The hotel discovery layer: Metasearch, eaten by what it invented
The original upstream comparison surface is being absorbed into the AI surfaces that learned from it
A traveler types "best boutique hotels Lisbon" into Google. The result isn't a list of OTAs anymore. It is an AI Overview that names properties, summarizes them, and links to hotel websites or Google Business Profile listings. Kayak doesn't appear. Trivago doesn't appear. Tripadvisor doesn't appear in the answer.
Metasearch invented this exact moment — upstream comparison before the booking channel — twenty years ago. It is being squeezed out of it by the surfaces that learned from it.
What's compressing metasearch isn't competition from a new entrant. It is the surfaces metasearch built. Two decades of upstream comparison taught the wider web — Google in particular — how to do this work. The wider web has now learned.
What metasearch was
The metasearch platforms are old. Kayak launched in 2004. Trivago in 2005. Google Hotel Ads in 2010. Tripadvisor pivoted into metasearch over the same period. Their proposition was simple and structurally novel — pull rates from many channels, show them side by side, let the traveler pick where to book.
This was upstream comparison before the booking channel. It was the discovery layer's first version, even if no one called it that.
Hotels learned to manage metasearch as a discrete distribution channel with its own commercial discipline. Partner mix, bid strategy, attribution windows, free booking links, commission models, parity exposure across the metasearch entry points — these were a craft, taught by agencies, optimized by tools, reported in distribution dashboards.
For fifteen years, metasearch was where the comparison work that mattered most happened. That is no longer true.
What's compressing it
Three structural shifts have moved upstream comparison off standalone metasearch.
The first is Google's absorption of the function. Google Hotel Ads became the dominant metasearch surface, then changed the rules. Commission-based bidding — the model that made metasearch accessible to smaller properties — was eliminated under DMA pressure in 2024 and 2025. The remaining models, CPC and tROAS, favor larger advertisers. Google didn't displace metasearch; it became the metasearch surface and reshaped the economics from inside.
The second is AI-driven displacement. The hotelrank.ai study of 4,000 Google AI Mode hotel queries found that 79% of hotel clicks stay inside Google's own ecosystem, primarily through Google Business Profile. OTAs receive 3.6%. Standalone metasearch isn't a meaningful destination in that flow at all. The comparison happens; the click goes elsewhere.
The third is the financial evidence. Booking Holdings took a $457 million writedown on Kayak in October 2025, with CEO Steve Hafner attributing it directly to Google's AI Overviews squeezing the category. Trivago and Tripadvisor have lagged pandemic recovery for years. The standalone metasearch business is structurally compressed, and the public financial reporting reflects it.
The asymmetry
With standalone metasearch, hotels had a discrete commercial relationship. Partner contracts existed. Bid auctions had published rules. Performance dashboards showed attribution. Hotels could see what they were paying for, what they were getting, and where the leverage points were.
That infrastructure is shrinking.
The function metasearch performed has not disappeared; it has migrated. AI Overviews don't auction visibility. Google Business Profile isn't a metasearch product hotels can manage as a distinct line item. AI summaries that name properties don't have a bid mechanism. The function got absorbed; the commercial controls did not come with it.
Hotels have spent fifteen years learning how to do metasearch well — bid management, attribution analysis, partner relationships. The skills still work. The surface they work on is getting smaller.
What this means
For two decades, metasearch was a discrete strategic question. Which partners. What bid strategy. What share of the distribution budget. The line item sat alongside OTA commissions and brand.com investment as one of several distribution decisions a commercial team made each year.
By 2026, that question is changing shape. The discrete decision is being absorbed into a larger one — how much of the discovery layer is now controlled by a single intermediary, and what that means for hotels that previously had multiple competing surfaces to choose between.
Standalone metasearch may persist for years. Trivago is still profitable. Tripadvisor still has scale. Kayak still has brand strength. But the structural ground they sit on is being absorbed, and the financial reporting already shows it.
Metasearch is no longer a separate question about partner mix. It is a question about how much of the discovery layer is now Google's.
Next in this series: Agentic discovery and the next intermediary
by Markus Busch, Editor/Publisher Hospitality.today
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 →