Hotel information is broken — and so is discovery
Google says AI is responsible for the decline in generic hotel search. The data points somewhere else.
Generic hotel searches are declining. Specific-name searches — travelers arriving at Google Search with a hotel already in mind — are up double digits. Google's explanation is that travelers now research in AI tools before returning to Google Search to confirm pricing and book.
It is a tidy explanation. It is also the most convenient one available to Google, because it frames the data as an orderly handoff — AI does discovery, Google closes the transaction — rather than what it more plausibly reflects: the accumulated result of Google Search becoming a poor product for hotel inspiration.
What Google Search did to its own discovery product
Generic hotel searches don't decline because travelers found something better. They decline when travelers stop expecting useful results. Over the past several years, generic hotel queries on Google Search returned pages dominated by paid placements — Hotel Ads, sponsored OTA listings, the Google Hotels module — with organic results pushed far enough down that content capable of serving genuine inspiration rarely surfaced where anyone would see it.
Travelers learned, through repeated experience, that typing "boutique hotel Milan" into Google Search returned a rate comparison they didn't ask for and a set of OTA links they could have reached directly. The product stopped answering the inspiration question. Generic searches declined because Google trained users to stop asking it those questions.
The rise in specific-name searches is therefore not primarily evidence that AI is doing discovery well. It is evidence that a significant share of hotel inspiration had already left Google Search — and had been gone for some time.
Where travelers actually get inspired
The largest driver of specific-name search growth that Google's framing omits entirely is creator content. A travel video on YouTube, a hotel reel on TikTok, a property featured on Instagram — these are surfaces where genuine inspiration happens at scale. A traveler moves from no intention to a specific named hotel without ever opening a search engine. When they do reach Google Search, they arrive with a name. The decision formed somewhere else.
The scale of this is not marginal. A TripIt survey of 1,000 Americans found that 69% of Gen Z and millennial respondents find their travel inspiration on social media, with 44% specifically naming TikTok. A British Airways Holidays and YouGov survey found that 23% of UK adults use TikTok for holiday inspiration, with 18% of those going on to book. TikTok-commissioned research conducted with Phocuswright found that 42% of users who engage with travel content on the platform have made a booking or travel-related purchase as a result.
The pathway from a three-minute video to a Google Search for a specific hotel name produces exactly the search pattern the data shows: specific-name queries from travelers who arrive informed, with nothing to do with AI.
What AI discovery actually is right now
The AI contribution to specific-name search growth is real, but it requires more precision than it usually receives. When ChatGPT recommends a hotel, it is largely surfacing signals from Expedia and Booking.com inventory — the same commercial stack that has long shaped OTA platforms, now delivered through a conversational interface. It is not independent curation. It is a different front end on familiar commercial infrastructure, inheriting the same distortions: properties with strong OTA presence surface; properties without it don't.
It is, however, better than Google Search for the inspiration question — not because it produces better results, but because it at least attempts to answer an open-ended question with a recommendation rather than a rate table. A traveler asking "where should I stay in Milan for a long weekend, something design-led and quiet" gets a response that engages with the question. Google Search, presented with the same intent as a keyword query, returns ads.
The AI recommendation layer will mature and the data shaping it will shift. But crediting it today with having replaced Google Search at the top of the discovery funnel overstates what the product currently delivers.
Three origins, one pattern
The specific-name search rise reflects at least three distinct upstream origins: creator content driving named-property interest, word-of-mouth operating entirely outside any platform, and AI interfaces doing a version of discovery that mostly recirculates existing commercial signals. Each produces a traveler who arrives at Google Search decided. None required Google Search to have been part of the process.
What connects all three is an information environment that failed the inspiration question. Google Search filled its results with paid signals. AI tools inherited an inventory shaped by OTA commercial logic. The organic content that might have done the discovery work was deprioritized on both surfaces.
Travelers didn't abandon structured search because they found a better version of it. They found inspiration where the commercial layer hadn't yet arrived — in creator content, in recommendation, in conversation. That is the more honest reading of what the data is showing.
by Markus Busch, Editor/Publisher Hospitality.today
Read also: The hotel ranking game has a new playing field
Enjoying this analysis? Hospitality.today delivers daily insights on hotel distribution, AI trends, and travel commerce — straight to your inbox. Subscribe for free at Hospitality.today →