Get a free AI visibility report
AI Overviews now appear on more than one in five Google searches. Click-through rate drops by nearly 60% when they do. This roundup pulls together verified 2026 research on AI Overviews coverage, click loss, and citation patterns. It breaks down what the numbers mean for how a brand should measure AI search visibility.
Three different data problems sit inside "AI Overviews statistics." Coverage measures how often an overview shows up at all. Click loss measures what happens to organic CTR once it does. Citations measure which pages, and which domains, the overview actually pulls from. A brand can win one of these and lose the other two. The numbers only mean something read together, not as a single headline percentage.
Most AI Overview citations point to a third-party domain rather than a given brand's own site, and even a page that does get cited earns a fraction of the clicks a plain organic result would. The visible, attributable click is no longer where most AI-search impact shows up. It shows up later, as a branded search or a direct visit that no UTM parameter ever ties back to the AI Overview that triggered it.
Google's own material is the clearest confirmation of scale. Following an expansion announced at Google I/O, AI Overviews are now available in more than 200 countries and territories and in more than 40 languages. That is a statement about reach, not about how often the feature actually renders on a given search. The two get conflated constantly in AI Overviews statistics roundups.
On the "how often" question, the most reliable figure comes from SparkToro's Similarweb-based research, covered by Search Engine Land. It found AI Overviews now appear on more than 20% of Google searches in the US. That figure is a market average. It says nothing about how often overviews trigger on the specific prompts a given brand cares about. Track your own AI Overviews coverage with a dedicated rank tracker instead of relying on an aggregate number.
Aggregate coverage numbers hide real variance by intent. Seer Interactive's analysis of 49,353 queries, reported by Search Engine Land, found comparison-style queries ("X vs. Y") trigger an AI Overview about 95% of the time. Question-format queries trigger one about 86% of the time. Informational queries trigger overviews roughly 36% of the time. Transactional queries trigger one only about 5% of the time. A brand selling in a transactional, purchase-ready query set will see far less AI Overview exposure. A brand whose prompts skew toward comparisons and open questions will see far more. Content quality has little to do with that gap.
A low AI Overviews coverage rate on transactional prompts is not a technical failure. It reflects how rarely Google renders an overview for that intent at all. The more useful diagnostic is coverage within intent, not coverage overall.
The AI Overview click-through rate story has two layers: the floor it hit, and the partial recovery since. Seer Interactive tracked 53 brands, 5.47 million queries, and 2.43 billion impressions from January 2025 through February 2026. CTR on AI Overview queries bottomed at 1.3% in December 2025, then rose to 2.4% by February 2026. That's an 85% increase in two months, per Search Engine Land's coverage of the study. It's a real rebound, but it still sits well below the roughly 3.3% CTR on searches where no AI Overview appears at all.
The same data breaks the picture down further. When an AI Overview appears and cites a given page, that page earns roughly 2.1% CTR. An uncited page on that same results page earns roughly 0.9%. Both numbers trail the no-AI-Overview baseline of about 3.3%. Being cited clearly beats being ignored. Neither outcome matches what a top organic result earned before AI Overviews existed.
CTR on searches with no AI Overview rose from about 2.8% in early 2025 to 3.8% by February 2026, per the same tracking. Ranking well on queries where no summary renders is becoming more valuable, not less, since those are increasingly the searches where a click still happens. Paid results held up better than organic over the same window. When an AI Overview appeared, paid CTR rose from 14.6% to 16.2%. On non-AIO searches, paid CTR actually fell, from 26% to 21.8%.
The CTR recovery has to be read against a larger trend moving the other way. Zero-click searches, meaning a Google search that ends without any click at all, reached 68% of US queries in early 2026. That's up from 60.45% two years earlier, per the same SparkToro research cited above. When an AI Overview appears, click-through rate drops by close to 60% relative to a search where no overview renders.
Raw click count is the wrong single number to anchor a measurement program on. Hat Club's AI search program illustrates why. Roughly one in fifty visitors arriving through AI referral traffic drove a 20x increase in AI-driven sales. A small, disproportionately valuable slice of traffic can carry far more revenue weight than its share of total sessions suggests. A pure CTR number misses that signal entirely.
