State of the GDS, Part 4: Travelport, the Rebuild

Eric Marketts

For about fifty years, booking a flight through any of the big three systems meant typing a string of nonsense into a green screen. Not English. Codes. Every travel agent, on every GDS, learned a private language of slashes and cryptic commands just to sell a seat. That cryptic terminal was the GDS. It was the product, and it was the same product at all three. Then a couple of years ago, the smallest of them looked at that fifty-year-old machine, the thing the entire industry was built on, and did what the two bigger, richer players would not. It threw its own version in the bin and started building a new one from scratch.

Steph Nell

Which, when you say it out loud, is a little insane. Because this is the smallest of the three. It is heavily leveraged, and its debt wall is the nearest of the three. It has the least room for error. And it is the one that chose to do the single most radical engineering thing any of them has done. The two richer rivals renovated. This one demolished.

Eric Marketts

Welcome back to The V1 Airline Retailing Report. This is State of the GDS, Part 4, the finale. I'm Eric Marketts.

Steph Nell

And I'm Steph Nell. Quick orientation if you are just joining, and then we are going to close this whole thing out. Four parts, one company at a time. Part 1 was the foundation, how the GDS got built as a weapon and forced neutral. Part 2 was Amadeus, owning the whole stack. Part 3 was Sabre, the survivor selling a limb to stay in the game. Today is Travelport, the rebuild. And then we answer the question this entire series has been circling. Can this industry even hold three of these companies.

Eric Marketts

And before we get into Travelport's bet, do the same thing we did with the other two, because the origin matters more here than anywhere. Part 2, Amadeus was built by four European airlines. Part 3, Sabre was the direct descendant of the machine American Airlines built. Where does Travelport come from?

Steph Nell

This is the richest lineage of the three, and almost nobody knows it. Go back to Part 1, the spin-outs. Travelport is not one airline's system. It is where most of the others ended up. Its brands are Galileo, Apollo, and Worldspan. Apollo was United's system, the very one United used in 1976 when it walked away from the neutral joint-industry effort and sold its own screen instead. Galileo was a consortium of nine European carriers, British Airways, KLM, Swissair, Alitalia, and more. And Worldspan was built by Delta, Northwest, and TWA. Every one of those was an airline-owned system. Amadeus and Sabre became the big two. Everyone else got swept into this one company. So if Amadeus carries the DNA of four airlines, and Sabre of one, Travelport carries the DNA of more than a dozen.

Eric Marketts

So it is the most airline-descended of the three, and also the most tangled.

Steph Nell

Both. And hold that, because the tangle is the whole reason the rebuild we are about to describe was the hardest of the three to pull off, and maybe the most necessary. When your company is more than a dozen old airline systems fused into one, you either keep paying to maintain that mess forever, or you gut it and start clean. Travelport is the one that chose to gut it.

Eric Marketts

So here is the frame for Part 4. Amadeus is betting on breadth. Sabre is betting on endurance. Travelport is making the most all-or-nothing bet of the three. It bet the company on the idea that if you build a genuinely cleaner pipe, the cleanest pipe in the industry, you can win from third place, even against two rivals with more money and more airlines.

Steph Nell

And the whole season comes down to whether that bet, and the two next to it, can all pay off at once. Or whether we are watching three companies fight over two chairs. So let's give Travelport a fair hearing, and then let's do the math on the industry.

Eric Marketts

Now the position, because that lineage left a mark, and it shows up in everything. What kind of company did all that fusing actually produce?

Steph Nell

For years, a struggling one. All that stitched-together legacy made Travelport expensive to run and slow to move. It carried more old plumbing than either of the other two, and it paid for that every year. But the thing that nearly killed Travelport was not the technology. It was the balance sheet.

Eric Marketts

And that financial story is the thing you cannot skip with Travelport, because it shapes every decision the company makes.

Steph Nell

It nearly did not survive the last few years. Travelport went private, owned by a group of investment firms. Siris, Elliott, Davidson Kempner, Canyon Partners. And in late 2023 it had to be rescued from its own balance sheet. The owners put in new equity, around five hundred and seventy million dollars, and did a big debt-for-equity swap, roughly two and a half billion in debt converted, to keep the company alive and push the maturities out. Then this year, in March 2026, the shareholders put in another fifty million dollars specifically to fund the AI and the content and the new platform.

