State of the GDS, Part 2: Amadeus and the Stack Play

Eric Marketts

In May of 2025, a passenger booked a Finnair ticket, and the booking did not create a PNR. No passenger name record. For sixty years, every airline booking on Earth created one of those. It is the little record locator, the six characters, the spine of how the industry has tracked you since the SABRE days. This booking did not make one. It created something else. An Order. The thing the industry had been talking about for a decade, and mostly not doing, quietly happened on a real flight, with a real passenger .

Steph Nell

And the rails that booking ran on were built by the biggest GDS in the world. Which is the strange part, and it is what this whole episode is about.

Eric Marketts

Welcome back to The V1 Airline Retailing Report. This is State of the GDS, Part 2. I'm Eric Marketts.

Steph Nell

And I'm Steph Nell. Quick orientation if you are just joining. This is a four-part series taking the global distribution system apart, one company at a time. Part 1 was the foundation, how a competitive weapon got forced into neutral infrastructure, and we closed on a warning. Control is concentrating again, and the last time one player had this much of it, they rigged the screen .

Eric Marketts

So Part 2 starts with the player doing the most to re-concentrate that control. Amadeus. The largest GDS in the world. And the strangest strategic story in this industry, because Amadeus is not defending the GDS. Amadeus is quietly building the thing that kills it. On purpose. From the inside .

Steph Nell

The market leader, dismantling its own most profitable product before anyone else gets the chance. Let's understand why a company would do that, and whether the timing works.

Eric Marketts

Before the scale numbers, one thing people forget about Amadeus, because it matters for everything that follows. Where it came from.

Steph Nell

Nineteen eighty-seven. Four European airlines, Air France, Iberia, Lufthansa, and SAS, put Amadeus together on purpose. Part 1 gave you SABRE, born inside one American airline in the nineteen fifties. Amadeus was the mirror image. A consortium of carriers building their own distribution rails so Europe would not have to run its business on American systems.

Eric Marketts

So hold that, because it reframes the whole episode. Amadeus was created by airlines, to give airlines control of their own distribution. Everything we are about to describe, the stack, Nevio, the Order, is that same company, forty years later, doing the same thing in a new shape. This is not a GDS pivoting. It is a company doing what it was founded to do, sit at the center of airline distribution. The difference is who holds the center. In 1987 that was four airlines. Today it is Amadeus, and only Amadeus.

Steph Nell

And it got big. So start with scale, because you cannot understand the bet without it.

Eric Marketts

Start with scale. How big is Amadeus, actually?

Steph Nell

Big. For the full year 2025, revenue was about six and a half billion euros, growing about eight and a half percent at constant currency. Strong margins EBITDA close to thirty-eight percent. This is not a company in trouble looking for a Hail Mary. This is the leader, growing, throwing off cash.

Eric Marketts

And here is the number that reframes everything people assume about Amadeus. Most of the industry still calls it a GDS. A booking pipe. But the GDS is no longer the majority of the company.

Steph Nell

Right. The distribution business, the GDS, is still the single largest line, about forty-eight percent of revenue. But that is under half. Add up everything that is not the GDS, the airline IT business and hospitality, and it is the bigger share now, just over half. So the pure GDS, the thing everyone still labels Amadeus, is no longer the majority of Amadeus. The growth and the strategy liv on the technology side.

Eric Marketts

Unpack the stack, because this is what makes Amadeus different from Sabre and Travelport. It does not own one layer. It owns three.

Steph Nell

Three layers. Layer one, the GDS, the distribution business, the pipe that sells airline content to agencies. That is the classic Amadeus. Layer two, Altéa. That is the Passenger Service System. It runs the airline's actual operation. Reservations, inventory, departure control, check-in. Altéa has served more than two hundred airlines and well over a billion passengers a year. When we said in Part 1 that the PSS is the operational nervous system of an airline, Altéa is the biggest one on the planet .

Eric Marketts

And layer three is the new one. The one that changes the game.

Steph Nell

Offer and Order. The retailing engine. Amadeus calls it Nevio. And the point of owning all three is that they reinforce each other. If you already run your operation on Altéa, Amadeus is your operating system. If you distribute through the Amadeus GDS, it is your sales channel. And now Nevio offers to be your retailing brain on top of both. Three layers, one vendor, each one making the others stickier.

Eric Marketts

That is the stack play. And it is why this is not really a GDS story anymore. It is a story about a company using its GDS as a foothold to own everything around the booking.

Eric Marketts

And the reach is bigger than Altéa. This is the part almost everyone misses.

