Airbnb Stops Saying 'AI Search' and Starts Building a Trip Agent
Q3 delivered the reacceleration management had promised for five straight quarters — and the vocabulary quietly rotated from search to agents, social graphs, and a loyalty program it now admits it lacks.
ABNB · Earnings Call · 2026-11-05
The promise, kept
For over a year, Airbnb's calls carried a recurring promise that the numbers kept failing to confirm. In August 2025 the CEO was explicit: “We are looking to reaccelerate the growth of Airbnb. We are not satisfied with the company growing approximately 10% year-over-year.” — Brian Chesky, Co-Founder and Chief Executive Officer · 2025-08-06 The market had heard versions of this before, and the stock spent years in neutral. This quarter, the promise finally met the tape: revenue of $4.9 billion, up 19%, gross booking value of $27.7 billion, up 21%, and 155 million nights and seats, up 16% — with double-digit nights growth in every region, the fastest North American pace in four years, and first-time booker growth the best since early 2022.
Airbnb is back in growth mode. It's rare for a marketplace of our size to reaccelerate growth, but that's exactly what we're doing. Q3 was our fastest revenue growth in more than 3 years.
The framing is itself new — growth mode did not exist as a theme in this company's keyword history until now, and neither did fastest pace or Gen Z as a demand-engine story. The other side of the ledger is the base management is compounding off: the last filed quarter showed total revenue of $3.6 billion, up 17% year over year, at an 82.5% gross margin. That is a business accelerating from an already profitable base, not a recovery off a trough.
The vocabulary pivot: 'AI search' disappears
The most telling signal in this dataset is not a new word — it is a missing one. AI search was the company's second-ranked theme in Q4 2025 and still a top-20 theme as recently as Q1 2026. It does not appear anywhere in the current top-30. In its place: chat agent, general agent, AI agent, operating system, and a distinction between a specialized agent and a general-purpose one that did not previously exist in the company's language.
That is a promotion, not a rename. For four quarters the AI story was a search-box toggle with a hedge attached — “AI search will eventually -- I can't put a time line on it because AI is obviously highly unpredictable.” — Brian Chesky, Co-Founder and Chief Executive Officer · 2026-02-12 Now there is a named product with a ship date: a proprietary travel agent rolling out in the first half of next year, built on MCP integrations so it can reach outside Airbnb's walls, with a deliberately non-chat interface of maps, photos, calendars and itineraries. The strategic argument has sharpened too: “I do think general chat agents are going to be popular for simple everyday things. I think as they get really, really complicated, Specialization is going to win.” — Brian Chesky, Co-founder and Chief Executive Officer · 2026-11-05
This is not a company-unique wave. The Muse agent and personal agent themes run through the global keyword set this quarter, and Pepijn Rijvers — Booking.com's former accommodations chief, now Airbnb's Chief Business Officer — shows up as a first-time keyword with an explicit Performance Marketing mandate. Airbnb is riding a market theme, but it has moved from talking about discovery to talking about ownership of the interface.
Worth noting the check against the fear: on the question of whether agentic traffic is already disintermediating them, the answer was blunt. “We do have nearly half a million bookings a day and a very non-material number of bookings happen on Muse and Instinct. So we're talking very, very low numbers... It kind of peaked in early October.” — Matthew Kost, Analyst · 2026-11-05 The threat is currently a rounding error.
Three options, one admission
The social graph is the genuinely new idea on the call, and the CEO sold it hard — as a trust proxy for hosts, a collaborative planning tool for group travel, and a hook into Gen Z. “I think this is maybe one of the single most underestimated parts of our business.” — Brian Chesky, Co-founder and Chief Executive Officer · 2026-11-05 Hotels, meanwhile, have graduated from a four-city pilot to thousands of boutique properties across 60+ destinations, with management asserting it will be a multiple-billion-dollar revenue line.
One statistic deserves scrutiny. In May, the CFO said “over 55% of people who book a hotel on the platform come back to book a home.” — Ellie Mertz, Chief Financial Officer · 2026-05-07 Today the anchoring number is approximately 35% within a year. That may be a definitional difference rather than a decayed conversion funnel — but it is the kind of anchor that migrates quietly, and it is worth flagging.
The admission is loyalty. Analysts have asked about it in call after call; the answer has been consistently evasive — "if we do something," "testing a lot of different tracks," "not imminently, but in the future." This quarter the CEO conceded ground: “We're like the only travel- major travel brand in the world that doesn't have one. And I think we have been at a disadvantage because we don't have a loyalty program.” — Brian Chesky, Co-founder and Chief Executive Officer · 2026-11-05 The loyalty program keyword has now appeared in four separate quarters without a launch — and the language has escalated from optionality to deficiency.
The bill for Reserve Now, Pay Later
All of this growth is not free. Implied take rate slipped to 17.6% as Reserve Now, Pay Later pushes the booking into the numerator of GBV while deferring the revenue, and cancellation rates rose 200–300 basis points to roughly 19% from about 16%. Management calls that healthy; the working capital drag is real enough that free cash flow growth in the quarter was muted by it. Gross margin held at 82.5%, but a $75 million step-up in property and equipment for a New York office is a small crack in the capital-light narrative.
The offset is a balance sheet that can absorb experimentation. Effective net cash stood at $21.8 billion, up 7% year over year, and the board layered a new $12 billion repurchase authorization on top of $2.7 billion remaining, after $700 million bought back in Q3 alone. Quarterly repurchases have hovered around $1 billion for eight straight quarters.
And the valuation is no longer the obstacle it once was. Price to revenue sits at 6.4x, down roughly 80% from the 32.7x peak of early 2021, with price to free cash flow at 26.8x. The tape agrees at the margin but not at the extreme: the stock is up 15.5% over 90 days, yet sits 10.7% below its August high and about 21% below its 2021 peak.
The reacceleration is real and the cash generation is real. What the market has not yet priced is whether a Trip Agent shipping into a crowded agentic field — with sponsored listings deliberately deferred until the interface exists — plus a social graph and a long-promised loyalty program can turn a good 2026 into a structurally different 2027. That is the bet this quarter asked investors to make.