Eric Marketts

So let me be blunt about what that means. The owners have had to keep feeding the company money.

Steph Nell

They have. And you can read that two ways, and we should read it both ways honestly. The bear read is that this is a company that could not fund its own turnaround, that needed repeated injections just to stay upright. The bull read is that sophisticated investors keep choosing to put more money in, which is not what you do with a corpse. You do not write a fifty-million-dollar check in 2026 to fund AI at a company you think is dead. So the balance sheet is fragile, and the people who know it best are still betting on it.

Eric Marketts

And there are some real numbers underneath the bet. It is not all struggle.

Steph Nell

No. The operating story has genuinely improved. Company-disclosed earnings, their EBITDA, the core profit measure, grew about twelve percent last year, from a bit over three hundred million to about three hundred and forty million dollars. That is real improvement off a low base. But hold the number that hangs over all of it, the same way the debt hung over Sabre in Part 3. After that swap, Travelport still carries around two billion dollars in debt, and it comes due in 2028. And be precise here, because this is the exact thing people get wrong. In raw dollars, two billion is less than Sabre's four billion. But Travelport is a much smaller company, so per dollar of earnings the weight is just as heavy, and the deadline is sooner. Sabre pushed most of its wall out to 2029 and beyond. Travelport's is 2028. So the clock in this story is 2028, and everything Travelport is building, it is building against that date.

Eric Marketts

And on size, where does it sit against the other two?

Steph Nell

Third. And not a close third. On the widely reported estimates, Amadeus is the biggest, Sabre is second, and Travelport is roughly a fifth of the global GDS market. So put the whole position together. Smallest share. Around two billion in debt against the nearest maturity wall of the three, in 2028. The least room for error. And a new CEO, John Mangelaars, who took over in April, whose pitch for the company is essentially, we are big enough to matter and small enough to move fast. Which is the only honest pitch a distant third can make.

Eric Marketts

Now the bet itself, because this is what makes Travelport genuinely interesting, and it is the exact opposite of what its size would predict. Walk through what they actually did.

Steph Nell

They rebuilt the pipe. Go back to that cryptic green screen from the top of the episode. The old GDS was built on decades-old architecture. Mainframe-era, cryptic-command, batch-driven plumbing that everything else got bolted onto over the years. Amadeus, in Part 2, is building the future on top of that installed base, layering the stack. Sabre, in Part 3, is modernizing it module by module, componentizing. Travelport did the thing the other two would not risk. It went underneath and rebuilt the core, cloud-native, from a clean sheet. That platform is called TripServices, and they formally launched it in June of 2026.

Eric Marketts

Make TripServices concrete, because the name means nothing to a normal person. What is it, actually?

Steph Nell

It is one modern connection to a lot of travel. Think about the old way. If you wanted flights and hotels and ancillaries out of the GDS, you were dealing with old formats, cryptic commands, separate systems, and a lot of translation glue. TripServices replaces that with a single, clean, modern API. One connection, structured data, machine-readable, with the content ranked by machine learning. Flights, hotels, add-ons, coming through one modern doorway instead of a museum of old ones. They put something like eighteen months of engineering into rebuilding the servicing layer alone.

Eric Marketts

And I want to connect that straight back to Part 1, because it is the same idea coming around again.

Steph Nell

It is. In Part 1 we said the GDS gave the industry a shared language, EDIFACT, the clunky old messaging standard that let every airline and agency talk to each other. That language is what made global distribution work, and it is also what makes it feel like 1985. What Travelport is doing with TripServices is trying to replace the old plumbing with a modern one, the same fight NDC is fighting, but at the infrastructure layer. Not a new screen on the old machine. A new machine.

Eric Marketts

So why does rebuilding from scratch matter so much right now, in 2026 specifically? The old pipe worked. Cryptic and ugly, but it worked.

Steph Nell

Because of who the new customer is. And this is the whole thesis. The next big buyer of travel is not a human agent typing commands. It is an AI agent. Software doing the shopping. And software does not want a cryptic green screen or a forty-year-old message format. It wants clean, structured, modern APIs it can plug into instantly. So Travelport's bet is that in the agentic era, the messy old pipe becomes a liability, and the clean new pipe becomes the advantage. They are betting that the thing that made them look small and behind, having less legacy revenue to protect, actually freed them to do the rebuild the bigger players cannot stomach.