Steph Nell

Amadeus also owns Navitaire, the platform that runs the low-cost carriers, the airlines that built their whole business to stay out of the GDS. Sell direct, unbundle everything, avoid the old pipe. Amadeus owns their operating system too. So its reach covers both ends of the market, full-service on Altéa, low-cost on Navitaire. The airlines that rejected the GDS still run on Amadeus, and still pay Amadeus. And there are now two retailing engines, not one. Nevio for the full-service carriers, Navitaire Stratos for the low-cost and hybrid ones, both built to the Offer and Order standard. Stratos already has launch customers, TUI and Volotea. The irony is the low-cost carriers invented modern retailing years before the term existed. Now Amadeus owns their platform and the one everyone else is migrating to.

Eric Marketts

Let's make Offer and Order concrete, because the industry drowns this idea in jargon, and underneath it is simple.

Steph Nell

It is simple. Today, when you book a flight, the airline creates a PNR, then issues a ticket, then handles your bag and your seat and your meal in a stack of separate systems that were bolted together over forty years. The fare is in one place, the ticket in another, the ancillaries in another. It is held together with tape. The Order model throws that out. One record. The Order. It holds everything you bought, the seat, the bag, the flight, priced and managed as one thing the airline controls directly.

Eric Marketts

And that is what happened on the Finnair booking we opened with. No PNR. A native Order. Why does that one technical detail matter so much?

Steph Nell

Because the PNR is a GDS-era artifact. It is the format the old pipe speaks. The minute an airline can create and manage an Order natively, in its own systems, it no longer needs the GDS to be the keeper of the record. Finnair created the world's first airline-native Order in 2025. That is the real production reference for this entire transition. Not a slide. A lye'v airline, issuing Orders.

Eric Marketts

Now here is where I want to be disciplined, because this industry constantly confuses a signed deal with a live system. Where is Nevio actually in production, and where is it just announced?

Steph Nell

Good, because the gap is large. Truly lye'v, in production. Finnair, the reference case. And Saudia, which is doing something clever called smart bridging, converting its legacy Altéa records into Orders so it can operate the new model without ripping everything out at once. Those two are real.

Eric Marketts

And the marquee names?

Steph Nell

British Airways is the clearest example of a major carrier retiring proprietary distribution infrastructure. They ran their own NDC platform for over a decade, built on IATA schema 17.2. In June 2026, they announced they were retiring it and moving to Amadeus Altéa NDC on version 21.3. Revenue management moved to Amadeus, selected in 2024 and cut over in 2025. Nevio is being implemented now for Offer and Order. BA is consolidating onto a complete Amadeus stack, PSS, revenue management, NDC, and Offer and Order, all from one vendor. The last piece, Nevio for Offer and Order, is still being implemented. That's not a distribution deal. That's a platform commitment. And it signals something important: the window to justify proprietary distribution infrastructure at a major carrier closed in 2023.

Eric Marketts

Sit on BA for a second, because it is not just another logo. Was that a smart commercial move, or a surrender?

Steph Nell

Both readings hold, and that is what makes it interesting. The smart-move read: BA was an NDC pioneer. For a decade they proved the thesis, built the direct relationships, took the economics, and now they hand the plumbing to a vendor who can maintain the schema versions and ship Offer and Order faster than any internal team. That is a mature build-versus-buy call. Stop paying to run middleware that is not your core business.

Eric Marketts

And the surrender reed?

Steph Nell

Harder to shake. The whole promise of NDC, the reason BA fought the GDS over it for years, was airline control of distribution. Own the offer, depend on no one. And the loudest evangelist for that promise just decided it could not, or would not, sustain it alone. BA did not win independence. It swapped GDS dependence for Amadeus dependence, and a deeper one, because now the offer engine sits in the same stack as the operation.

Eric Marketts

Which raises the uncomfortable question for the whole industry. If the airline that wanted distribution independence most cannot run it in-house, then maybe the NDC endgame was never airline independence at all. Maybe it was just airlines choosing which vendor to depend on. That is a very different future than the one NDC was sold as.

Steph Nell

Hold that, because it is the series thesis in miniature. And BA is not alone. Lufthansa Group committed to Nevio in January 2026. Nine carriers across the group. Not a single headline announcement. Amadeus is rolling it out carrier by carrier, Lufthansa, SWISS, Austrian, and Brussels Airlines among the named carriers. The breadth is the signal. Nine carriers from one group moving to the same retailing stack is a platform decision, not a pilot.

Steph Nell

Air France-KLM put over a hundred million euros behind it. Multi-year, multi-product. Revenue management, NDC, and Offer and Order all in scope under the MOON program. Three of the largest carrier groups have made the same call. BA retired proprietary NDC infrastructure. Lufthansa Group is deploying Nevio across nine carriers. Air France-KLM committed nine figures. The pattern is not ambiguous.

