The V1 Airline Retailing Report

Episode 024: "Delta's Calendar, Lufthansa's Clock"

Published: Monday, October 5, 2026

Episode Description

Three weeks ago, in Episode 21, we asked whether Delta's servicing-first NDC would work on launch day. Delta has now published its schedule. Eric and Steph read it, then follow one question across three narratives. Who does the work around the sale, and who pays?

Delta answers the servicing question with a calendar. Business Travel News reported Delta's launch list and 2027 roadmap from a digital Delta Business Showcase event. At launch, agencies are due to be able to shop flights, see private fares, settle through ARC and cancel orders. Changes to private fare bookings arrive in the first quarter of 2027, corporate loyalty identifiers in the second, international sales in the third. The bull case is a defined core with dates. The bear case is that the corporate features come after launch. Our read: partly. The corporate half of servicing is phase two. That is our reading of a reported list, not Delta's claim.

Lufthansa puts a clock on the corporate approval. Lufthansa Group has introduced a 24-hour price hold on NDC bookings for agencies and TMCs serving corporate clients, with a valid Corporate Location Identifier required. Its factsheet says September 2026 to January 2027, and trade outlets report October 1 to January 18. The stated reason is time for internal approvals. The bull case is a concrete fix for a real corporate pain. The bear case is a short trial, listed exclusions, and a hold that only applies in Lufthansa's NDC channels. We found no corporate-only volume figure. Put it next to Delta: two airlines, two problems, one message. Both are offers, not proof.

The same AI, two ledgers. Spotnana's Steve Singh says he believes AI can cut servicing costs by 50 percent or more. Skift reports Amadeus's Elena Avila saying the bills for AI are arriving now and revenue will not simply follow. Both agree on the cost. Only Singh claims a saving. The saving is labor at the seller. The cost is search at the shopping layer. Different companies hold each side. That is our inference. And nobody has published a price for an AI agent's search.

The Bottom Line

Delta put a dated list of servicing features in front of agencies, and the corporate features sit at the back of it. Lufthansa put a 24-hour clock on corporate approvals, as a trial. Servicing is the service department, and that is where cost and complaints live. Until agencies show corporate volume moving through these lanes, the new lane stays a second lane.

Stories Referenced

Delta Answers the Servicing Question
- Business Travel News, "Delta Details Initial NDC Capabilities, Enhanced Features for Agencies" (Sept 25, 2026): https://www.businesstravelnews.com/Distribution/Delta-Details-Initial-NDC-Capabilities-Enhanced-Features-for-Agencies

Lufthansa Puts a Clock on the Corporate Approval
- Lufthansa Group NDC Partner Program factsheet, Price Guarantee Time Limit (Sept 2026): https://lhgroupairlines.com/fileadmin/ndc-assets/news/PGTL_Factsheet_NPP_Website.pdf
- Business Travel News Europe, "Lufthansa to pilot 24-hour price 'guarantee' for NDC corporate bookings" (Sept 14, 2026): https://www.businesstravelnewseurope.com/TMC-Distribution/Lufthansa-to-pilot-24-hour-price-guarantee-for-NDC-corporate-bookings
- TRAVEL INSIDE, Lufthansa Group price guarantee for corporate bookings (German): https://abouttravel.ch/reisebranche/laengere-preisgarantie-fuer-firmenbuchungen/
- L'Agenzia di Viaggi Magazine (Sept 17, 2026, Italian): https://www.lagenziadiviaggimag.it/ndc-lufthansa-estende-il-price-guarantee-time-limit/

The Same AI, Two Ledgers
- Skift, "Agentic AI Is Raising Travel's Tech Bill" (Sept 23, 2026): https://skift.com/2026/09/23/agentic-ai-is-raising-travels-tech-bill/
- Skift, Spotnana CEO Global Forum preview (Sept 19, 2026): https://skift.com/2026/09/19/skift-global-forum-preview-spotnana-ceo-after-booking-servicing/

Licensed under a Attribution 4.0 International License.
The V1 Airline Retailing Report publishes every Monday. Subscribe on Apple Podcasts, Spotify, or wherever you listen.
V1 Advisory LLC | v1advisory.co

Episode 023: "Same Toll Booth, New Road"

Published: Monday, September 28, 2026

Episode Description

Last week ended on one line: the number that decides the outcome is the one nobody publishes. This week Sabre published a number, nearly eighty customers on its AI agent layer, and left out the one that matters: what it charges them. Eric and Steph follow one argument across three stories from the same week. The protocol changed. The intermediation economics didn't.

Sabre says MCP is the new NDC. On September 21, Sabre declared the AI connection protocol the successor to the airline content standard. The deployment is real: a year in production, nearly eighty customers, and a Skift award for Virgin Australia. The claim is positioning. MCP is a pipe. NDC is what flows through it. Calling the pipe the standard puts Sabre at the center of the agent channel and treats a decade of airline offer work as secondary. Watch the fee schedule, not the headline.

T2RL Engage opens the same day, and the airline side answers. In London, the industry measured Offer and Order against "The Journey to Value." Accelya brought Lufthansa Group numbers showing half of indirect bookings now on NDC, and a line of its own: NDC is the foundation, not the destination. Infosys put AI traffic costs on stage. The airline-side architecture is structurally better for margin control. It is also years behind the agent infrastructure the GDSs have already shipped.

Kenya Airways picks a foundation. On day two of T2RL, Kenya Airways moved its reservations, inventory, and Offer and Order core to Sabre Mosaic, kept NDC distribution with Amadeus, and put its direct channel on Branchspace. Three vendors, one real risk: if Sabre's MCP becomes the default path agents use, Kenya pays to modernize and still pays a toll at the agent layer.

