Your SEO Strategy Still Works. The Clicks Are Just Moving Elsewhere.

TL;DR

• AI search (Google AI Overviews, ChatGPT, Gemini, Perplexity) is intercepting top-of-funnel travel queries — your rankings stay, but click-through is collapsing.
• Most OTA marketing dashboards still measure impressions and ranks, missing the click-through drop entirely.
• The fix isn't more SEO. It's optimizing your content to be cited inside AI answers — what's now called Generative Engine Optimization (GEO) and Answer Engine Optimization (AEO).

Quick disclaimer before you dive in: this piece will be interesting, slightly annoying, and oddly comforting — it contains bad news, good news, and a small dose of group therapy for anyone experiencing AI fatigue (or mild AI rage).

Now, back to the regularly scheduled existential dread: what AI search is actually doing to OTA (online travel agency) discovery, and why most distribution teams are tracking the wrong metrics.

For a decade, the OTA organic playbook had one line everyone agreed on: rank for the long-tail informational queries, capture the user before they decide where to book, and funnel them to inventory. "Best beach hotels in Crete." "Where to stay in Lisbon with kids." "Is Booking.com cheaper than the hotel website?" Every one of those queries used to send a real person to a real OTA page. Most of them still rank for the same OTAs they did in 2020.

Almost none of them are sending traffic anymore.

This is not a ranking problem. It's a click problem, and it's mostly invisible if you're still measuring SEO the way you measured it in 2022.

The funnel got shorter and nobody got the memo

Google's AI Overviews now appears on a meaningful share of travel-related queries — particularly the informational, comparative, and "how does X work" questions that used to be the OTA top-of-funnel bread and butter. Studies from Authoritas (2025), BrightEdge (2025), and Ahrefs (2025) have all pointed in the same direction: when an AI Overview appears, click-through rate (CTR) to the top organic result drops sharply, even when the rank holds steady. Estimates vary widely by query type — some put the CTR drop in the 30% range, others closer to 60% — but the direction is consistent across every study and every vertical that's been measured.

Travel is uniquely exposed. The queries AI search handles best are exactly the ones that historically sent OTAs the most cheap, intent-rich traffic:

• "Best time to visit [destination]"
• "Where to stay in [city] for [traveler type]"
• "Is [hotel chain] better than [hotel chain]"
• "How does refundable booking work"
• "Difference between bed-and-breakfast and half-board"

A user asking ChatGPT, Gemini, or Perplexity any of these now gets a synthesized answer with maybe three citations. Sometimes one of those is an OTA. Often none of them are.

And it isn't just Google. Perplexity, ChatGPT search, and Gemini's standalone search experience are pulling search volume out of the open web entirely. Pew Research and Similarweb data through Q1 2026 both show AI-native search adoption rising fastest among exactly the millennial and Gen-Z travel decision-makers OTAs spent fifteen years acquiring through SEO.

The metric that's lying to you

Here's why most distribution teams have missed the alarm: their dashboards still look fine. Rankings are stable. Impressions are stable, sometimes up. The keyword tracker still shows you owning page one for the queries you spent the last decade building toward.

What's collapsing is the gap between impression and click, and most SEO reporting still treats those as roughly correlated. Well, they aren't anymore.

If your weekly SEO review is still focused on "we hold position 2 for 'best hotels in Tuscany'" — congratulations, you're tracking a vanity metric. The relevant question is: how many people who saw that result actually clicked it, and how has that number moved over the last six months?

Across the OTAs and bedbanks (wholesale hotel inventory aggregators) we work with at Gimmonix, the pattern repeats: rankings hold, impressions hold, clicks quietly fall quarter over quarter. Most teams aren't looking for it. The dashboards they're looking at were built in 2019.

What's still working

Hey, we know it might feel like the world is on fire, but not all of it…yet.

The bottom of the funnel is in fact healthier than it's been in years.

Transactional, high-intent searches like "Madrid hotels May 12-15," "[hotel name] book," "cheap flights LHR JFK" still send users to conversion-ready pages. AI search hasn't replaced these because there's nothing to summarize: the user already knows what they want, they just want the booking interface.

Branded search is holding up too. "Booking.com Lisbon," "Expedia weekend deals," "TripAdvisor Tokyo reviews" — these queries assume the destination, and AI doesn't intercept them.

What's eroding is the layer above: the discovery, comparison, and "tell me about" queries that used to feed the top of the funnel. That's the layer most OTAs, bedbanks, and TMCs (travel management companies serving corporate clients) built their organic acquisition strategy on top of.

What distribution teams should actually do

Three things, in order.

One: audit clicks, not ranks. Pull six months of Google Search Console data and look at impressions versus clicks for your top 100 informational queries. If impressions are flat or up and clicks are down meaningfully, your AI search exposure is real. Most teams haven't run this audit. It takes about an hour.

Two: optimize for being cited, not for ranking first. AI search engines pull from the web, but they don't pull from everywhere. They pull from sources with structured data, clear answers to specific questions, and recognizable authority. This is what generative engine optimization (GEO) and answer engine optimization (AEO) are about — and they aren't a rebrand of SEO. The unit of work is different. You're writing for a machine that's writing for a human, not for a human directly.

Three: stop waiting for the traffic to come back. It isn't. The shift in how people start travel research is structural, not temporary. The teams that recognize this in 2026 will rebuild their content strategy around getting cited inside AI answers. The teams that wait will spend another year defending impressions that don't convert.

The honest version

SEO isn't dead. It changed jobs.

For fifteen years, organic search was the cheapest, highest-intent traffic source in OTA marketing. That window is closing — not because search is going away, but because the part of the search journey that used to belong to your blog post or category page now belongs to a model that summarizes ten of those pages and ships the answer back to the user.

The OTAs, bedbanks, and TMCs that figure this out fastest will look like they got lucky with traffic in 2027. They didn't get lucky. They got their measurement right early.

Related reading: Agentic AI Is Coming to Cut OTAs Out. Or Is It?