Airline retailing is finally becoming real — and the GDSs are not invited

For most of the last decade, "airline retailing" has been a phrase that gets a polite nod at industry conferences and very little actual investment. NDC was the asterisk on every distribution roadmap — the thing that would happen eventually, after the next fare filing project, after the next core upgrade.

That asterisk is finally being cashed.

In the last twelve months, three of the largest network carriers — Air France-KLM, United, and Lufthansa Group — have moved retailing from a side project to a board-reported program. American Airlines, Delta, and British Airways are not far behind. The work is no longer about checking the NDC compliance box for IATA. It is about owning the offer construction layer end-to-end — fare, ancillary, loyalty redemption, dynamic price, channel-specific presentation — and pushing GDSs into a narrower role as a connectivity utility.

This is the shift the industry has been talking about since 2014. It is now actually happening, and the strategic implications are larger than what's currently in sell-side models.

What changed

Three things changed at roughly the same time.

First, cloud-native offer and order management systems are finally production-grade. Amadeus Nevio, Sabre SabreMosaic, and the in-house platforms built by Lufthansa and Singapore Airlines have all reached scale. Carriers can now build or buy a retailing platform that genuinely sits outside the PSS without rewriting it. That removes the single biggest excuse for delay.

Second, the largest OTAs accepted NDC as a fact of life. The grinding distribution war that dominated 2017–2021 ended quietly: Booking and Expedia signed connection deals with American, Delta, and United through 2024 and 2025, agencies adapted commission models, and the channel-mix risk that airline CFOs feared turned out to be manageable. PhocusWire reported in late 2025 that more than 80% of Booking.com's flight inventory now flows over NDC pipes for participating carriers.

Third, AI changed the math on the cost side. Personalized retailing used to require a six-figure team to operate per region. Carriers are now building offer optimization with materially smaller teams because the data-science work that took quarters takes weeks.

The combined effect is that the marginal cost of being a real retailer fell, and the marginal cost of not being one — in lost ancillary capture and direct-channel margin — rose.

Where margin moves next

If you believe airlines own the offer, you have to believe a meaningful pool of distribution economics moves with it. We think three pools are in play.

• Ancillary capture in indirect channels. Today, a material share of agency and OTA bookings sell only the base fare. Genuine NDC offer construction lets carriers present bundles, seats, bags, and lounge access through every channel. Even modest attach rate gains compound quickly across hundreds of millions of segments.
• Loyalty redemption as a margin lever. When the offer engine controls cash-plus-points pricing in real time, mileage liabilities become more actively managed and breakage assumptions get sharper. This is a quiet but real P&L story for the carriers with the largest loyalty programs — Delta SkyMiles, United MileagePlus, and the AmEx-linked Krisflyer.
• GDS fees per segment. This is the most visible line. Lufthansa's €18 distribution surcharge, introduced in 2015, has become the industry template — IAG, Air France-KLM, and most Asian network carriers now run a version of it. As more bookings shift to direct-connect NDC, the per-segment economics of legacy GDS distribution compress. The Financial Times reported in late 2024 that Sabre and Amadeus are now running their own retailing platforms as a defensive response.

None of these are revolutionary alone. Taken together over a three-year horizon, they materially change the unit economics of distribution at the network carrier level.

The read-through

For airline strategy and IR teams: the carriers furthest along on retailing have a story to tell that the market has not yet priced. Make sure your investor materials are explicit about the margin pools you are unlocking and on what timeline. This is one of the few credible levers for non-volume EBITDA growth this cycle.

For GDS investors: the bear case here is no longer about NDC adoption — that fight is over. The bear case is about per-segment fee compression as direct-connect share grows. The bull case rests on Sabre and Amadeus successfully repositioning as offer-and-order infrastructure, which is plausible but execution-dependent. Sabre's announced spin-out of its retailing unit is the clearest signal that incumbents see the same picture we do.

For OTAs and metasearch: the medium-term risk is not losing inventory — carriers want to be on every channel that converts. The risk is becoming a thinner layer in the stack as more of the offer logic lives upstream at the airline. Booking's continued investment in flight infrastructure is the right defensive posture.

For PE and growth investors in airline retailing tech: the buyer universe just got real. We expect 2026–2027 to be an active year for both strategic and sponsor M&A in offer management, order management, and the surrounding data infrastructure. Watch for moves on PROS, Travelfusion, and the smaller European order-management vendors.

The asterisk is being cashed. The carriers who treat this as a board-level program will compound on it; the ones still treating it as an IT project will not.