The Bottom Line

Everyone in this week's news put themselves in the middle of the transaction and called it enablement. The GDSs took in roughly seven billion dollars in distribution revenue last year. The agent channel is where that toll gets reset or renewed. The number that will tell you which is the per-booking fee for MCP-originated bookings, and nobody has published it yet.

Stories Referenced

Sabre Says MCP Is the New NDC
- Sabre, "MCP Is the New NDC: Is This Travel's Shift-to-Streaming Moment?" (Sept 21, 2026): https://www.prnewswire.com/news-releases/mcp-is-the-new-ndc--is-this-travels-shift-to-streaming-moment-302884118.html
- Hospitality.today, "Airlines spent a decade building NDC. Sabre says AI makes it yesterday's standard": https://www.hospitality.today/article/airlines-spent-a-decade-building-ndc-sabre-says-ai-makes-it-yesterdays-standard

T2RL Engage: The Airline Counter
- T2RL, "The Journey to Value | T2RL Engage 2026": https://www.t2rl.com/events/journey-to-value/
- Accelya, Engage-week NDC data (Sept 22, 2026): https://w3.accelya.com/resources/press-releases/modern-airline-retailing-gap-widens/
- Infosys at T2RL Engage 2026: https://www.infosys.com/newsroom/events/2026/travel-technology-research.html

Kenya Airways Picks a Foundation
- Sabre, "Kenya Airways Partners with Sabre to Enhance Customer Experience through Upgrade to Modern Retailing Platform" (Sept 22, 2026): https://investors.sabre.com/news-releases/news-release-details/kenya-airways-partners-sabre-enhance-customer-experience-through
- Branchspace, "Kenya Airways Takes Off as a True Retailer, with Branchspace": https://www.pr.com/press-release/979772
- The Star (Kenya), "Kenya Airways ditches 20-year-old booking system in tech overhaul" (Sept 24, 2026): https://www.the-star.co.ke/business/markets/2026-09-24-kq-ditches-20-year-old-booking-system-in-tech-overhaul

GDS economics
- Amadeus FY2025 results: https://amadeus.com/en/newsroom/press-releases/amadeus-fy-2025-results-revenue-growth
- Sabre FY2025 results: https://www.sec.gov/Archives/edgar/data/1597033/000162828026008796/a2025q4earningsrelease.htm
- Travelport is privately held. Its share of the distribution revenue figure is a V1 Advisory estimate.

Licensed under a Attribution 4.0 International License.
The V1 Airline Retailing Report publishes every Monday. Subscribe on Apple Podcasts, Spotify, or wherever you listen.
V1 Advisory LLC | v1advisory.co

Episode: 022

Published: Monday, September 21, 2026

Episode Description

Sabre says its cached airline offers are up to 95% accurate, and publishes that as a win. Read it the other way: as many as one answer in twenty may not be what the airline would have said. Eric and Steph pick up last week's line, "the number is right, and the value is somewhere else," across a cached offer, a discovery market, and the industry's own NDC scoreboard.

The live personalized offer is being cached from both ends of the chain. Last week a model lab repriced memory and gave the demand side a reason to cache. This week, the supply side. Sabre's Cache-powered Intelligent Shopping markets up to a 28% look-to-book reduction and "up to 95% accuracy versus live polling." Amadeus's NDC Offer Repository does the same thing without the word. IATA Resolution 787 built NDC for "an intelligent response for all products based on who is asking." You can cache a segment. You cannot cache an individual, in-the-moment offer. And the airline has no vote on whether the agent on the other end caches its answer.

The AI front door is splitting three ways, and trust never caught up. Phocuswright's September 4 European research puts standalone AI platforms at 44 to 51% of AI trip research, AI inside OTAs and metasearch at 21 to 33%, and AI in search engines at 25 to 34%. Only 17 to 20% of travelers trust AI results. If discovery never consolidates into one agent, leverage belongs to whoever is legible to all of them. A capability you build, not a protocol you join.

Scoreboard check: 82 airlines, 73 capabilities, and a ticket underneath. AltexSoft's 2026 datasheet shows NDC growth concentrated in payments and order servicing. Real progress. But a yes in the payments column is not a settled order, and nobody publishes how many orders clear with no ticket underneath.

The Bottom Line

At every layer the scoreboard is accurate and measures the wrong variable. The incumbents publish look-to-book savings, not freshness. The industry tracks discovery share, not offer legibility. The capability grid counts what airlines say they can do, not orders that settle. In all three, the number that decides the outcome is the one nobody publishes.

Stories Referenced

Both Neighbors Already Decided
- Sabre, Cache-powered Intelligent Shopping: https://investors.sabre.com/news-releases/news-release-details/sabre-solves-look-book-constraints-cache-powered-intelligent
- Amadeus, NDC Offer Repository: https://amadeus.com/en/blog/articles/enabling-innovative-travel-inspiration-with-ndc-offer-repository
- IATA Resolution 787: https://www.iata.org/contentassets/6de4dce5f38b45ce82b0db42acd23d1c/ndc-resolution-787.pdf

The AI Front Door Splits Three Ways
- Phocuswright Europe research via Travel Professional News:https://travelprofessionalnews.com/new-phocuswright-research-european-travelers-are-using-ai-more-but-trusting-it-less-than-friends-family-and-human-advisors

Scoreboard Check
- AltexSoft, NDC Airlines List and Their Retailing Capabilities (2026): https://www.altexsoft.com/infographics/ndc-airlines-list/
- IATA, Distribution with Offers and Orders fact sheet (June 2026): https://www.iata.org/en/iata-repository/pressroom/fact-sheets/fact-sheet-ndc/

Licensed under a Attribution 4.0 International License.
The V1 Airline Retailing Report publishes every Monday. Subscribe on Apple Podcasts, Spotify, or wherever you listen.
V1 Advisory LLC | v1advisory.co

Episode: 021
Date: September 14, 2026

Three layers of airline retailing got declared finished in the last two weeks. In all three, the number is right and the value is somewhere else. Nobody made an error in any of them, which is what makes them worth an episode.