Citation data has moved further from organic rankings than most SEO teams expect. An Ahrefs analysis of 863,000 keywords and 4 million AI Overview URLs looked at this directly. Reported by Search Engine Journal, it found only 38% of cited pages also ranked in Google's organic top 10 for the same query. That's down sharply from 76% in a July 2025 version of the same study. The rest split almost evenly between pages ranking 11 to 100 (about 31%) and pages ranking beyond position 100 (about 31%). A separate BrightEdge analysis, using a different methodology and dataset, put the top 10 overlap closer to 17%.
Search Engine Journal's coverage flags an important caveat. Some of that 76%-to-38% drop reflects improved citation detection in Ahrefs' own tooling, not purely a change in how Google selects sources. The two study waves aren't perfectly comparable. What both waves agree on is direction.
Google's query fan-out process likely explains part of the shift. A single search gets split into related sub-queries before the overview is generated. A page can get cited for relevance to one of those sub-queries even if it never ranked for the original exact-match term. Tracking which competitors get cited across that broader prompt set matters more now than tracking rank for one exact-match keyword.
The clearest single trend in the citation data is video's growing weight. Among AI Overview citations that didn't rank in Google's top 100 for the same keyword, 18.2% were YouTube URLs. YouTube accounted for 5.6% of all AI Overview citations in the Ahrefs dataset. Ahrefs' broader Brand Radar tracking found YouTube is now the single most-cited domain in AI Overviews overall. That citation share has grown 34% in six months, per the same Search Engine Journal coverage. A separate study of German health queries found YouTube outranking official medical sources in citations for that vertical. This is a format shift, not a fluke in one dataset.
Put the three data sets together and one conclusion holds. Coverage is rising, so a brand's queries are more likely to trigger an overview than a year ago. Click-through rate on those queries is recovering slightly but still trails a plain organic result. Citations increasingly come from pages that never ranked in the traditional top 10 at all. None of that adds up to a clean, attributable click a standard analytics setup can track back to AI search.
This is the gap Cognizo's Answer Engine Insights module is built to close. Visibility Score, the percentage of tracked prompts where a brand is mentioned, works as the AI-search equivalent of impressions rather than clicks. Citation share splits into owned citations, where the AI links directly to a brand's domain, and earned citations, where it links to a third party covering the brand instead. The data above shows earned citations dominate in practice. Source mention rate and sentiment fill in the rest: which third-party domains an AI platform already trusts on a topic, and whether the brand's mentions there read positively, negatively, or neutrally.
Most AI Overview citations go to earned, third-party domains rather than a brand's own site, and even a citation converts to a click at a fraction of the rate a top-10 organic listing once did. A brand's real AI-search impact increasingly shows up downstream instead: a user sees a mention, doesn't click, and later runs a branded search or types the domain directly. Standard UTM tracking misses that path entirely, since there's no referral parameter to attribute it to. A "how did you hear about us" field with an explicit AI option, paired with citation-share tracking over time, closes more of that gap than any click-based dashboard can alone.
Every figure in this roundup is a market average pulled from a specific sample: Seer's 53 tracked brands, Ahrefs' 863,000 keywords, SparkToro's Similarweb panel. None of them tell a specific brand what its own AI Overviews coverage, CTR, or citation share looks like on the prompts that matter to its category. That number only comes from continuous monitoring of a brand's own tracked prompt set, not a quarterly industry benchmark.
Continuous tracking matters here in a way a monthly snapshot cannot replicate. The citation data above shifted from 76% top-10 overlap to 38% inside roughly eight months. CTR moved from a 1.3% floor to a 2.4% recovery inside two months. A measurement cadence that only checks in occasionally will miss exactly the kind of swing that should trigger a content or PR response. Tracking brand mentions as they accrue, rather than sampling them periodically, is what actually catches a citation shift before it shows up in a quarter's traffic numbers.
