On August 27, 2026, Google began rolling out hotel booking directly inside AI Mode in Search for users in the United States.
A traveler describes what they want in plain language, receives a visual list of properties with reviews and side-by-side comparisons, and completes the reservation through a “Continue on Google” flow paid with Google Pay.
Ten companies are in the initial rollout: Booking.com, Expedia, Hotels.com, Priceline and Trip.com on the platform side, and Choice Hotels, Hilton, IHG, Marriott and Wyndham on the hotel side. Independent and boutique properties are not among them.
The practical consequence is not that independents are shut out of AI discovery. It is that the transaction layer and the information layer have separated, and independents now compete entirely in the second one, where content, structured data and Google Business Profile accuracy determine whether a property is named at all.
Before the Booking: Closing the Hotel AI Discovery Gap to Drive Total Revenue
The new book from Amber S. Hoffman of The FS Agency. Travelers now plan entire trips — where to stay, eat, and spend — in conversations with AI, before they ever reach a booking site. Before the Booking shows hotel owners and operators how to make sure AI can see, understand, and recommend their property.
The book is available on Amazon via Kindle download or paperback. Secure your copy here.
The Distribution Layer Just Consolidated, and the Infrastructure Is Already Built
Look at who is in the launch group. Five online travel agencies and five major chains. For twenty years these two groups have been described as opponents, one fighting for direct hotel bookings and the other collecting commission. Inside AI Mode, they are on the same side of the table.
The plumbing underneath is worth understanding. Google’s Universal Commerce Protocol (UCP), extended to hotels at Google I/O in May 2026, is what allows a conversation to become a booking without leaving the interface.
Amadeus is Google’s foundational B2B partner for lodging on that protocol. Its head of hospitality product has stated it is Google’s only B2B technical partner on the hospitality roster for UCP. Sabre is advising on the airline side.
This is not a story about a vendor choosing chains over independents. Amadeus launched its AI Commerce product at HITEC in June, and it runs on Amadeus Central Reservation System, iHotelier and the Amadeus distribution platform, available to hotel customers using any of those. iHotelier is the booking engine behind a great many independent and boutique hotel websites. The road is being built and it does reach smaller properties.
What no protocol can do is make a model choose you. Being technically connected to a booking rail and being named in the answer that precedes the booking are different problems, solved by different work. The first is a distribution question. The second is a description question.
How Does Your Hotel Appear in AI Tools? Take our quick 6‑minute self‑assessment today
Your property can be excellent and still be invisible. The gap isn’t quality — it’s legibility: how clearly AI can read what you offer, across rooms, dining, spa, events, and the practical details guests actually search for. This walks you through the audit, then scores the gap.
Before you begin
Open ChatGPT, Gemini, and Perplexity in three tabs. Use them side by side.
For each section, paste the prompt — but swap in your city, neighborhood, and the details a real guest would mention. Talk to it like a person, not a search box.
Score honestly — you’re checking whether AI can describe you specifically enough to recommend. And if something doesn’t apply to your property — no bar, no spa — tap N/A; it won’t count against your score.
Want a second pair of eyes on your result?
Book a free 30-minute call with The FS Agency. Bring your score and we’ll walk through where your gap is, which guest searches you’re losing, and what’s worth fixing first. No pitch, no charge — just a clear read on where you stand.
Book a free 30-minute call →Google Pay Solved a Trust Problem That Was Holding AI Commerce Back
The most quoted statistic in this conversation is that 53% of travelers are comfortable letting AI suggest a hotel while 66% say they would not trust AI to book on their behalf. That gap is real, but it describes autonomous booking, where an agent transacts for you.
That is not what Google launched. The traveler still describes the trip, still sees the options, still compares, still chooses, still confirms. What has been removed is the moment where an unfamiliar interface asks for a credit card number.
Two mechanisms are doing the trust work:
- Google Pay. The credentials are already stored, the brand is known, and no new card details are entered into a chat window.
- Merchant of record. The hotel or booking platform remains responsible for the transaction and handles customer service. The traveler is not buying from an AI. They are buying from Hilton, with Google as the interface.