Eric Marketts

One precision on Finnair before we move on. Because that native Order is the headline, and the headline carries a qualifier.

Steph Nell

It does. The native Order is the direct channel. Book on Finnair.com or the Finnair app, the booking creates an Order in Nevio, no PNR as the primary record. Book that same flight through a GDS or an agency, and it still creates a PNR. That is not a failure. It is the architecture. The legacy pipe keeps moving legacy-pipe bookings. And this is a coexistence measured in years, not a transition quarter. Expect Altéa and Nevio running side by side for a decade or more.

Eric Marketts

And the underlying settlement?

Steph Nell

Unresolved, in public. We know the booking creates an Order. Whether the settlement, the actual money movement, runs through an Order ID or still runs through the industry's ticket-settlement system, the BSP and the IATA ticket underneath, that has not been publicly confirmed by Finnair or Amadeus. The ticket may not be gone. It may be invisible, operating as a settlement rail in the background while the Order handles the surface. Native Order: real . Direct channel only: confirmed . Settlement via Order ID versus BSP ticket: publicly unresolved .

Eric Marketts

So the slide looks like dominance, and the production reality is two airlines.

Steph Nell

Two airlines, and be precise even about those two. Remember the native Order is the direct channel only, the website and the app. For a legacy carrier like Finnair, that is a minority of total bookings. The rest still flows through agencies and OTAs as a PNR. So even at the reference airline, Orders are a slice of one channel, not the whole airline. Not a hundred percent of its bookings. Not half. That is how early this genuinely is.

Eric Marketts

And yet.

Steph Nell

And yet. Two true production references and a very long, very real pipeline. Both things are true, and you have to hold both. The commercial momentum is genuine, the deployed reality is early. Anyone who tells you Offer and Order is done is selling something. Anyone who tells you it is vaporware is not watching Finnair.

Eric Marketts

Here is the part that I think most people miss about Amadeus, and it connects straight back to Part 1. Everyone assumes the GDS is the moat. It is not. The moat is the integration.

Steph Nell

Say more, because that is the strategic core.

Eric Marketts

Think about what it takes to leave. If you are an airline running your whole operation on Altéa, switching your PSS is not a software upgrade. It is a multi-year, nine-figure program that touches every part of how you run flights. We said that in Part 1 about Altéa specifically. Now layer Nevio on top. The airline that runs Offer and Order on Amadeus is, by design, running even more of its business on Amadeus. Every layer you adopt makes the next one harder to refuse and the whole thing harder to leave.

Steph Nell

So the GDS was never the lock. The lock is depth. The more of the stack you run on one vendor, the more switching costs compound, and the more that vendor can move you to whatever comes next, on its timeline.

Eric Marketts

And that is the genius and the danger of the stack play in one sentence. Amadeus does not have to win the Offer and Order market in the open. It can win it inside its own installed base, airline by airline, because those airlines are already on Altéa and already find it easier to say yes than to shop the market.

Steph Nell

Which raises the obvious question for any airline listening. If the moat is depth, are you choosing Nevio because it is the best retailing engine, or because you are already too deep in the stack to seriously evaluate anyone else? That is not a rhetorical jab. It is the actual decision on the table in a lot of boardrooms right now.

Eric Marketts

And there is no clean answer. Depth is convenient and depth is a trap, and usually it is both at once .

Eric Marketts

Now the hard part. The strategic tension at the center of the whole Amadeus story. And I want to be careful here, because this is our analysis, not an Amadeus admission. Amadeus does not describe it this way.

Steph Nell

Frame it.

Eric Marketts

Amadeus makes money every time a booking flows through its GDS. A segment fee, a charge for every flight leg booked. That is the classic, high-margin, decades-old revenue. Now look at what Nevio and the Order model do. They move bookings out of that GDS pipe and into a direct, Order-based relationship between the airline and Amadeus, where the economics are different. So Amadeus is building the thing that erodes its own most profitable product. That is self-cannibalization. And to be precise, Amadeus does not call it that. Officially the framing is additive, more revenue per passenger inside the IT business. The cannibalization read is ours, and the trade's, not a quote from the company.

Steph Nell

So let's do the bull case and the bear case honestly, because this is the crux of whether the stack play works.

Eric Marketts

The bull case is that Amadeus is doing the only rational thing a market leader can do. If the GDS is going to get disrupted anyway, by NDC, by Offer and Order, by AI agents, then better to be the one disrupting it, on your own timeline, capturing the new revenue as the old revenue declines. And the numbers, so far, support them. Revenue up about eight and a half percent at constant currency. The IT side growing double digits and now driving the growth. GDS share of the business shrinking while total revenue grows. That is exactly what a managed transition is supposed to look like from the outside. They are not bleeding. They are rotating.

