A small case study of one website suggests that clicks from Google’s AI Overviews and AI Mode run at only about 0.9%, a modest rate that still carries a surprising sting. When the site opted out of appearing in those AI features, its AI-driven impressions fell to zero, as expected, but its regular organic clicks fell as well. The result, published on October 30, 2026, hints that visibility inside AI answers may support clicks elsewhere on the same results page.
How the experiment worked
The study used Google’s own controls for opting out of Search generative AI features. By switching those controls on for one site, the owner blocked it from AI Overviews and AI Mode, while leaving the rest of the page untouched. That created a before-and-after comparison: with AI visibility, without it, and then with it again.
The comparison is narrower than it first sounds, and the distinction matters. AI Overviews kept appearing on the results pages for the relevant queries throughout the test. What changed was only whether this one site showed up inside them, so the study measures the value of a site’s own AI citation, not the effect of AI Overviews existing at all.
The headline number, and its limits
Across the periods when the site was visible in AI features, its AI impressions carried an estimated click-through rate of about 0.9%. That figure was not read directly from a single report. It was estimated by comparing overall clicks and impressions before and after the opt-out, and subtracting the known AI impressions from the period when the site was excluded.
AI impressions made up about 8% of the site’s total impressions, so the AI slice was small beside its regular search presence. Overall impressions also dropped during the opt-out period, which fits a site losing one more place to appear. The study’s author described the work as a small case study, resting on a limited sample from a single site.
That reach is what makes even a small AI click rate worth watching. AI Overviews now appear on more than 80% of searches tracked by independent measurement firms, so a citation inside those answers is increasingly part of how readers first meet a brand. A low rate on a large number of impressions can still add up to meaningful traffic.
The surprise: organic clicks fell too
The most unexpected finding was not the 0.9% figure itself, but what happened to the site’s ordinary results. When AI visibility was switched off, regular organic clicks dropped as well, even though those listings had not been removed or demoted. When AI visibility returned, the affected pages recovered, which points to a connection rather than a coincidence in timing.
The author called this a “Search SERP Halo Effect”: the more places a brand appears on a results page, the more likely a reader is to click one of them. An AI citation does not only send its own clicks; it keeps the brand in front of the reader, which may make a later click on a listing more likely.
Other search analysts echoed that interpretation, noting that appearing in AI Overviews may build brand familiarity that pays off in later searches and clicks. That effect would not show up consistently for every site or query, but it offers a plausible reason why removing one surface could drag down another.
Why the caveats matter
A single-site test cannot settle how AI answers affect search traffic in general, and several details argue for caution. Sitewide average position actually improved during the test, helped by new posts and brand searches, yet the pages with the most AI visibility fell and then rose again. That split suggests the effect is concentrated where AI visibility is highest.
The estimation method adds another reason to treat 0.9% as an approximation, not a benchmark. Because the rate was derived from before-and-after totals rather than measured directly, small changes in traffic, rankings, or seasonality during the test could shift it. Until other owners run the same test, there is no way to know whether this site’s experience is typical.
Why It Matters
For publishers and search marketers, the study reframes the opt-out question. Refusing to appear in AI Overviews and AI Mode might protect a site from having its content summarized, but this test suggests it could also cost clicks it would otherwise have kept. That trade-off deserves a test on each site’s data, rather than a decision made on principle alone.
The finding also lands at a moment when publishers are already reacting to AI search draining referrals, as shown by Reddit’s move toward a logged-in experience after its search traffic fell sharply. If AI visibility quietly supports organic clicks, losing it could hurt twice: through the missing AI clicks, and again through weaker performance in the classic results below.
For readers, the practical message is simpler. The presence of a familiar brand inside an AI answer may shape which ordinary result they trust and click next. Search visibility is becoming a combined effect across the whole page, not a contest between one listing and another, and measurement that looks at only one surface may miss that interaction.
Background
Questions about click-through rates from AI search have run ahead of hard data for more than a year, because Google does not publish a separate click rate for AI Overviews or AI Mode. Small, shared tests like this one are valuable even when they cannot be generalized, and the author has urged other site owners to try the same experiment.
Source: Search Engine Roundtable’s report.
0 Comments