Zero-Click Search: The Numbers Worth Trusting

The fastest way to lose credibility in an AI search conversation is to quote the right number with the wrong caveat.

Zero-click search is real. AI Overviews are reducing clicks. Travelers are asking longer, messier questions inside answer engines. None of that requires exaggeration. The numbers are already sharp enough if we use them carefully.

This is the field guide I would want in front of me before a panel, a board meeting, or a sales call with a destination or hotel team.

- The safest behavior study: Pew

- The stale stat: 34.5 became 58

- The big zero-click number: 68.01

- The travel implication: traffic is not the whole scorecard

- Where Drifter fits

- FAQ

The Safest Behavior Study: Pew

Pew Research Center published one of the cleanest public looks at how users behave when Google's AI summaries appear. The study used March 2025 browsing data from 900 U.S. adults and looked at 68,879 unique Google searches. About 12,593 of those searches produced an AI summary when Pew collected the result pages in April 2025.

The headline finding is simple: users clicked a traditional result on 8 percent of visits when an AI summary appeared, compared with 15 percent when it did not. Users clicked a source inside the AI summary itself on only 1 percent of visits. Pew also found that 18 percent of Google searches in the March 2025 dataset generated an AI summary.

Those numbers are useful because they are based on observed behavior, not a survey about intentions. But they need caveats.

The comparison is descriptive, not causal. AI summaries tend to appear more often for longer, question-shaped, informational searches. Those searches may have different click behavior even without an AI answer. Google also disputed Pew's methodology. Pew's AI summary prevalence number is also a March 2025 snapshot. It is useful historically, but it should not be treated as today's prevalence.

The safe sentence is this: in Pew's March 2025 browsing dataset, Google users were about half as likely to click a traditional result when an AI summary appeared, and almost never clicked the summary's cited sources.

That is enough.

The Stale Stat: 34.5 Became 58

The number many people still quote is 34.5 percent. It came from Ahrefs' April 2025 analysis of AI Overviews and position-one click-through rate.

Ahrefs reran the study with December 2025 data and published the update on February 4, 2026. The new finding: AI Overview presence correlated with a 58 percent lower average click-through rate for the top-ranking page.

The methodology matters. Ahrefs used 300,000 keywords: 150,000 with an AI Overview present and 150,000 informational keywords without one. The team compared aggregated Google Search Console desktop CTR data across December 2023 and December 2025.

This is not a randomized experiment. It is correlational. AI Overview keywords are not interchangeable with all keywords. The analysis is desktop-based. It comes from an SEO vendor with its own product interests, even if the method is transparent.

Still, the direction is hard to ignore, and the February 2026 number is the one to use if you are going to cite Ahrefs. Anyone quoting 34.5 percent without mentioning the update is ten months stale.

The safe sentence is this: Ahrefs' December 2025 desktop analysis found AI Overview presence correlated with a 58 percent lower average CTR for the position-one page, up from its earlier 34.5 percent estimate.

The Big Zero-Click Number: 68.01

SparkToro published the freshest broad zero-click figure on June 8, 2026, using Similarweb panel data for January through April 2026. The headline number: 68.01 percent of U.S. Google searches ended without a click to any website.

SparkToro compares that with a 2024 U.S. zero-click figure of 60.45 percent. Rand Fishkin is careful about the caveat: the comparison uses different panels. The 2024 data came from Datos, while the 2026 data came from Similarweb. That makes the trend direction useful, but the exact delta should not be treated like a clean apples-to-apples lab result.

The other caveat is mobile. SparkToro notes that true zero-click behavior may be even higher because of data exclusions around the Google app, but that is a reason for caution, not for embellishment.

The safe sentence is this: SparkToro and Similarweb estimated that 68.01 percent of U.S. Google searches from January to April 2026 ended without a website click, while noting that historical comparisons use different data panels.

That is a big enough number. It does not need decoration.

The Travel Implication: Traffic Is Not The Whole Scorecard

Travel marketers grew up with a clean funnel: search query, website visit, conversion path. The AI layer breaks that neat sequence.

A traveler can ask Google, ChatGPT, Gemini, or another assistant where to stay, which neighborhood fits a family trip, whether a destination is worth visiting in shoulder season, or which attractions work without a car. The answer can influence the shortlist without creating a website session. In some cases, the source that shapes the answer may never receive the click.

That does not make the website irrelevant. It makes the website more like infrastructure.

The official site still needs to be crawlable, specific, and current because it can shape the answer. It needs to explain the place better than generic listicles do. It needs to give models the facts they need to represent the destination, hotel, venue, or attraction accurately.

But traffic can no longer be the only proof that the source layer is working. If an AI answer recommends your destination but the user books later through direct, paid, OTA, group sales, or brand search, the influence may not show up as organic search traffic.

This is why AI visibility reporting needs different metrics. Answer Share tells you whether you appear in the answer. Authority Share tells you which sources shape the answer. Competitor displacement tells you who gets recommended when you do not. Source gaps tell you what to fix next.

Clicks still matter. They just do not describe the whole market anymore.

Where Drifter Fits

This is the measurement gap Currents is built around. If fewer searches become clean website sessions, travel teams need a scorecard that still shows whether they are being recommended, cited, and framed correctly.

Drifter tracks Answer Share, Authority Share, competitor displacement, and source gaps across the AI systems travelers use. It does not replace traffic analytics. It explains what traffic misses.

That matters in board conversations. Instead of arguing over whether organic sessions fully capture demand, a DMO or hotel team can show where AI answers include them, where competitors are winning, and which source fixes are next.

FAQ

What is zero-click search?

Zero-click search describes searches where the user does not click through to an external website. The user may get the answer on the results page, search again, click another Google property, or end the session.

Are AI Overviews reducing clicks?

Multiple studies point in that direction. Pew observed lower click behavior when AI summaries appeared. Ahrefs found AI Overview presence correlated with a 58 percent lower average CTR for the top-ranking desktop result. Both need caveats, but the direction is consistent.

Is the 34.5 percent AI Overview CTR stat still current?

No. Ahrefs updated its analysis on February 4, 2026 using December 2025 data and reported a 58 percent lower average CTR for position one when an AI Overview was present.

What should travel marketers measure instead of only organic traffic?

Keep measuring traffic, but add Answer Share, Authority Share, competitor displacement, source gaps, and assisted demand. AI answers can influence the shortlist even when they do not send a click.

Does this mean destination websites matter less?

No. They matter differently. The official site may receive fewer clicks for some queries while becoming more important as source material for AI answers. The site has to be readable, specific, and trusted enough to shape the recommendation.