AI vs the Click: Travel Search in 2026, By the Verified Numbers

How people find travel changed faster in the last eighteen months than in the previous fifteen years — and for once, the change is measurable. This briefing collects the verified numbers: a million-keyword study, Pew’s behavioural tracking of 68,879 real searches, Google’s own AI Mode disclosures, and Alphabet’s SEC filings. Every figure below carries its source. The short version: most searches now end without a click, travel is the vertical AI is reshaping fastest without shrinking — and nearly one in five consumers has already bought something on an AI’s say-so without checking anywhere else.

The Click Is Dying — The Numbers

The behavioural evidence is unusually strong because Pew didn’t survey anyone — it tracked 68,879 actual searches. When an AI Overview appears, organic clicking nearly halves, and only about 1% of users click a source link inside the Overview itself. BrightEdge’s twelve-month analysis frames the squeeze from the publisher side: search impressions up 49% since AI Overviews launched, clicks down roughly 30% — more people seeing results, fewer arriving anywhere. The scale is not niche: Google reports AI Overviews reaching two billion monthly users across more than 200 countries, and ChatGPT passed 700 million weekly users in the same period. And the commercial engine hasn’t flinched — per Alphabet’s Q4 2025 SEC filing, Google Search took $224.53 billion in FY2025 revenue, with Q4 search advertising up 17% year on year. The clicks are leaving; the money, so far, is not.

Demand Moved; It Didn’t Shrink

The largest public test of the "AI is killing search" thesis arrived on 2 July 2026, when Fractl and Search Engine Land published an analysis of 1,010,848 high-volume keywords (10,000+ monthly searches each) across 379 brands in eight verticals — 35.4 billion monthly searches in aggregate — alongside a survey of 1,004 US consumers. Gartner had predicted in 2024 that traditional search volume would fall 25% by 2026. The study measured the real figure at 29% of high-volume demand in decline — worse than forecast, but with a twist Gartner missed entirely:

Where it moved matters most for anyone reading this site: the study found non-branded queries — 90% of all tracked volume — are the most vulnerable, because a question with no brand in it can be answered entirely inside a chat window. And the consumer survey landed the adoption paradox in one pair of numbers: 70% of Americans say they use AI more than a year ago, but only 17% say they use search engines less. People are adding AI, not substituting it.

Why Travel Is the Exception

Across the eight verticals, the Fractl/SEL study found a clean dividing line: categories where a chatbot can deliver the complete answer (health information, financial definitions) are losing search demand hardest, while categories where people must ultimately compare prices and transact — explicitly including travel — are growing or holding near flat. Travel is simultaneously the most AI-assisted category and one of the most resilient, because an itinerary suggestion still ends in a booking somewhere. The strain shows on the publisher side, not the demand side: in Fractl’s earlier 2025 research, 39% of marketers reported traffic drops after AI Overviews arrived, with travel and hospitality among the hardest hit at 43%.

Google’s own AI Mode disclosures complete the picture. The company reports AI Mode has passed one billion monthly active users globally, with queries more than doubling every quarter since launch — and the average AI Mode query running roughly three times the length of a traditional search. People aren’t typing keywords at it; they’re describing trips. Google names travel among the top categories, with itinerary and vacation planning among the leading uses of its interactive Canvas feature. The behaviour we documented from our own logs in What Travellers Ask AI — machine-decomposed research briefs on operators, compensation, and status — is the retrieval layer under exactly this planning boom.

The Trust Gap and the 18%

Trust hasn’t transferred as fast as usage: 46% of surveyed consumers still trust traditional search more, 33% trust both equally, and 20% trust AI more — with 56% at least somewhat sceptical of AI product recommendations. Asked what would send them back to search, the top answer was AI giving unreliable answers (35%). For purchase research specifically, search engines and retailer sites tie at 47% each as the starting point; only 13% start with a chatbot, and shoppers still consult around three sources before buying. The five-year outlook is evolution, not exodus: 52% expect Google to remain their primary tool, while 20% expect to leave — and the reasons people prefer AI are prosaic and durable: better summaries (21%), faster answers (20%), conversational follow-ups (19%).

For a traveller, the practical read is the one the sceptical majority already applies: let AI shortlist, then verify the thing that costs real money against a primary source — the operator’s own terms, a policy document, a live price. Comparing actual coverage wording is exactly the step the 18% skip; standalone policies like SafetyWing publish theirs in plain language, which is what verification is supposed to look like. The same applies to experiences: an AI can suggest the tour, but availability and reviews live where the booking does — GetYourGuide shows both before you commit.

What Our Own Logs Confirm

This publication is a small, clean test case for every number above, because it was built in the AI era: roughly 100 days old, with Google contributing almost nothing yet. In that window our pages were cited about 31,000 times by AI assistants — documented query by query in our grounding-query report — while Google organic delivered clicks in the low hundreds. The Pew and Semrush zero-click figures aren’t abstractions from where we sit; they’re the operating environment. And the Fractl redistribution finding matches what the machine layer asks for: not the shrinking generic keywords, but dense, specific, comparative briefs — the demand that moved. Our Travel Demand Index tracks where it moved to, city by city and category by category.

Frequently Asked Questions