a reality. That’s a big deal, but it was just a matter of time before someone made the move. The potential benefit is there.

The funny thing is that it doesn’t seem like much has to change. The diagram I showed on Monday just needs a couple more lines (below), but it’s the nature of the connections that are so difficult to alter.

GDS Setup Proposed

I thought the most effective way to do this post would be to go through each group that has a stake in this thing to explain what this would mean for them and why they would or wouldn’t like it. This might help explain why we find ourselves still selling tickets the same old way. Let’s naturally start at the top, with the airlines…. (Grab some popcorn, we’ll be here for awhile.)

The airlines have a huge reason for trying to change the way things work today; they want to sell more at a lower cost. An airline’s website does this quite well, and those channels are moving full steam ahead, but there’s no way full-service airlines will ever get everyone to buy directly from the website. So they need to find a way to do what they do on their website with third parties.

Could they file all their ancillary rules through the central fare clearinghouse and then let all the third parties use that data? Maybe, but I really doubt it. Each airline will create ancillary options differently and a very structured way of formatting this wouldn’t necessarily be a good plan. Even if it was, the process would still rely heavily on the Global Distribution Systems (GDSes) to do all the processing. That would not only give the GDSes more leverage in continuing to maintain their position, but it would prevent the airlines from personalizing offers as they’d like because they wouldn’t control what the customer was given.

That’s why airlines like this new way of selling, because it would let them handle the processing themselves. Effectively, what they present on their websites could be delivered to any third party channel that wanted to sell travel. It’s in XML format so it’s pliable and can be used in all kinds of ways by third parties. That would help reduce distribution costs by reducing the importance of the GDS (but not eliminating its relevance, as we’ll discuss below). It would also give much more control to the airline.

But how can they do this? Well, the plan that’s moving forward would have a standard set up by the International Air Transport Association (IATA) unimaginatively called “New Distribution Capability” (NDC). Despite all the scare tactics I’ve seen put out by opponents, I’ve yet to see anything that suggests NDC will be anything but a standard for allowing airlines to communicate their offers to third parties and then take bookings. It doesn’t say anything about the content itself – just provides a way for it to be sent back and forth. IATA just filed this with the Department of Transportation for approval, so you can read all 87 pages with the background and details if you’d like.

How would this reduce costs? Well, development work could be focused on one channel so that improvements would impact the website, travel agencies, and third parties. It would also give the airlines more bargaining power with the GDSes. Why? Because it’s really hard to create a GDS today, but if the GDS becomes more of an aggregator of various feeds from different airlines, then it would reduce some barriers to entry for competition and put price pressure on the GDSes. The airlines like that, but of course, the GDSes don’t. And that brings us to…

Global Distribution Systems
Outside of the airlines, the GDSes have the most at stake with these changes, and different GDSes have taken different approaches. Today, the GDS revenue model is kind of backwards, and that’s a quirk of history. Let’s compare it to an ice cream company called, say, Mr Whippy. Mr Whippy has several suppliers from which it buys the ingredients that go into making its ice cream. Once it has made the ice cream, it sells to third parties (supermarkets, etc) for a profit. Those third parties then sell to the end consumer. It’s pretty simple to understand.

Now think about the GDSes. They are in the business of selling airline tickets (and hotels, etc, but we’ll just focus on airlines). They don’t have seats to sell themselves, but instead they have suppliers, the airlines. But instead of paying the suppliers for the inventory, they make the airlines pay them for the right to be sold in the system. Airlines pay booking fees that can range from a couple bucks to a lot more for each flight segment. (Recently, it was revealed that American was paying Sabre $2.73 a segment, but when they got into a business dispute, Sabre jacked that up to an incredible $7.31 a segment.) For a typical roundtrip with 4 segments, that’s a lot of money.

What do they do with that money? They go and pay the third parties (travel agents) that sell their products. Goofy, right? But it’s because way back in the day, the airlines owned the GDSes. And as we know from the history on Monday, the airlines could get an upper hand by getting their systems into agencies. They could bias the results in their favor. And of course, they wouldn’t want to charge those agencies. Instead, since they had the agencies captive, they could charge the other airlines to play. They were making money even if the agent booked another airline. Brilliant!

But things are different today because the airlines no longer own the GDSes. Though the business has changed, the model never has. For airlines, that means they pay a lot of money for something that they don’t think is providing enough value now that technology is starting to open up new ways of selling.

