The hotels that will win in AI Search are being decided now
Hotel owners are starting to ask a new question of their marketing teams: when someone asks an AI assistant for the best luxury hotel in a given location, does our property appear?
It is a reasonable question, but it often assumes that AI works like search. It does not.
Why traditional search measurement does not translate cleanly to AI
In search, measurement is relatively clear. A phrase such as “best boutique hotel Edinburgh” is a keyword. It has search volume. People type it into Google. A hotel ranks for it, or it does not. Rankings can be checked regularly, and while search results change, the measurement is repeatable enough to be useful.
AI discovery works differently.
In a conversational channel, there is no single keyword. The same guest intent can be expressed in many different ways:
“good hotel near the castle”
“dog-friendly boutique stays in Edinburgh”
“somewhere to stay this weekend that isn’t a chain”
“we’re coming for Hogmanay, where should we stay?”
Each version of the question can produce a different response. The same question may also return different answers at different times. Model version, context, retrieval mode, browsing access and other factors all affect the result.
That makes AI visibility harder to measure using search-style reporting.
The problem with AI visibility reports based on test queries
Some AI reporting tools work by asking a set of questions and checking whether a hotel appears in the answers. The result may look useful: “your hotel appeared in 73 AI results this month.”
The follow-up question should be: appeared in response to what?
If the questions were chosen by the supplier, phrased by the supplier and asked from the supplier’s own server, then the report is not showing real guest behaviour. It is showing the results of a controlled survey designed by the reporting provider.
That does not mean the data is worthless, but it should not be treated as audience data.
The more useful question is whether AI platforms are sending real users to the hotel, and whether that referral activity is increasing.
AI visibility is where website analytics used to be
AI reporting is still at an early stage. Much of it is being measured as if it were traditional search because that is the framework businesses already understand.
We expect this to change. Closed conversational networks are likely to provide more direct reporting over time, just as website owners gained clearer visibility once tools such as Google Analytics became widely available.
Until then, hotels have a choice. They can measure AI visibility using familiar methods that do not fully reflect how AI systems work, or they can start building the technical information assets that help AI systems understand, trust and cite their properties.
The hotels investing in that infrastructure now are more likely to be recommended in the future.
Why we built AthenaGeo
We built AthenaGeo to give hotels a more practical way to understand and improve their AI visibility.
It is not based on modelled estimates or traffic projections built on assumptions. It is designed to show whether AI networks are sending real traffic, enquiries and revenue to a property, and whether that activity is growing.
AthenaGeo also builds a complete, machine-readable profile of a hotel in the technical language used by Google, ChatGPT, Perplexity and other discovery platforms.
Without that structure, AI systems have to infer details from marketing copy or rely on third-party platforms that already provide structured information. In many cases, that means an OTA is better placed to answer the guest’s question.
That matters commercially. If a guest asks about check-in time, amenities, location or suitability, and an OTA provides the clearer answer, the booking may happen there instead of on the hotel’s own website.
Many major hotel platforms already use structured data in this way. Most independent hotel websites do not. That gap helps explain why OTAs are well positioned in conversational search results today.
AthenaGeo helps hoteliers see which AI platforms are sending real visitors to their property, while also improving the information structure that AI systems use to understand and recommend hotels.
As guests move from traditional search to AI assistants, hotels need to be readable by both.
How Internet Affected Works with Hotels
We work with hotels to build direct channel strategies grounded in verified booking data, not modelled attribution. From campaign structure and tracking setup to ongoing management at the individual property level, our approach is designed to give Revenue Managers the clarity they need to make confident investment decisions.
If the patterns described in this article look familiar, we’re happy to conduct a no-obligation review of your current campaign setup and identify where improvements can be made.
Get in touch to find out how we can help your hotel.
About the author
Glyn Spencer Hopkins is the owner of Internet Affected and has been working exclusively with hotels and luxury brands for over a decade.
Internet Affected provides digital revenue services tailored to the individual characters of hotels; a complete range of services designed to help them take back ownership of their hotel brand from the OTAs. Specialized marketing solutions to increase guest loyalty, food & beverage bookings, events and wedding inquiries, clearly reported in straightforward language.