Delta arrives at NDC late, and by arriving proves the offer layer is table stakes. A Delta spokesperson told Business Travel News on August 27 that the carrier plans to launch NDC by year-end. It is last among the big three legacy carriers and it refused both levers the others used, no Distribution Cost Charge and no NDC-exclusive fares. What it built instead is servicing, the part live NDC has consistently done worst. Critical take: ARC puts NDC at 18.4% of settled transactions in December 2023 and 21.2% in December 2025, then flat around 21% for a year. The offer layer stopped growing, so the only part of Delta's build that can matter is servicing parity on day one.

Amadeus announced openness just under seven weeks after closing its own developer door, and nothing has shipped.What is actually open is a waitlist. A developer cannot call a single Amadeus endpoint from Claude Code today, and on July 17 Amadeus decommissioned the self-service portal that used to be its most open channel. Critical take: watch the access-control layer, not the headline. Who gets off the waitlist, what they can transact and what it costs are the product, and none of it is written. That leverage is also not new, and the episode traces it back to 1984, when regulators took the biased CRS screen away and the advantage simply moved.

A cache-read price cut made the airline's live offer four times more expensive to ask for. On September 1 Anthropic cut cached-context reads by 75%, $1.00 to $0.25 per million tokens, and left fresh input at $10. The ratio between remembering and asking went from 10:1 to 40:1. One artifact in agentic shopping can never be cached, and it is the airline's live, personalized, built-for-this-traveler offer. Critical take: the personalization-caching scissors used to be an argument about incentives. It is now an argument about a published price list. Add one line to the distribution dashboard, the ratio between cache read and fresh input, every time a lab ships.

The Bottom Line

This industry keeps declaring layers finished, and every time the value quietly relocates somewhere the airline does not sit. Delta will finish the offer layer and find the value already gone to servicing. Amadeus took the headline and left the terms that constitute the actual product unwritten. And the airline's most advanced retailing capability, the live personalized offer, is now the most expensive thing in the chain for a machine to request. We built the one thing the buyer's agent has the strongest financial reason to skip. Nobody planned that. It is just what the price list says.

Corrections

Three corrections owed on the State of the GDS series, all discharged here.

State of the GDS, Part 1 said the GDS companies built the industry's settlement and messaging plumbing. They did not. BSP is IATA's, running since roughly 1971. ARC is owned by seven airlines. EDIFACT is a UN standard, ISO 9735. What the GDSs did, still a real achievement, is integrate it at scale from one agency terminal.

Part 3 said Sabre had "returned to profitability." That is continuing operations only. Net income attributable to common stockholders fell, $8.1M in Q1 2026 against $35.3M in Q1 2025.

Part 2 said Finnair's native Order was "direct channel only: confirmed." That was our inference, not a confirmation. Neither Amadeus release defines channel scope.

Stories Referenced

Delta plans NDC launch by year-end (Aug 27, 2026)
https://www.phocuswire.com/news/distribution/delta-plans-ndc-solution-launch-by-year-end
https://www2.arccorp.com/about-us/newsroom/2026-news-releases/december-2025-air-ticket-sales/

Amadeus brings travel intelligence to AI agents (Sept 3, 2026)
https://amadeus.com/en/blog/articles/anthropic-trusted-travel-intelligence-ai-agents
https://www.phocuswire.com/amadeus-shut-down-self-service-apis-portal-developers
https://skift.com/2026/07/21/sabre-hackathon-ai-agentic-voice/

The price of memory (Sept 1, 2026)
https://www.anthropic.com/claude-fable-and-mythos-5-1
https://platform.claude.com/docs/en/docs/build-with-claude/prompt-caching

Why the industry is built this way: IATA Type B Messaging Whitepaper v2.5; IATA Resolution 787; GAO RCED-92-130.

About the Show

Produced by V1 Advisory LLC, every Monday. Hosted by Eric Marketts and Steph Nell.