Steph Nell

And sit on that word, rotating, because it is the mechanism people miss. When a booking leaves the GDS pipe and becomes an Order, the airline stops paying the old per-segment GDS fee. But the distribution cost does not vanish. It gets rebuilt into the offer the airline constructs, into the price, and it gets paid to whoever runs the retailing engine and assembles that offer. For most of the industry that is a new intermediary. For Amadeus, it is Amadeus. The segment fee leaving the GDS line reappears on the IT line as a per-passenger, per-transaction platform fee. The money does not leave the building. It changes desks.

Eric Marketts

And one thing to say plainly, because it is a common confusion. This is not the cost moving from the airline onto the traveler. The traveler funded distribution the whole time. It was always inside the fare, since the GDS was a per-segment charge in the airline's cost base. What changes is not who pays. It is which desk collects, and who controls how that cost shows up in the price. And that is exactly why the cannibalization is survivable for Amadeus and dangerous for the other two. Sabre and Travelport face the same fee migration, but they sold off or under-built the IT side. When the fee moves, it can move away from them. Amadeus built the desk it is moving the money to.

Steph Nell

And the bear case?

Eric Marketts

The bear case is timing and depth of proof. The GDS revenue is real today, banked, high margin. The Nevio revenue that is supposed to replace it is mostly still a pipeline. Two airlines in true production against a book of multi-year rollouts. If the GDS line compresses faster than Nevio scales, there is an air pocket, a period where the old business is shrinking and the new one has not caught up. And the hardest part of Offer and Order, true dynamic offers generated in real time at scale, is the part the whole industry is still worst at. Amadeus is betting it can run that transition faster than the decline. That is a bet, not a certainty.

Steph Nell

And the agentic layer sits right on top of this, because it changes how fast the old model erodes.

Eric Marketts

It does. Amadeus knows the AI shopping wave is coming, and it knows what we covered in Part 1, that agents shopping endlessly blow up look-to-book economics, the ratio of searches to actual sales. Their CTO, Sylvain Roy, has been blunt that adopting MCP alone is not enough, it is only a first step, because it cannot handle the real retailing workflows, shopping, booking, servicing. So Amadeus is investing in things like precomputed fares to absorb that query volume, and it bought a company called SkyLink to put conversational, AI-first booking inside tools like Slack and Teams for corporate travel. They are trying to own the agentic entry point too, the same way they are trying to own every other layer.

Steph Nell

So the critical take.

Eric Marketts

The critical take is that Amadeus has the best position and the hardest math. Best position, because it is the only one of the three that owns the operation, the distribution, and the retailing engine all at once, and it is funding the transition from genuine strength. The hardest math, because it is the one with the most high-margin GDS revenue to lose, which means it has the most to cannibalize and the most to time correctly. Sabre and Travelport are fighting to survive the transition. Amadeus is trying to own it without tripping over its own cash cow on the way. That is a harder problem than it looks, and the next two or three years of Nevio production, not announcements, production, is how we will know if the timing held.

Eric Marketts

Pull it back to the series. In Part 1 we said the GDS got forced into neutrality, and that control is now re-concentrating. Amadeus is the clearest example of an incumbent trying to pull that control back to itself, from the airline side, by owning the whole stack.

Steph Nell

And the elegant part is that it does not look like a power grab. It looks like helpfulness. We will run your operation, distribute your content, and build your retailing engine, all in one place, at your own pace. But the result is the same shape we saw in 1976. More and more of the industry's plumbing, controlled by fewer hands. Just this time the hand is a technology vendor, not an airline, and the tool is a stack, not a screen.

Eric Marketts

Whether that is good for airlines depends entirely on a question we cannot answer yet. Does owning the whole stack give airlines a faster path to modern retailing, or does it just rebuild the old dependency in a new shape. Amadeus is betting it is the former. History says watch the lock-in.

Steph Nell

Next time, Part 3, the opposite story. Sabre. The survivor. The company that did not get to play offense, that had to sell off part of itself just to stay in the game, and is now betting it can modernize faster than its debt can catch up with it.

Eric Marketts

Amadeus is the incumbent trying to own the future. Sabre is the incumbent trying to lyv long enough to see it. Two very different bets, born from the same machine.

Steph Nell

If this was useful, follow the show so Part 3 lands in your feed, and share it with someone trying to make sense of where airline retailing is actually going. The V1 Airline Retailing Report is on Apple Podcasts, YouTube, and Spotify. I'm Steph Nell.

Eric Marketts

And I'm Eric Marketts. Same machine. Three bets. Stay sharp out there.