The GDSes, however, realize that they still currently control a huge piece of the market. Airlines make a large percentage of their money on business travelers, and big businesses use GDSes (we’ll talk more about that later). Though it varies by geography, in the US, Sabre is the big boy with over half the market. Travelport is next with something in the 30 to 35 percent range. Amadeus has the rest of the market (less than 10 percent), though Amadeus is huge in other parts of the world.

Sabre seems to have taken a stance that its role is to, as described to me by Shelly Terry, VP of Airline Merchandising for Sabre, “ensure we’re meeting the needs of the agencies and corporate customers that utilize our system.” It’s not just that. There was talk about balance between the supplier and customer needs, but Sabre seems to think that this new plan by the airlines would tip the balance too far in favor of airlines and would harm agencies.

From what I can tell, that means Sabre wants to keep doing the processing itself instead of letting the airlines handle it, because the company thinks that will make it easier to compare offerings across airlines. It paints a picture of airlines withholding fares from travelers unless they fork over their personal information. As discussed Tuesday, that idea seems silly to me.

But for Sabre, it has a lot to gain by maintaining its position. As the market leader, it wants to keep what it has, but it realizes it can’t sit completely still. Sabre has moved to allow for merchandising in small steps, even using connections that have an XML feed and allow the airline to process the sale. But those are built as one-offs for specific ancillary services instead of addressing the entire merchandising/selling strategy for the airline. Those one-off connections can be very slow to set up (over a year to sell US Airways Choice Seats alone – just imagine the labor costs), and this method forces those ancillary fees to be separate from the fares themselves in the shopping experience, by design.

Sabre wants to keep fares sold the way they are today because it thinks that’s a better way to compare across airlines. As it suggests, this new proposal from the airlines is about “changing the way fares and pricing information is made available to consumers and travel agencies.” The disagreement is on whether that’s a good thing or a bad thing. And Sabre clearly thinks it’s bad.

Not all GDSes feel the way Sabre does. Travelport, for example, has introduced Agencia in Canada, and then just yesterday announced it would do something similar with American here in the US.

Travelport Agencia Fare Display

Agencia incorporates Air Canada’s direct connect content with traditional GDS content in a very user-friendly interface. Travelport does all the processing for the airlines in the GDS, but Air Canada handles processing for its bookings. You can see a nearly 11 minute video on the Travelport website. What’s even more amazing? I understand that agencies actually pay for Agencia. This flips the current model on its head and makes it closer to a more traditional business/supplier model.

Travelport will do the same thing for American in the US with one big difference. It won’t force the desktop upon agents. The plan is to find a way to offer all these various options through the existing interface, though possibly in other ways as well. It is VERY early in this process, but I imagine we’ll hear more soon. I was recently invited to a conference in April to learn more about this and I’m hoping to attend. I will keep you posted if that comes together, naturally.

For Travelport, this is a great move. It’s a distant second in the US market, but by offering this, it will be able to provide a distinct benefit to agents that Sabre doesn’t want to offer today. Sabre has to be steaming. With Travelport offering to become an aggregator that puts direct connections right next to GDS options in one place, there is a real alternative to Sabre now. Though it’s too early to know, this announcement yesterday could be huge in terms of pushing distribution methods to change in the future.

Travel Agencies
We know GDSes seem to be split at best on this issue, but what about travel agencies? Again, it’s the same kind of thing. Many brick and mortar agencies rely on the revenues coming in from GDS incentive payments, so the thought of losing that is a scary one. The idea of having to actually pay for the content is even more harrowing. Of course, travel agents like the idea of having all this information so easily accessible. It makes the job much easier. But if you have to forego a lot of revenue, then it wouldn’t be worth the switch.

That’s why it really falls on to the airline to make sure the agencies are made whole. There has to be a reason for the agencies to take the plunge. And this is really why working with a GDS instead of trying to reinvent the model is so important for airlines. It’s why the Travelport/American announcement is such a big deal. If a GDS doesn’t want this to happen, it can cut off agency payments or start biasing airline displays (at least in the US) and that’s a really tough thing to overcome. But working with the GDS, that becomes a non-issue, especially if the new GDS agreement involves lower costs for the airlines in the process.