Powered by Jellypod.com: https://go.jellypod.com/rQZEXiO

Subscribe on Apple Podcasts, YouTube or Spotify. © 2026 V1 Advisory LLC. | v1advisory.co has the strongest financial reason to skip. Nobody planned that. It is just what the price list says. **Corrections** Three corrections owed on the State of the GDS series, all discharged here. State of the GDS, Part 1 said the GDS companies built the industry's settlement and messaging plumbing. They did not. BSP is IATA's, running since roughly 1971. ARC is owned by seven airlines. EDIFACT is a UN standard, ISO 9735. What the GDSs did, still a real achievement, is integrate it at scale from one agency terminal. Part 3 said Sabre had "returned to profitability." That is continuing operations only. Net income attributable to common stockholders fell, $8.1M in Q1 2026 against $35.3M in Q1 2025. Part 2 said Finnair's native Order was "direct channel only: confirmed." That was our inference, not a confirmation. Neither Amadeus release defines channel scope. **Stories Referenced** Delta plans NDC launch by year-end (Aug 27, 2026) https://www.phocuswire.com/news/distribution/delta-plans-ndc-solution-launch-by-year-end https://www2.arccorp.com/about-us/newsroom/2026-news-releases/december-2025-air-ticket-sales/ Amadeus brings travel intelligence to AI agents (Sept 3, 2026) https://amadeus.com/en/blog/articles/anthropic-trusted-travel-intelligence-ai-agents https://www.phocuswire.com/amadeus-shut-down-self-service-apis-portal-developers https://skift.com/2026/07/21/sabre-hackathon-ai-agentic-voice/ The price of memory (Sept 1, 2026) https://www.anthropic.com/claude-fable-and-mythos-5-1 https://platform.claude.com/docs/en/docs/build-with-claude/prompt-caching Why the industry is built this way: IATA Type B Messaging Whitepaper v2.5; IATA Resolution 787; GAO RCED-92-130. **About the Show** Produced by V1 Advisory LLC, every Monday. Hosted by Eric Marketts and Steph Nell. Powered by Jellypod.com: https://go.jellypod.com/rQZEXiO Intro music: Lundstroem, CC BY 4.0. Subscribe on Apple Podcasts or Spotify. © 2026 V1 Advisory LLC. | v1advisory.co

Episode: 020 Date: September 7, 2026

Two layers of airline retailing moved this year. One hasn't moved since 2024. The order started leaving the PSS, a machine finally booked a trip end to end, and NDC has sat flat for two years while the industry called it scaling on a number no institution publishes. Where retailing moved, somebody else owns it.

Three narratives, one structure underneath all of them.

NDC didn't scale this year. It plateaued two years ago, and the number everyone quotes to say otherwise has no source. The 2026 trade-press figure — roughly 24% of indirect sales globally, up from 11% in 2023 — traces to one travel-technology vendor's blog post, attributed to no research firm, data provider or industry body. IATA's own June 2026 Distribution with Offers and Orders fact sheet carries no adoption statistic at all. What does exist is ARC's monthly NDC share of ARC-settled US transactions: 21.5% in June 2024, 20.8% in March 2026, 20.1% in April, 21.6% in May and June, 21.2% in July. Two years inside a two-point band, and the only audited series in the industry. Critical take: the plateau is the disintermediation thesis reporting its own result. Share stopped climbing exactly where growth required airlines to build direct connectivity at scale, and the content routed through the aggregator instead.

Every vendor will now sell you an order management system, and that's the warning, not the milestone. The heart of Offer and Order was never the NDC message. It was pulling the order out of the PSS into a system the airline controls, because whoever holds the order holds servicing, settlement and the customer record. In 2026 the OMS became the “gravitational centre” of retailing, with nine or ten vendors selling one and half the field advertising it's “PSS-neutral” — while the same coverage adds they're “not fully independent from the PSS.” The dated deal that shows what that's doing to ownership sits one layer over: on June 25, Amadeus published that British Airways is retiring the NDC platform it built in-house on the 17.2 standard, after a decade running it, for Amadeus Altéa NDC on 21.3. Critical take: the carrier that went furthest toward owning its retailing technology decided the vendor should carry the next version.

A machine finally booked a trip end to end this year. It did it inside somebody else's policy. For consumer connectors the received wisdom holds: Expedia's connector in Claude searches flights and hotels, returns live pricing and availability, and stops there. But in July, Amex GBT launched the Egencia AI Connector in Claude, which searches, books and manages air and hotel within corporate travel policy — what Amex GBT calls the first deployment of its agent-to-agent architecture. Machine-executed booking didn't start in leisure and trickle into managed travel. It started in managed travel. Critical take: an agent can be held to a policy. It cannot be held to a preference. That's why managed travel went first, and why a decade-long push toward the personalised per-traveler offer just met its first machine buyer inside the one channel built to override individual preference with corporate rules.

The Bottom Line

This is not an industry that can't get started. It's an industry that modernized without renegotiating, then measured itself with a number nobody publishes. A better offer is not control of the offer. A dedicated order system is not independence if the order can't leave it. Being bookable by an agent is not owning the transaction if somebody else writes the rules. One question at every layer: when you buy the modern version of anything this year, what share of the value do you control, and what share are you renting? Get the honest number before you sign.

Stories Referenced

The NDC plateau and the number with no source https://www.prnewswire.com/news-releases/july-us-travel-agency-air-ticket-sales-total-9-6-billion-302856518.html https://www.iata.org/en/iata-repository/pressroom/fact-sheets/fact-sheet-ndc/ https://www.travelport.com/press-releases/royal-jordanian-airlines-goes-live-with-ndc-on-travelport

The order management system became the category https://travelinmotion.ch/2026/01/06/modern-airline-retailing-outlook-2026-steady-momentum-and-initial-breakthroughs/ https://amadeus.com/en/blog/articles/british-airways-adopting-altea-ndc

The machine learned to book, inside somebody else's policy https://www.amexglobalbusinesstravel.com/press-releases/amex-gbt-launches-business-travel-agent-to-agent-infrastructure-debuts-egencia-ai-connector-in-claude/ https://claude.com/connectors/expedia

About the Show

Produced by V1 Advisory LLC, every Monday. Hosted by Eric Marketts and Steph Nell.