Eric Marketts

That is a genuinely elegant argument. The underdog reframes its weakness as the reason it can move.

Steph Nell

It is elegant. Now let's find out if it is true.

Eric Marketts

Here is where we apply the exact same discipline we used on Amadeus and Sabre, because it is the one rule that has held all season. We separate what is lyve from what is announced.

Steph Nell

And it matters more here than anywhere, because Travelport is telling the best story of the three and has the least deployed proof to show for it. So be precise. The platform is real. TripServices is built and launched, that is not vaporware. But when you ask who is actually running on it, the answer as of now is thin. There are no marquee named customers live on TripServices yet. Travelport talks about advanced discussions with major travel management companies and promises more announcements later in the year. That is a pipeline, not a customer list.

Eric Marketts

So how do we know the rebuild is real and not just a press release? Amadeus had Finnair. Sabre had a Q1 earnings beat. What is Travelport's hard proof?

Steph Nell

The one number that is actually measurable, and it is a good one. By Travelport's own reporting, the share of its transactions running through modern APIs rather than the old cryptic pipe went from about forty-three percent in 2022 to about sixty-three percent last year. Now, that is company-defined and self-reported, so hold it at arm's length. But it is the honest signal. It says the plumbing is genuinely being used, not just built. The traffic is actually moving to the new pipe. That is more than a slide. It is the strongest evidence any of the three has that a modern rebuild is being adopted in production, and not just demoed.

Eric Marketts

But that number cuts both ways, and I want to be honest about it. If Travelport threw out the old machine and rebuilt from scratch, why is a third of its traffic, that other thirty-seven percent, still running through the old cryptic pipe? That sounds like the old machine is very much still on.

Steph Nell

Because building a new pipe and moving every booking onto it are two different jobs, and this is the single most important thing to understand about the whole rebuild. Travelport built the new platform from scratch. That part is real and it is done. But you do not flip a global booking system that thousands of agencies run their business on from old to new overnight. You migrate it, connection by connection, over years. So the old cryptic pipe keeps running in parallel, carrying whatever has not moved yet, until the last customer is across. Sixty-three percent is not a finish line. It is the read on how far the migration has come and which way it is going.

Eric Marketts

So rebuilt from scratch does not mean the old one is already gone.

Steph Nell

Right. And it is the same truth we hit with Amadeus in Part 2. Even there, the old world and the new world run side by side, a coexistence measured in years, not a switchover quarter. Travelport built the cleanest new pipe of the three. It is still draining the old one into it. Both things are true at once. So the honest line is not the old machine is gone. It is the new machine is built, it is real, and the traffic is moving onto it faster than anyone else in this category.

Eric Marketts

So the fair scorecard is what?

Steph Nell

The fair scorecard is that Travelport did the boldest engineering and has the thinnest roster of brand-name go-lives to prove the commercial payoff. The pipe is genuinely modern and genuinely in use by its own agency base. The big, industry-defining customer wins that turn a rebuild into a franchise are still ahead. Built, yes. Adopted at scale, not yet. Anyone who tells you Travelport has already won the agentic pipe is ahead of the evidence. Anyone who says the rebuild is just marketing has not looked at that API number.

Eric Marketts

Travelport has put down two big chips that it hopes turn the rebuild into a winning hand. The first is an airline. The second is an AI company. Start with the airline.

Steph Nell

In December 2025, Travelport announced a long-term strategic relationship with United Airlines. And this one matters more than a normal distribution deal, because of what it is. United is giving Travelport early access to its modern retailing and NDC roadmap, and the two are co-developing merchandising and servicing capabilities together. In plain terms, a major U.S. airline looked at the rebuilt Travelport pipe and said, we will build the future of our retailing with you, not just sell through you.

Eric Marketts

Now hold on. Co-development is exactly the kind of word that shows up in a press release and means nothing. Every vendor claims it partners with its big customers. Why is this any different from a normal, large distribution deal Travelport could dress up the same way?

Steph Nell

Fair, and most of the time you would be right to roll your eyes at it. The difference is what United is actually handing over. Early access to its retailing and NDC roadmap means Travelport gets to see and shape where a major carrier's future offer is going before it ships, and the two are building the merchandising and servicing pieces together. That is not a reseller agreement. That is a carrier choosing to co-own a piece of its retailing future with a vendor. And here is the tell. You do not co-build your roadmap with a company you think is going away. For a distant third that needed exactly one marquee airline to validate the rebuild, United is the name, and it had every other option on the table.