This goes for online travel agents as well. Priceline took the plunge and now works with American on a direct connect. How did it manage to do that? I would assume that American made it worth their while, but Priceline also has a ton of business that’s not air-related. I can only assume it was able to use its heft to avoid any serious loss. And now it can sell American’s Preferred seats while others can’t… yet. Though we know that’s changing.

Corporate Business
Often called “managed travel,” this is the bread and butter business for any airline. These are usually the high fare-paying road warriors that airlines love, but above those road warriors is a large organization, usually one that has staff overseeing the travel program which is often administered by an agency (called a Travel Management Company). So how the corporate folks feel about this should matter greatly. How do they feel? I had the chance to talk to two people who run the corporate travel groups of very large organizations. For them, there are really three things that matter most.

1) Security – For any business, security is important. But when it comes to travel, you’re freely passing around credit cards, birthdates, passport numbers, etc. No, there’s no social security number, but there is still a lot of sensitive information flowing. For any technology to even be considered, the security has to be there.

2) Data – Big businesses run travel departments on data. They need all kinds of data in a way that they can slice and dice it to make decisions on a broad level. So any solution must be able to provide that kind of data. The GDSes provide a great deal of data, and that’s why corporates like them. They can get all kinds of data. So if direct connect doesn’t provide that same level of data to the GDS or some other aggregator to give to the companies, then it’s not going to fly.

3) Discounts – Businesses negotiate corporate deals with the airlines, and that usually means discounts with extra perks. This kind of information needs to be easily accessible when booking regardless of the source. And all the discounts need to be in one place with access to all flights and fares. People aren’t going to hop around to various tools depending upon the airline.

You would think that credit cards would have found a way to extract the data and attach discounts to that number by now, right? But they don’t. So GDSes are still the best bet for these companies to do what they need to do. Could they benefit from direct connect? Yes. If the airlines started including all the ancillaries and provided robust tracking data for everything, it could be beneficial. Then again, with a lot of these corporate deals, the ancillaries aren’t purchased anyway because they are frequently negotiated as part of the overall agreement.

I definitely heard some skepticism about airlines being able to pull this together. After all, they do some things that make things more difficult for the corporates. For example, there are different merchant codes used by some airlines for various ancillaries. It makes pulling the data together tough. But with Travelport integrating the direct connect, that’s the best of both worlds.

Last but not least, we have the traveler. Travelers don’t currently receive incentives from anyone, so they should be the easiest to sway with new technology. Ultimately, it’s the traveler who may stand to benefit the most here. A clean interface that allows people to truly compare apples to apples by knowing how much each ancillary option costs (if at all, depending upon status) would be a huge leap forward. And getting offers tailored to a traveler’s need is helpful if done right. Naturally, just because the airlines put the direct connect out there, however, doesn’t mean it gets to the traveler. That only happens on airline-controlled channels.

As we’ve seen with Priceline, this can be done by third parties today. Even with limited funcionality, what Priceline has done is good for travelers, and benefits will only increase as more features are rolled out. Tech work is definitely a hurdle. IATA’s NDC has yet to roll out and then there will be development work required. But the barriers will continue to fall in time. Something that the airlines want which also benefits most of the stakeholders isn’t going to fail. I just wish things didn’t move so slowly in this industry. But after this week, you probably have a better understanding of why that’s the case.

(Visited 1,817 times, 1 visits today)

Get Posts via Email When They Go Live or in a Weekly Digest

Leave a Reply

24 Comments on "Why Changes in Airline Distribution Take So Long – A Look at What the Players Stand to Gain and Lose"

newest oldest most voted
Great series, so much content and so much to think about. Nice work. I work for one of those large corporate customers. Because of our size we have discounts with American, Delta, United, Southwest and US Airways in the US (any many more worldwide). I am assuming that that all of the data aspects you describe is not going to be possible without the GDSes. We book through a dedicated travel portal, but frankly it sucks (it is from a major player you mention). Up until recently for one of the 5 approved carriers we could not reserve seats (name… Read more »
David SF eastbay

Seems one website with all this would be better then old school GDS systems. But to have one website would have the big brother fall back of one company controling it all.

But if each GDS used a website and with come standard industry formats, it would see the information could be presented better and things change faster then trying to reprograms decades old GDS systems.


Cranky, you neglected to mention one player, who is probably the biggest customer of air travel in the U.S. and also plays several other key roles: the U.S. Government. How do they figure into the mess?