Powered by Jellypod.com: https://go.jellypod.com/rQZEXiO

Subscribe on Apple Podcasts or Spotify. © 2026 V1 Advisory LLC. | v1advisory.coetter offer is not control of the offer. A dedicated order system is not independence if the order can't leave it. Being bookable by an agent is not owning the transaction if somebody else writes the rules. One question at every layer: when you buy the modern version of anything this year, what share of the value do you control, and what share are you renting? Get the honest number before you sign. **Stories Referenced** *The NDC plateau and the number with no source* https://www.prnewswire.com/news-releases/july-us-travel-agency-air-ticket-sales-total-9-6-billion-302856518.html https://www.iata.org/en/iata-repository/pressroom/fact-sheets/fact-sheet-ndc/ https://www.travelport.com/press-releases/royal-jordanian-airlines-goes-live-with-ndc-on-travelport *The order management system became the category* https://travelinmotion.ch/2026/01/06/modern-airline-retailing-outlook-2026-steady-momentum-and-initial-breakthroughs/ https://amadeus.com/en/blog/articles/british-airways-adopting-altea-ndc *The machine learned to book, inside somebody else's policy* https://www.amexglobalbusinesstravel.com/press-releases/amex-gbt-launches-business-travel-agent-to-agent-infrastructure-debuts-egencia-ai-connector-in-claude/ https://claude.com/connectors/expedia **About the Show** Produced by V1 Advisory LLC, every Monday. Hosted by Eric Marketts and Steph Nell. Powered by Jellypod.com: https://go.jellypod.com/rQZEXiO Intro music: Lundstroem, CC BY 4.0. Subscribe on Apple Podcasts or Spotify. © 2026 V1 Advisory LLC. | v1advisory.co

Episode: 019 Date: August 31, 2026

Delta's president described the destination as "offer management" — a price available on that flight, at that time, to you, the individual. Delta's lawyers told the Senate there is "no fare product Delta has ever used, is testing or plans to use that targets customers with individualized prices based on personal data." Read those carefully and they do not contradict each other. One says offer. One says price. Whether that distinction is a real defence or wordplay is the most consequential unanswered question in airline retailing.

Three stories, three chokepoints, unwritten terms at every one.

Offer or price? Delta's whole defence sits on one word. On August 11, Rep. Frank Pallone — Ranking Member of House Energy and Commerce, so a letter, not a subpoena — sent 19 questions to eight carriers. It expands the same letter he sent 25 corporations in May, and two of the nineteen are about tariffs, which tells you how airline-specific it is. No reporting shows what any carrier filed, so we read the question, not the answers. This is Delta's third round: senators in July 2025, two dozen House members that November quoting its president back at it, now this. Critical take: the industry already tried to fix this and got halfway. A4A's Chris Sununu told House Judiciary in June he would back banning surveillance pricing, "100%," and that no member does it — then never defined the term he volunteered to be banned from, while every vendor deck still sells personalization as the point of Offer and Order. Nor is the risk hypothetical: JetBlue has been defending a surveillance-pricing class action since April, and JetBlue is on Pallone's list.

The rail went to whoever a machine could safely call. The shorthand everyone uses is wrong. OpenAI's March walkback was retail checkout settled through Shopify — travel never got that far, because Expedia and Booking were apps that always handed you off. The largest consumer AI audience on the planet has never once closed a flight. In May, Mindtrip launched what its partners call travel's first all-in-one agentic AI flight booking — their phrase, attributed not endorsed — on Sabre Mosaic's agentic-ready air APIs and PayPal's checkout. Mindtrip owns the conversation and is the most replaceable piece. Sabre owns the rail. PayPal owns the money. The airline is content. Critical take: Sabre won not by having better AI than OpenAI, but by having built governed, authenticated access before anyone showed up wanting it.

Nobody wants the commission, and that should frighten you. The plumbing shipped — the Agentic Commerce Protocol in September 2025, Google's Universal Commerce Protocol in January — but who is owed a commission and who is merchant of record when an agent books remains untouched in public. Critical take: everybody is asking who gets the commission and it is the wrong question. The platforms decline it on purpose, because merchant of record is liability without leverage. The fight is for the conversation, and it monetises through attention — worse than the OTA era, because an ad auction has no IATA, no standard, no auditable settlement rail, no published take rate.

The Bottom Line

Three chokepoints, unwritten terms at every one, and in all three the party writing them is not the airline. This industry is very good at building things and very bad at defining them: the engineering ships and the definitions never do. Two metrics. New this week: can your commercial lead and your general counsel state the line between offer-level and person-level pricing in one sentence, and is it the same one? Standing: authentication, and a price for the query. An unwritten rule is not neutral space. Somebody else will fill it.

Stories Referenced

Delta AI pricing and the Congressional inquiry https://pallone.house.gov/media/press-releases/pallone-presses-us-airlines-answers-surveillance-pricing https://news.delta.com/delta-responds-misinformation-around-ai-pricing https://skift.com/2026/06/24/airline-lobby-backs-ban-on-surveillance-pricing/ https://www.constellationr.com/insights/news/delta-starting-scale-its-ai-driven-dynamic-pricing-system

Mindtrip / Sabre / PayPal agentic flight booking https://investors.sabre.com/news-releases/news-release-details/mindtrip-launches-travels-first-all-one-agentic-ai-flight https://skift.com/2026/05/06/sabre-mindtrip-paypal-launch-agentic-ai-travel-booking/

Agentic commission and merchant of record https://www.hospitalitynet.org/opinion/4133767/the-agentic-booking-question-no-one-has-answered-who-gets-the-commission https://skift.com/2026/03/05/openai-chatgpt-checkout-walkback/

About the Show

Produced by V1 Advisory LLC, every Monday. The stories that matter in airline and travel retailing. Hosted by Eric Marketts and Steph Nell.