Eric Marketts

Okay, but now give me the other side, because you are being generous.

Steph Nell

I am, so let me take it back a notch. It is early, and it is a partnership, not a live system moving millions of bookings. It is phased across 2026. No financial terms were disclosed. And a roadmap collaboration is a promise about the future, which is exactly the kind of thing this series has taught us to discount until it is in production. So it is the best possible starting signal. It is still a starting signal.

Eric Marketts

Now the second bet, the AI one, and this is the one that got the most attention. Set it up.

Steph Nell

In May of 2026, Travelport announced a collaboration with Cognizant, the big IT services firm, and Anthropic, the AI company behind the Claude models. And it has two layers. The first is inward. They are using the AI to help rewrite and modernize Travelport's own code faster, which for a company doing a giant rebuild against a 2028 clock is not a gimmick, it is speed. The second layer is the interesting one. They are building an agentic layer on top of TripServices using MCP, the model context protocol, the same standard Amadeus was weighing in Part 2 and Sabre built its translator on in Part 3. The idea is that an AI agent can express a travel intent in plain language, and that layer turns it into a real, confirmed booking through Travelport's pipe.

Eric Marketts

And Travelport was pointed about one thing. They said this is not a limited pilot.

Steph Nell

They were, and that phrasing is deliberate, because everyone in this industry is drowning in pilots. Travelport is trying to plant a flag as the clean, structured pipe for the agentic era. When the AI agents come shopping, plug into us, we speak your language natively. First customer-facing capabilities are expected within 2026.

Eric Marketts

But here is where I have to connect it to Part 3, because we have heard this exact song before, three weeks ago.

Steph Nell

We have, and you should call it out. In Part 3, Sabre made the identical claim. Be the clean pipe for the agentic era. Sabre launched its own agentic APIs and its own MCP translator and claimed first-mover status. So clean pipe for AI is not a Travelport moat. It is a crowded street. Amadeus is investing in it from the stack side, Sabre is claiming first-mover, and Travelport is betting the whole rebuild on it. Everybody is racing to be the doorway that AI agents walk through. What Travelport can argue, fairly, is that it is the only one whose entire core was rebuilt clean for exactly this, rather than retrofitted. That is the honest version of its edge. Not that it is alone. That it is cleanest.

Eric Marketts

And there is a deeper vulnerability sitting under all of this, the one that ties the whole season together. The fee.

Steph Nell

The fee. This is the thread from Part 2 and Part 3, and Travelport sits on the wrong side of it, right next to Sabre. Remember the mechanism. In the Offer and Order world, the distribution fee does not disappear. It gets rebuilt into the offer and paid to whoever assembles and prices that offer, whoever runs the retailing engine. For Amadeus the fee migrates inward, because Amadeus owns the airline's operating system and its retailing engine. It built the desk the money moves to. Travelport did not. Like Sabre, it sold or never built that deep airline IT layer. It does not run the airline's core. So when the fee migrates, it can migrate away from Travelport too.

Eric Marketts

So what is Travelport's answer to that? Because that is an existential problem.

Steph Nell

Its answer is the whole bet in one sentence. Travelport is wagering that it does not need to own the airline's stack to catch the fee. It needs to own the connection. If it becomes the cleanest, most neutral pipe that every AI agent plugs into to actually book travel, then it recaptures value at the doorway instead of inside the airline. Amadeus recaptures the fee by owning the engine. Travelport is trying to recapture it by owning the interface to the agents. That is a real strategy. It is also a narrower ledge to stand on, because owning the doorway only pays if the agents actually choose your doorway, and two bigger rivals are building doors right next to yours.

Eric Marketts

So let's land Travelport honestly, bull and bear, before we close the season. The bull case first.

Steph Nell

The bull case is that Travelport made the most intellectually honest bet of the three. It did not renovate a legacy it was afraid to touch. It rebuilt. It has the cleanest, most modern architecture in the category at exactly the moment the industry's new customer, the AI agent, rewards clean and modern. It has real adoption of that pipe by its own base, the sixty-three percent number. It landed United as a co-development partner and a genuinely differentiated AI collaboration. Its earnings are growing. And its owners keep funding it, which tells you the people with the most information have not given up.