Even the language reflects it. “Continue on Google” signals continuity rather than a handoff to something unfamiliar.
This matters more than the feature itself. The barrier to commerce inside AI conversations was never really about the recommendations. Travelers have been comfortable with those for a while.
It was about payment security in an environment that felt experimental. Once one major platform makes that transaction feel routine, the behavior generalizes, and every AI surface becomes a place where booking is plausible rather than novel.
What Independent Properties Still Control
Bookings through conversational AI agents currently represent less than 1% of total room nights industry-wide, according to Booking Holdings’ own CFO. That number is the reason to act rather than the reason to wait. The behavior is forming now, and the properties that become legible to these systems early will be described accurately while the volume is still small.
Independent research indicates roughly 79% of hotel links inside AI Mode currently route through Google Business Profiles. That single figure should redirect a great deal of hotel marketing attention. The profile is not a legacy local SEO asset. It is a primary input into how AI describes a property.
What a property without a booking integration still controls:
- Google Business Profile. Categories, attributes, descriptions, photos, hours for every outlet, and the specific language used to describe rooms, dining, spa and event space.
- Structured property pages. Hotel schema, FAQ schema, and pages that state facts a model can extract rather than atmosphere it cannot use.
- Destination and area content. When AI builds an itinerary, it draws on material about the area, not only about the hotel. This is where a property with genuine local knowledge outperforms a chain running templated copy.
- Reviews and management responses. Both are machine-readable. A response that names the restaurant, the treatment or the room type adds retrievable detail the review itself may lack.
- Third-party sources. Regional publications, directories, planner resources and creator content are corroboration, and corroboration is what moves a property from known to recommended.
None of that requires a partnership. All of it requires specificity.
This is the point most hotels miss. A property can be genuinely excellent and still be invisible, because the gap is not quality. It is legibility. A page describing “an unforgettable escape” gives a model nothing to recommend on. A page describing a rooftop bar with harbour views, a spa with four treatment rooms including two for couples, and a ballroom that seats 300 gives it everything.
Will Google Open This Up?
Probably, and the shape of it is worth anticipating. Google has said it does not intend to become an online travel agency and plans to leave servicing to industry experts, which is difficult to square with a permanently closed list of ten.
The more likely path is the one Google has always taken: paid access. Ads now appear in roughly 25.5% of AI results, up from 5.17% in early 2025, and Google has already introduced shopping ads with Direct Offers inside AI Mode. If that pattern holds, smaller properties may eventually buy their way into AI placement much as they buy search ads today, and much as advertisers are now doing on ChatGPT.
Which makes the organic work more valuable, not less. A property that is already described accurately, cited by credible third parties and structured for extraction will pay less for the same visibility than one starting from nothing. That has been true of every advertising channel Google has built.
Frequently Asked Questions
Yes. The booking integration and the recommendation are separate systems. AI Mode still surfaces properties in its informational layer, drawing on Google Business Profiles, structured property pages, reviews and third-party sources. Roughly 79% of hotel links inside AI Mode route through Google Business Profiles, so profile accuracy and detail are the highest-leverage starting point for a property without a partnership.
UCP is Google’s open standard for agentic commerce, extended to lodging at Google I/O in May 2026. It allows an AI conversation to complete a transaction directly rather than sending the user to another site. Amadeus is Google’s foundational B2B partner for lodging on the protocol, and its AI Commerce product runs on Amadeus Central Reservation System, iHotelier and the Amadeus distribution platform. If your property uses one of those systems, ask your Amadeus representative what is available to you and on what timeline.
AI systems build their picture of a property from sources that accumulate at their own pace. Reviews arrive as guests write them. Third-party coverage has to be published, then found, then treated as reliable. Structured data has to be crawled and re-crawled before it is trusted. A Google Business Profile edit may be reflected within days, but corroboration from independent sources typically takes months. This is why a property that recently rebranded, changed management, or opened new outlets is often still described using outdated information — the older sources outnumber the newer ones until enough new material accumulates to shift the balance.