Powered by Jellypod.com: https://go.jellypod.com/rQZEXiO

Episode: 18 Date: April 24, 2026

An airline went live on next-generation NDC. Another launched zero percent financing. Google bought a dead airline's data for ten million dollars. Read the document underneath each and the airline that went live is on a portal its own executive calls a pilot, the introductory zero percent offer expired August 15, and Google has not bought anything because a judge has not signed. Live, free, and sold. Not one survives its own paperwork.

Three stories, one question: who actually holds the thing?

Amadeus says Finnair is first live with Nevio NDC. The agent docs say pilot. The products launch on FAST, Finnair's agent portal, which its VP of digital, revenue and retail calls "our pilot platform." FAST still issues and services tickets; Amadeus's own words put ticketless, Order-based settlement "in the future." FAST does not support interline at all. Finnair remains on Altéa with no published retirement date, and neither the Q4 2025 nor H1 2026 investor report presents this as a material programme. Set that against IATA's own Readiness Survey, page 13, where the fourteen responding Consortium airlines rate their PSS vendors 0% "operational and proven" and 29% "mainly vision and slideware." The announcement and the survey are one finding from two directions. Riyadh Air is the strongest counter-example and proves the point: Order-native at its core, translating back to legacy at every boundary a counterparty still needs.

JetBlue put a lender inside its loyalty programme. The lender is not JetBlue. The July 15 ClarityPay launch surfaces instalment terms while the customer is still shopping, six weeks to 48 months, TrueBlue points earning. Three things the coverage moved past: the date is July, not August; the blanket introductory 0% APR expired August 15, with programme loans still running 0% to 36%; and loans are provided by DR Bank, with ClarityPay Program Services explicitly not the lender. Critical take: financing moved up the funnel, and anything shaping an offer before the offer is formed is retailing, not payments. JetBlue did not become a lender. It rented one, kept the loyalty relationship and the shelf position, and left the lending to a bank.

A court is about to price an airline's data. Google won a section 363 bankruptcy auction for Spirit Airlines' internal data at $10 million, with AI recruiter Mercor.io backup at $7.5 million. Roughly 100 million emails, 30 million lines of source code, 190 million PNRs and 7.5 billion transactions with years of booking curves. Passenger and loyalty records are carved out. It is not approved: the Association of Flight Attendants objected on reidentification grounds and the August 19 hearing moved to September 9. Careful with the over-claim. IATA already sells Direct Data Solutions, built from BSP and ARC settlement transactions, and ARC's Travel Intelligence Program sold agency ticketing data to CBP and ICE until press and congressional pressure shut it down. Section 363 has covered data sales since 2005. A price on airline data is not new. What we could not find priced anywhere is an airline's internal operating corpus, in a public docket instead of a concealed contract.

The Bottom Line

The announcement is a marketing document. The contract is the product. To find what is real, find the party who had to be accurate because somebody would otherwise have sued, audited, or refused to pay them. Then ask who holds the thing. Finnair holds an Order and the ticket still holds the settlement. JetBlue holds the customer and a bank holds the loan. Spirit's estate holds the data and a judge holds the decision. The party in the headline is not the party carrying the risk, and the party carrying the risk is where the strategy lives. Standing ask, unchanged: authentication, and a price for the query. A court has now priced an airline's data corpus. We cannot find a carrier that has priced its own.

Stories Referenced

Amadeus / Finnair Nevio NDC (Aug 13, 2026); Finnair FAST agent docs; IATA Readiness Survey, 3rd ed.

JetBlue and ClarityPay (July 15, 2026)
https://www.prnewswire.com/news-releases/jetblue-and-claritypay-launch-first-personalized-pay-later-program-with-trueblue-points-earning-302825749.html

Spirit Airlines data sale, Case 25-11897 (SHL), SDNY
https://ppc.land/google-wins-bankrupt-spirit-airlines-data-for-10-million/
https://nydailyrecord.com/2026/08/19/us-court-delays-hearing-on-googles-purchase-of-spirit-airlines-data-as-union-objects/

About the Show

Produced by V1 Advisory LLC, every Monday. The stories that matter in airline and travel retailing. Hosted by Eric Marketts and Steph Nell.

Powered by Jellypod.com: https://go.jellypod.com/rQZEXiO

Subscribe on Apple Podcasts or Spotify. © 2026 V1 Advisory LLC.

Episode: 017 Date: August 17, 2026

Lufthansa Group charges 23 euros to book through Travelport on EDIFACT. Eight euros for the identical seat through Travelport's NDC channel. Zero through almost any other door. Same company, two roads: what is priced is the road, not the seller. And every number this industry quotes about NDC adoption counts all three as one.

Two stories this week, and each is a number describing something that does not exist.

Lufthansa publishes the price of access. Amadeus rose to 19 euros this year on EDIFACT, Sabre to 22.50; Travelport has been 23 since 2022. NDC Public held flat at eight. And the tier nobody covered: Bilateral carries no DCC. The names mislead — Public is, in LHG's words, "available through GDS NDC aggregators only," while Bilateral is everything else and needs no scale and no engineers. Critical take: follow the money. Lufthansa collects the eight on the ticket as a carrier ticketing fee, so the traveller pays it. The airline uses it to offset what the GDS charges for that booking — Amadeus told investors it expects to earn as much per NDC booking as EDIFACT, if not more — and Travelport confirms GDS incentives still apply to NDC, so part flows back to the agency. Traveller pays, airline passes it on, GDS keeps a margin and rebates the seller. On the standard sold for a decade as removing the middleman. Structural proof: bilateral is zero and includes independent aggregators, who charge the seller not the airline. Nobody publishes how much. And the fee is the smaller half anyway — the Base Offer has no Economy Basic or Light and no group bookings. Cost recovery explains a charge. It does not explain removing group bookings.