Eric Marketts

And the bear case.

Steph Nell

The bear case is time and size. Travelport is the smallest player, carrying around two billion in debt against the nearest maturity wall of the three, in 2028, and the marquee customer proof for the rebuild is still mostly ahead of it, not behind it. Its signature clean-pipe claim is contested by both larger rivals. It sits on the wrong side of the fee migration, with a narrower recapture strategy than Amadeus. And it is owned by investment firms who, by definition, eventually need an exit. So the rebuild is real, and the runway is short, and those two facts are in a race.

Eric Marketts

So the critical take.

Steph Nell

The critical take is that Travelport did the right thing and may have the least time to be rewarded for it. Rebuilding the pipe from scratch was probably the correct architectural call. Ripping out the legacy instead of layering on it is what you would do if you were designing for the next twenty years. But strategy and survival are different clocks. Travelport has to prove that content and customers and agentic volume arrive on the new pipe before the debt maturity forces the question. It built the best machine for the future and it has the least cushion to wait for the future to show up. That is the whole story. Boldest bet, least margin for error.

Eric Marketts

And that sets up the real reason we made this a four-part series instead of one episode. Because you cannot judge any one of these bets in isolation. You have to put all three on the table at once.

Steph Nell

So let's do exactly that. Let's close the season.

Eric Marketts

Pull all the way back to Part 1. For fifty years, all three of these companies sold the same product. Aggregated airline inventory, delivered through a terminal, for a fee. That was the GDS. Same business, three logos. The single most important thing that has happened in this industry is that the shared product finally broke apart, and each company made a different bet on what replaces it.

Steph Nell

And say the three bets plainly, because this is the spine of the whole season. Amadeus is betting on the stack. Own the airline's operation, its distribution, and its retailing engine, all at once, and rotate the fee inward as the old pipe fades. Sabre is betting on focus and endurance. Get smaller, get cleaner, protect the North American corporate fortress, and modernize before the debt catches up. Travelport is betting on the rebuild. Throw out the legacy, build the cleanest pipe, and own the connection the AI agents plug into. Three answers to one question. Does the function survive, and in whose hands.

Eric Marketts

And here is the thing people keep getting wrong. They keep asking whether the GDS is dying. That is the wrong question.

Steph Nell

It is the wrong question. The GDS is not dying. It is differentiating. The jobs the GDS did are permanent. Somebody still has to aggregate competing airlines' content, rank it fairly, settle the money, and give the buyer an honest view of the market. Those jobs do not go away in the Offer and Order world or the agentic world. They just get done by a different kind of company, in a different shape. So the real question is not survival of the function. It is who owns the function, and how many of them survive doing it.

Eric Marketts

And that is the question this series has been building to. Can this industry support three of these companies?

Steph Nell

And here I want to be careful and label this clearly, because it matters. What I am about to say is our analysis, our opinion, not a reported fact. No one has announced a merger. Travelport's own messaging is organic growth, next phase, standalone. So this is us reading the board, not reporting a deal.

Eric Marketts

Flag noted. Now read the board.

Steph Nell

The economics are pointing toward consolidation pressure, and it falls hardest on exactly the two players we would expect. Go back to the fee migration. Amadeus built the desk the money moves to. It owns the airline's operating system, so as the fee migrates inward, Amadeus catches it. Sabre and Travelport do not own that deep airline IT layer. When the fee migrates, it can migrate away from both of them. So the two companies most exposed to the transition are the second and third players, the ones without the stack to recapture the fee. And both are carrying debt. That is not a coincidence. That is the structure of the industry telling you where the strain lands.

Eric Marketts

But let me push on that, because it sounds darker than the facts support. The owners keep writing checks. TripServices is lyve. EBITDA is growing. United signed on. That is a company moving forward, not one circling a drain. Are you not talking yourself into a merger story the numbers do not actually show?