David M

The government falls under the “Corporate Business” category in Cranky’s breakdown. They have negotiated fares in the form of the City Pair Program and they have their own web-based booking portals (E-Gov Travel Service and Defense Travel Service) that ultimately book through a GDS.


[…] Flyer is putting together a great series on distribution in the travel industry. Today’s post focuses on Global Distribution Systems and how they’re impacting the […]

Not surprised to hear you say this was the most difficult of these related posts to write. Be assured, it was the most difficult to read. As you point out, it is difficult to try to untangle the web of relationships. You probably do it as good as anyone to explain all this. But, do I, a potential customer of the airline industry, really care? Not really. Relationships exist in just about every industry, but do we discuss it, care about it as much as we seemingly do with the airline industry? I don’t think so. I, like many, want… Read more »
David M
These new direct connect options that better expose ancillary services should help you feel less like a sucker. By exposing the ancillary services and their costs, you’re better aware up front of what your options are. And if airlines know who you are, they can customize things based on status (free checked bags, for example). Cranky’s part 2 article from Tuesday had a great example of this. He had a quote from the CEO of Spirit Airlines, stating that while 25% of their bookings come through online travel agencies, 100% of their complaints come from the customers who book through… Read more »
Thanks, David. Cranky has done something on just about everything of interest here, and done it better than anyone else I’ve seen. My concerns do center on pricing, and availability of a seat at a price. There was a time when a potential customer could research all the carriers’s fares, make sure he/she met every term and condition applicable to a particular fare, and assuming there was a seat on the particular flight, got the fare. Now, that’s not the case. Trying to meet every term and condition of any fare is basically useless when the airline can simply say… Read more »
Nick Barnard

Airlines (in general) don’t have antitrust immunity. There are a few specific markets that they do have this, but that is the exception rather than the rule.

Nick Barnard
CF have any of the newer startups in the travel distribution business (i.e. Hipmunk) mentioned any thoughts about moving closer to being a full reseller? Right now they forward customers over to the airlines to book, but it’d be a cleaner experience for the customer to have people just book at one site. The big thing is moving the data to XML allows smaller disruptive competitors to emerge. We owe the VoIP telecom provider growth to Level 3 and other CLECs reselling services to other smaller innovative companies, opening this up allows a better spot for smaller companies to interconnect… Read more »

[…] it’s time to talk about how they’d like to sell them. Then Thursday we’ll look at why this is so slow to take shape. […]

Cranky, thanks for an excellent post. This is the type of post that I really enjoy reading, because it provides insights you can’t get anywhere else. While I agree with much of your analysis, I have to disagree with the claim that there?s no way full-service airlines will ever get everyone to buy directly from the website. Many people used to say that about all airlines, until Southwest came in and proved them wrong. In many small countries with only one or two major airlines, almost everyone buys direct, both leisure and business customers. The US airline industry is rapidly… Read more »
Have to agree with Cranky on this one. Even if all but the largest companies move to a “self service” travel system, those largest companies (my former employer included) hold a ton of influence because of the volume of business they do. We were strictly prohibited from booking our own air tickets, even if the price at say AA.com was less than what was offered on our travel portal. I don’t see this changing anytime soon, for the sole reason that forcing bookings through a single source gives HR the ability to more effectively monitor travel patterns and keep costs… Read more »
David L
RE: Mr. Whippy comparison. I continue to see these comparisons and they are not apples to apples. The supermarket pays Mr. Whippy $x. The supermarket then charges the customer $2x (actually any multiple they want, although Mr. Whippy will suggest the price it should be sold at). With the airlines, the 3rd party reseller doesn’t actually ever buy the airline seat, all they can do is sell it for the airline at $x (with no discount or markup – although some backend incentives do exist). Also the 3rd party can’t sell the seat to Bob, then have Bob give it… Read more »

It looks to me that you got paid by the airlines, a lot of a** kissing! Airlines are building monopoly!


Do you guys really think that airlines will lower their fare if they get rid of GDS?
They don’t care about passengers. Why should they.


[…] my series about why airline distribution takes so long to change. At the time, I mentioned that American had signed a deal with the Global Distribution System (GDS) Travelport that would allow it to sell tickets the way it does on its website. This deal would allow travel […]


From this passage, I still don’t understand why changes in Airline distribution need take long time, any idea why?