Two direct connects — and do they survive AI agents? Navan upgraded Singapore Airlines to a direct connection on August 10; Spotnana followed on the 13th. Navan's 70% NDC figure is an unsourced vendor claim with two incompatible published versions. Then a correction we owe on our own watch item. Four episodes running we said no carrier publishes authentication, rate limits or a price for the query. Two stand. The rate limit does not: Lufthansa's NDC terms, section 7.1, cover shopping costs "until a maximum Look-to-Book Ratio of 500:1," above which the group reserves the right to deactivate access. Unchanged since May 2023. Not an agent policy — no AI or bot language, written for agencies polling badly — but we spent four episodes asking for a document we had not read. Industry look-to-book is 10,000 to 20,000 to one and agentic projections run to 200,000, so a seller at agent-era ratios is 20 to 400 times over the published cap. Access stops being technical and becomes contractual, and only sellers the airline can identify hold a contract. Direct connects survive as a credential, not a pipe. Which reframes that zero tier: Lufthansa is not giving away access. It is buying identity.

The Bottom Line

This industry is running several realities at once and quoting them as one, and most are built on wishes rather than data. We are not outside that: we spent four episodes asking for a rate limit that had been sitting in a contract since 2023. So go to the document somebody had to be accurate in: the rate card, because the airline bills against it; the terms, because they must be enforceable; the agent docs, because sellers have to transact; the investor report, because shareholders have to be told. Every real finding here came from one of those, not from a press release. Read the rate card, not the roadmap.

Stories Referenced

Lufthansa DCC Guideline and NDC terms (sec. 7.1) https://business.lufthansagroup.com/content/dam/b2b/experts/files/LHG_DCC_Guideline_05MAY26.pdf https://lhgroupairlines.com/fileadmin/ndc-assets/TermsConditions/GTC_NDC_Offer_EN_version_26MAY.pdf

Navan / Singapore Airlines https://www.businesswire.com/news/home/20260807666419/en/Navan-Upgrades-Singapore-Airlines-NDC-to-Direct-Connection

About the Show

Produced by V1 Advisory LLC, every Monday. The two or three stories that matter in airline and travel retailing, with the analysis behind them. Hosted by Eric Marketts and Steph Nell.

Powered by Jellypod.com: https://go.jellypod.com/rQZEXiO

Subscribe on Apple Podcasts or Spotify. © 2026 V1 Advisory LLC. | v1advisory.co

Episode: 016

Date: August 10, 2026

May 28, 2025: Southwest starts charging for a checked bag. January 27, 2026: fifty-four years of open seating end. August 3, 2026: Southwest announces NDC. Read those three dates in order and the story tells itself.

Last week we argued the pipe stopped being scarce, and what is scarce now is the offer and the shelf. This week the last major US holdout proved it, then published an architecture through 2027 that says nothing about the buyer it is getting. Three narratives, one sequencing error.

Southwest folds, and the reason is the seat map. NDC connectivity plus Business Priority, a corporate tier landing early 2027. The direct-connect API is due by end of 2026, Altéa NDC later in 2027. Bull case: corporate demand forced it, and Amadeus gets the US reference it needed. Bear case: soft timelines, a doubled integration surface, and transaction economics Southwest spent thirty years avoiding. Critical take: everyone has the causation backwards. Southwest changed the product, then found no channel could sell what it built. NDC adoption tracks product complexity, not standards maturity.

The agent layer gets a production floor, and "MCP versus NDC" is a category error. Travelport's TripServices went live with Travelsoft on July 1 across 400+ European agencies. Neither company said MCP; the deterministic layer is live, the framing on top is analysis. Bull case: it closes the fulfillment gap NDC never targeted, with no re-standardization. Bear case: MCP is an integration convention, not a settlement or liability framework, and Expedia's March survey found just 8% of travelers comfortable booking through AI. Critical take: asking whether MCP beats NDC is like asking whether English beat the telephone. That framing sells, and it costs you.

IATA standardizes the back office while the front end moves. The Consortium's readiness work is in its third edition: 20 IT providers surveyed, seven full-scope airline contracts, forty-eight pilots, legacy retirements from 2027. IATA's vision paper cites McKinsey at up to $7 per passenger, against an aspirational 2030 date. Bull case: Settlement with Orders is the sleeper. A CFO funds a cash-cycle improvement long after they stop funding a retailing vision. Bear case: every workstream is internal, and none says how an agent authenticates against an Order. Critical take: Offers and Orders is necessary and oversold. Fund it as an operational upgrade, not customer acquisition.

The Bottom Line

Southwest ran the correct sequence in plain sight and the industry called it conversion. Change the product, then the channel. Not the reverse. Run a three-year NDC program for a product nobody needs an offer to sell and you have bought a channel you cannot use. And notice what is missing from the year's biggest distribution announcement: a full architecture through 2027, not one word about machine buyers. Third episode running, we have looked for a carrier that published agent terms and found none. Authentication. Rate limits. A price for the query. We also withdraw Episode 15's partial credit to Air France-KLM: it is Amadeus tech from December 2025 and targets NDC search generally, not agents. Change the question you ask of your roadmap. Not what percentage of Offers and Orders is complete. Whether it names the buyer you will have in 2028.