Steph Nell

It is a fair challenge, and I want to be precise about what I am, and am not saying. I am not saying Travelport is failing. Everything you just listed is real, and it is exactly why I keep saying the rebuild was the right call. What I am saying sits underneath the quarterly numbers. Growing earnings do not change the fact that the fee is migrating toward whoever owns the airline's engine, and neither Sabre nor Travelport owns that layer. A company can execute well and still be on the wrong side of where the money is moving. That is not pessimism about management. It is arithmetic about position. And when two capable companies are squeezed by the same structural force, both carrying debt, the market's usual answer is that they combine, to take out cost and buy scale. I am not predicting a deal. I am telling you the pressure is real, and it points one direction.

Eric Marketts

And say the quiet part, because we literally titled Part 3 The Survivor. Is Sabre actually the only survivor in this story?

Steph Nell

No. Both of them survived a near-death moment. They just did it differently. Sabre survived by selling a healthy business, its hotel unit, to pay down debt. Travelport survived by handing its lenders about two and a half billion dollars in equity instead of paying them back in cash. Two different ways to cheat the same fate. The word survivor fit Sabre's move, selling a limb to stay whole. Travelport's move was to restructure the balance sheet and then rebuild the product. But make no mistake, Amadeus is the only one of the three that never had to be rescued. The other two both looked over the edge.

Eric Marketts

So put a sharp point on it. What does three-becomes-two actually threaten?

Steph Nell

It threatens the exact thing Part 1 said the GDS was forced to provide in the first place. Go back to the original sin. In 1976 the industry had a chance at one neutral, shared system and the biggest players torched it, because a neutral pipe is worthless to you if your whole advantage is owning the pipe. It took a decade of antitrust to force neutrality onto the GDS. Price transparency, a level screen, a fair view of the market. Those were not gifts. They were imposed. Now play consolidation forward. If three aggregators become two, or two become one, you are re-concentrating control of the neutral marketplace into fewer hands. And the last time control concentrated like that, the screen got rigged.

Eric Marketts

And this time there is no obvious referee. In the eighties the systems were visible, regulated, and everyone knew who owned the screen. Now the control is moving into stacks and pipes and AI layers that most travelers will never see.

Steph Nell

That is the uncomfortable close. The neutrality that took a decade to force onto this industry was never actually paid off. Part 1 said it. It was regulated into a truce. And truces hold only as long as the balance of power that made them necessary. We now have one player trying to own the whole stack, one fighting to survive, and one betting the company on a rebuild, all at the same time that the buyer is becoming a machine and the fee is quietly migrating to whoever owns the engine. Fewer hands, less visibility, and the same incentive that started the whole story in 1976. Own the pipe, tilt the screen.

Eric Marketts

So the series verdict, in one honest breath.

Steph Nell

The GDS did not die. It split into three bets, and the industry probably cannot reward all three equally. Amadeus has the best position and the hardest math. Sabre has bought itself runway and still has to prove the business it saved was worth saving. Travelport built the best machine for the future and has the least time to wait for the future to arrive. Watch production, not press releases. Watch which fee migrates where. And watch whether anyone remembers, this time, that neutrality in this industry has never once happened on its own. It has only ever been forced.

Eric Marketts

That is the series. Four parts, one machine, three bets, and a fifty-year fight that is very much not over. If you came into this thinking the GDS was a boring back-office relic, I hope you leave seeing what it actually is. The contested center of a trillion-dollar industry, being rebuilt in real time by three companies who do not agree on what it should become.

Steph Nell

And if you are new to this industry and wondering why any of this matters, here is the one thing to carry out. Every fight you are going to read about this year, NDC, Offer and Order, AI agents, airline direct, all of it, is really a fight about who controls the space between a traveler and a seat, and who collects a fee for standing there. That is the whole game. Now you can see it. And once you can see it, the headlines stop being noise and start being a map.

Eric Marketts

So here is the ask. If this series sharpened how you see the industry, follow The V1 Airline Retailing Report so you do not miss what comes next, and send this season to one person who needs the history to make sense of the headlines. We built these four parts to be shared, and to still make sense a year from now. We are on Apple Podcasts, YouTube, and Spotify. Rate it, and it reaches more people building the future of this industry.

Steph Nell

And tell us where we got it right and where you think we got it wrong. If you are inside one of these companies, or betting against one of them, we want to hear it. This is the conversation, and you are part of it now.

Eric Marketts

I'm Eric Marketts.

Steph Nell

I'm Steph Nell. Thanks for spending the season with us.

Eric Marketts

Same machine. Three bets. Stay sharp out there.