Are Metasearch Engines Dead, or Have They Simply Become Invisible?

This is a really thought-provoking take. I agree that metasearch isn't dead. it's evolving, quietly powering AI-driven interfaces rather than existing as a visible marketplace. The first to benefit will be those who can move fast and think at scale likely OTAs while traditional booking engines risk becoming just another costly layer.

For hospitality leaders, the challenge remains to rethink how we engage guests, embrace the invisible infrastructure, and adapt before this post-metasearch world becomes the new normal.

Rather than disappearance, we're witnessing transfiguration: the logic of metasearch is alive, but its form is shifting from a visible comparison table to an invisible, integrated intelligence layer.

Simone's argument resonates strongly with me. If metasearch has dissolved into the invisible infrastructure of distribution, then the next challenge is redefining where real value lies.

With conversational AI agents now capable of pre-filtering, recommending, and contextualizing travel options, simply consolidating data from various sources is no longer enough. These systems don't just compare rates — they interpret intent, summarize across a broader pool of travel and accommodation products, and deliver personalized recommendations in a natural, human-like way.

That means platforms will need to move beyond aggregation and bookability. The new competitive edge is about who activates more or better proprietary and first-party data — the kind of data that can enrich these AI-driven conversations with unique context, insight, and differentiation.

It"s an exciting shift: as AI agents become the new interface, the question is no longer how visible your distribution channel is, but how indispensable your data becomes in shaping the conversation itself. 

How can this be achieved? Find out more about Dynamic Inventory: https://www.gauvendi.com/post/to-be-found-by-machines-you-must-sell-like-a-human.

Metasearch is not dead, but has become a niche channel, a hybrid advertising/distribution channel.

Hotel metasearch has existed for over 22 years now (SideStep, acquired by Kayak). Today, Google Hotel Ads (GHA) has bigger market share than all metasearch platforms combined.

So, how big is hotel metasearch as an advertising/distribution channel? Less than 5% of hotel roomnights are booked on or referred by metasearch platforms. The same percentage was 10 years ago, 5 years ago and today.

In my view, with the ascendancy of AI search and agentic AI, this percentage will shrink further. Recently, Booking Holdings reported a $457 million accounting writedown for its Kayak brand due to increased customer acquisition costs, caused by the rise of AI search, requiring more spending on paid Google ads.

So, what should hoteliers do?

Hotel price is one of the four criteria travelers use to find the property for their hotel stay; location, customer reviews and hotel information being the other three. By participating in Google HPA. it's free and paid booking links, the property provides an answer to all four criteria in real time, since Google serves pricing information in its SERPs, Google Maps, Hotel Business Listing, etc.

I, too, have not been a believer in MetaSearch for quite some time dating back to early Covid (see here). Most hotels, especially as we go into 2026, should be focused on stealing market share, not channel shift of business. MetaSearch only drives 3.8% of total website visits (Q3 2025), ad that has been decreasing each quarter, down from 5.3% Q4 2024. As agentic AI plugs into LLMs, the comparison functionality of MetaSearch will continue to diminish.

If you are looking for the horse, it's no longer in the stable.

As an industry, we must face the reality that Round One of the AI-Meta War is over: the AI Browser has already absorbed search and booking. When Google Gemini directly leverages the entire Google Travel ecosystem, guess who is left on the outside of these foundational partnerships? The hospitality operator.

If a user's personal AI Browser bypasses the need to click deeper—be it Meta, OTA, or direct-to-provider—then the distribution channel has effectively disappeared into the "mether". We lose control of the customer, the commercial terms, and our Competitive Advantage.

The war for Round One is lost to the scaled AI tech companies. Our only hope for Round Two is to stop fighting the interface and become the indispensable data source for it.

We must transition from fragmented inventory to a cohesive, attribute-rich digital product, amplified by technology. The winning strategy is to change the playing field, making ourselves an unavoidable, high-value partner by delivering superior strategic Execution. The war is not lost if we redefine the battle as a fight for control over the underlying data infrastructure.

Gentlemen, choose your weapons!

Metasearch, in its traditional form, may be fading, with several metasearch travel sites showing slower traffic. Instead of declaring its "death," I agree with Simone that we are witnessing a quiet evolution. Today, metasearch is shifting from a simple comparison marketplace to a background intelligence layer embedded in AI-driven travel ecosystems.

Aggregation still matters, and it continues to happen invisibly, moving even deeper underground. Intelligent agents and large language models now aggregate, interpret, and rank data behind the scenes, shaping travel options without direct consumer awareness. Since most consumers never fully understand metasearch to begin with, they are unlikely to notice how it continues to influence their travel decisions today.  

For hotels, this shift presents both risk and opportunity, as other market transformations do. Without visibility and authentic brand storytelling, bidding strategies will have less impact. What matters more now is data quality, integration, and trust signals that make a property "machine-favored." The new ecosystems will reward data transparency, authenticity, and connectivity.

I have long argued that metasearch has never been a business all on its own. Metasearch has always been a feature. And given that hotels get most of their metasearch volume from Google and TripAdvisor (as opposed to from pure-play "metasearch companies"), the data backs my argument up. 

Those metasearch bookings also have long carried an additional toll, albeit less expensive than traditional paid search. The greatest trick Google ever played was swapping hoteliers' free organic results for paid metasearch clicks… and because those paid clicks cost less than OTAs, making us feel like we'd won. 