Stories Referenced in This Episode

Southwest Commits to NDC

https://www.travelweekly.com/Travel-News/Airline-News/Southwest-to-add-NDC-connectivity-Business-Priority-option

The Agent Layer

https://www.travolution.com/news/travel-sectors/air/travelsoft-offers-travelports-airline-content-to-agencies-via-orchestra/

https://traveldistributionnews.com/travelport-and-travelsoft-bet-on-mcp-now-comes-the-real-test-across-400-agencies/

IATA and 100% Offers and Orders

https://airlines.iata.org/2025/12/18/progress-journey-100-offers-and-orders

https://www.iata.org/en/programs/airline-distribution/retailing/consortium/

https://www.travelweekly.com/Travel-News/Travel-Technology/ARC-reports-rising-NDC-adoption-2025

About the Show The V1 Airline Retailing Report is produced by V1 Advisory LLC and publishes every Monday. Every episode surfaces the two or three stories that matter most in airline and travel retailing, and delivers the 360-degree analysis that helps commercial leaders, distribution professionals, and travel technology executives understand what's really happening and what to do about it.

Powered by Jellypod.com, the AI podcast platform behind this show. Check it out and use our referral link: https://go.jellypod.com/rQZEXiO

Hosted by:

Eric Marketts — Tech and aviation journalist, co-host

Steph Nell — Airline distribution expert and consultant, co-host and analyst

The V1 Airline Retailing Report publishes every Monday. Subscribe on Apple Podcasts or Spotify. © 2026 V1 Advisory LLC. All rights reserved. | v1advisory.co

Episode 015 | August 3, 2026

Early online era: 100 to 200 searches per ticket sold. Today: 10,000 to 20,000. OAG's estimate for an agent-driven world: 200,000. Same one ticket at the end.

First weekly episode after our State of the GDS season, which ended on one line: same machine, three bets. This week they showed up at once, and you can see what they are aimed at. Not the pipe. The offer layer above it.

The GDS oligopoly stopped running the same playbook. Amadeus bought SkyLink in February, a working conversational booking engine. Sabre alleged in May that Amadeus holds a dominant monopoly position in PSS. Travelport launched TripServices, wired in Anthropic's MCP, and moved API-driven transactions from 43% of volume in 2022 to 63% today. Bull case: real architectural choice, first time in fifteen years. Bear case: three incompatible integrations for every TMC, OTA, and aggregator. Critical take: none of these is a bet on the pipe, and they are not the same bet. Above the pipe are two jobs: building the offer, and owning the shelf it lands on. Amadeus is not choosing; it can fund both. Sabre is hedging. Travelport cannot hedge, has no airline PSS, and went all in on the shelf. Last on the scoreboard everyone quotes, first on modern-API share.

We asked who would govern agent traffic. Amadeus answered, and the answer is a retreat. In Episode 8 we covered the traveler who pointed an AI agent at Etihad and pulled 881,076 fares for one ticket, and left a watch item: the first airline to publish an explicit AI traffic policy. Eight weeks on, the first real answer came from Amadeus, and it is not a policy. It is precomputed fares. Bull case: machine-scale demand forces the authenticated, metered offer API that is the most durable control point in distribution. Bear case: you cannot precompute context, so the answer to the agent era is a step back toward the cached-fare world NDC existed to escape. Critical take: the industry is optimizing the cost of the look instead of pricing it. An engineering answer to a commercial problem. The watch item stands.

The OMS is being named the new center of gravity, as PSS contracts expire. Travel in Motion argues airlines and vendors should treat the order management system, not the PSS, as retailing's "gravitational centre". Amadeus positions Nevio as an OMS; Navitaire holds IATA ONE Order Capable status and unveiled Stratos. Bull case: order-native platforms free retailing from the fare-class straitjacket, and contract expiry is the only moment a carrier has leverage and budget at once. Bear case: "we're going to replace the PSS" is the most dangerous sentence in airline IT. Critical take: order-native is necessary, not sufficient. If the new OMS cannot authenticate an agent, meter it, or price for it, you have replaced your back office and still cannot compete in the channel coming.

The Bottom Line

For thirty years the power in this business sat with whoever controlled the pipe between the airline and the traveler. This month the pipe stopped being scarce. What is scarce now is the offer, a correct, current, governed, machine-readable answer to a specific question from a specific buyer, and the shelf that answer lands on. So change your metric. Booking share told you who won the last thirty years. Watch instead for a carrier publishing real agent terms: authentication, rate limits, and a price for the query. That is an industry treating machine demand as a customer rather than as weather.

Correction

The figure of 200,000 searches per ticket sold in an agent-driven world is OAG's estimate. On air we called it OAG's projection. It is neither a formal forecast nor measured data, and our own production notes said so before the script overrode them. Corrected August 2026.

Stories Referenced in This Episode

3 GDS, 3 Bets: https://traveldistributionnews.com/three-gds-three-bets-amadeus-sabre-and-travelport-are-fighting-over-who-owns-the-ai-era-booking/

200,000 Searches for One Ticket: https://www.oag.com/blog/airline-ai-interface

881,076 searches, Skift, June 1, 2026: https://skift.com/2026/06/01/ai-impact-travel-search-costs/

The Order as Center of Gravity: https://travelinmotion.ch/2026/01/06/modern-airline-retailing-outlook-2026-steady-momentum-and-initial-breakthroughs/

Navitaire unveils Stratos: https://www.travolution.com/news/exclusive-first-look-navitaire-unveils-next-gen-stratos-solution/

About the Show

Produced by V1 Advisory LLC, every Monday. The stories that matter in airline and travel retailing, with the analysis behind them. Hosted by Eric Marketts and Steph Nell.

Powered by Jellypod.com: https://go.jellypod.com/rQZEXiO

Subscribe on Apple Podcasts or Spotify. © 2026 V1 Advisory LLC. | v1advisory.co