That said, I don't think metasearch will go away anytime soon. If anything, metasearch will remain a useful feature within AI tools, just as it has become a useful feature within Google's search results. AI answer engines will sometimes display metasearch results when travelers ask questions like, "Show me hotel options in [location]." AI assistants will be sophisticated enough to present options that match guests' known and inferred preferences and earning a fee for those clicks will help those platforms monetize guest activity. 

In short, metasearch isn't dead. It's simply migrating into new customer experiences. It's too useful and profitable a feature to do anything less.

Metasearch is not disappearing; it is dissolving into the data layer that fuels the new conversational distribution model. And while OTAs are clearly positioned to capitalize first, I am not convinced they are destined to dominate.

Human travelers still crave variety, uniqueness, and authenticity. That demand will influence how LLMs evaluate sources and surface answers. Because of that, the field remains wide open for booking engines, channel managers, and metasearch providers that can deliver clean, structured, LLM-ready repositories of rates, inventory, and truly differentiating content.

In this next phase, success will not hinge on visibility in a marketplace; it will hinge on who can become the most trustworthy and comprehensive data source for the systems shaping traveler intent.

The perspective presented in the write-up is both insightful and timely, especially given the rapid evolution of user interfaces and the adoption of large language models (LLMs) in information searches. As pointed out, "metasearch"(once a disruptor) has not vanished, but it has become less visible as aggregation shifts behind the scenes. In this new landscape, conversational AI serves up answers sourced from aggregated data, effectively embedding metasearch capabilities into the foundational fabric of digital distribution.

This shift fundamentally changes how consumers interact with travel and commerce platforms. Rather than manually comparing options, users now receive curated responses in natural language, obscuring the mechanics of aggregation. The analogy of absorption versus erasure is apt: the value of aggregation remains, but its form has changed.

Are OTAs, with their robust platforms and integration capabilities, the best positioned to capitalize on this shift? While others (like brand booking engines) may find themselves subject to renewed intermediation, despite appearances. To prevent this, hotels must own the guest experience and adopt and experiment in AI technologies while also collaborating and not competing with tech enablers.

Ultimately, the question posed is thought-provoking. Metasearch is not dead, but evolving, its "ghost" now animates the invisible engines supporting new AI-driven paradigms, prompting us to consider whether this is progress or simply a wolf in new sheep’s clothing.

If AI collapses comparison into a single recommended answer, the only part of metasearch that matters to hotels disappears. Metasearch worked because a hotel's direct rate appeared next to an OTA's rate. That visibility gave hotels a real chance to compete. If that comparison goes away, the advantage shifts to whoever has the most LLM-ready data. Today, OTAs have a significant edge. In an AI-driven interface, they become the default source of truth.

The deeper risk isn't that metasearch disappears but rather that hotels fail to enter the consideration set at all. If the model views the OTA version of the hotel as more complete or more trustworthy, that becomes the recommended answer.

The better question is how hotels influence the signal behind that recommendation before AI ever looks at rates. The strategic path forward is shaping traveler intent early, by feeding models authoritative content and trusted signals that guide what the system sees as best, long before price enters the picture.

Controlling the upstream demand signals AI relies on will determine whether hotels or OTAs win. Hotels succeed by showing up in the sources AI trusts most, not by fighting for a metasearch placement that no longer exists.

Metasearch isn't dead—it's evolving. The familiar marketplace of bids, clicks, and placements is fading, but its essence remains. Rate comparison and inventory aggregation haven't disappeared; they've simply moved behind the scenes, powering AI-driven interfaces that deliver fast, personalized answers. 

Google's work on agentic AI for travel booking underscores this shift. While Google Hotels and other metasearch models still exist, the future may look less like a visible comparison site and more like an intelligent assistant making choices for you in real time. What was once a consumer-facing experience is now embedded in conversational tools like chatbots, smart travel planners, and platforms such as ChatGPT, Perplexity, or Gemini. 

This isn't just a technical change—it's a turning point for the industry. Brands that adapt quickly and integrate will lead the way, while those clinging to traditional models risk higher distribution costs and missed opportunities. The right partnerships—like connecting through Amadeus—help hoteliers stay competitive without adding extra fees, ensuring bookings and searches flow efficiently through trusted systems. For hoteliers, the challenge is clear: embrace this transformation to remain discoverable in a post-metasearch world, where visibility and success depend on agility and a willingness to rethink distribution. 

If metasearch looks like it is dying, it is only because it is melting into something wider and harder to see. The familiar theatre of grids, bids, colours, and frantic CPC gymnastics is disappearing from the surface, not because the mechanism is gone, but because the interface that once exposed it has become unnecessary. Aggregation still happens, only now it lives beneath conversational layers that are quickly becoming the main gateways to travel discovery.

What Google is doing with AI Mode is simply another step along a path that was already clear years ago. When I wrote my book "We Are the Glitch," I described the drift toward a world where AI becomes the primary UI. When that happens, I argued, metasearch stops being a destination and turns into a silent layer that feeds the model the ARI it will narrate back to the user.

This is why the question "are metasearch dead?" misses the point. What is fading is their visible shell. What persists is the underlying logic, now absorbed into the generative web, where the AI becomes the interface, and the web becomes an API driven substrate.