There's also a measurement-integrity reason to track the rendered answer directly instead of reconstructing it from an API response. AI Overviews vary by account, region, and query phrasing in ways an API sample can miss. That's part of why Ahrefs, BrightEdge, and other trackers report different top-10 overlap and coverage numbers from similar-looking methodologies. Capturing the actual rendered page a real user sees, the approach Cognizo uses across all ten platforms it monitors, closes that gap more reliably than sampling an API alone.
Coverage is not plateauing. Click-through rate on AI Overview queries is recovering but still structurally lower than a plain organic result. Citation composition keeps drifting further from the organic top 10, with YouTube absorbing a growing share of the overflow. None of the three trends is likely to reverse on its own. A brand that only watches organic rank, or only watches one AI referral traffic number, is reading a partial picture of its own visibility.
The teams that adapt fastest stop treating this as a one-time audit. They run it as a standing, always-on program instead, where Autopilot handles the research, prompt planning, and content updates the shifting citation data calls for. Waiting for the next industry report to notice a trend means the trend already moved weeks ago.
The clearest 2026 data points in three directions at once. Coverage keeps climbing, with AI Overviews now appearing on more than one in five US Google searches. Click-through rate on those queries is recovering from a late-2025 low but still trails a plain organic result. Citations are drifting away from the organic top 10: only 38% of cited pages still rank there in the most recent Ahrefs data, down from 76% a year earlier. Read together, the pattern shows AI Overviews rewarding a wider set of pages than traditional rankings did, while sending each of those pages fewer clicks than it would have earned before.
Aggregate industry figures matter less than a brand's own numbers on its own tracked prompt set. The figures worth watching continuously are AI Overviews coverage rate by query intent and citation share split between owned and earned mentions. Add source mention rate for the third-party domains an AI platform already trusts on the topic, and sentiment on any mention that appears. CTR is useful context, but it shouldn't be the primary KPI. Even a cited page earns a fraction of a pre-AI-Overview click rate. Visibility Score, the percentage of tracked prompts where the brand appears at all, is the closer analogue to an impression metric.
The figures here come from three primary research efforts. Seer Interactive tracked 53 brands, 5.47 million queries, and 2.43 billion impressions from January 2025 through February 2026 to measure click-through rate. SparkToro used Similarweb desktop and mobile clickstream panel data covering US Google searches from January through April 2026 to measure zero-click rate and AI Overview prevalence. Ahrefs analyzed 863,000 keywords and 4 million AI Overview URLs to measure how often cited pages also rank in the organic top 10. Each study uses a different sample and methodology. That's why exact percentages vary slightly between sources even when the underlying trend agrees.
Click-through rate dropped because AI Overviews answer many queries directly on the results page. That removes the reason to click through for users who just wanted a quick answer. Seer Interactive's tracking found CTR on AI Overview queries bottomed at 1.3% in December 2025, then climbed to 2.4% by February 2026, an 85% increase in two months. That is a genuine rebound. It remains well below the roughly 3.3% CTR on searches where no AI Overview appears at all, so the underlying click loss hasn't reversed. It has only partially stabilized.
Google's AI Overviews increasingly rely on query fan-out, where a single search gets split into related sub-queries before the overview is generated. A page can be cited for relevance to one of those sub-queries even if it never ranked for the original exact-match term. Ahrefs' most recent analysis found only 38% of cited pages ranked in the organic top 10 for the same query, down from 76% about a year earlier. Part of that drop reflects improved citation detection in Ahrefs' own tooling rather than a pure behavioral change. Either way, ranking well for one keyword no longer guarantees citation for the topic around it.
A zero-click search is any Google search that ends without the user clicking anything at all. That includes organic results, paid ads, or Google-owned properties like Maps and YouTube. An AI Overview appearing is one specific cause of a zero-click outcome, not the only one. Featured snippets, knowledge panels, and People Also Ask boxes have driven zero-click behavior for years before AI Overviews existed. Research found zero-click searches reached 68% of US queries in early 2026. AI Overviews likely contribute to that rise, though the underlying study doesn't isolate how much of it is attributable to AI Overviews specifically, versus other zero-click features.
Not necessarily. Treating click count as a revenue proxy is one of the more common measurement mistakes in this data. A small share of AI-referred visitors can convert at a rate that outweighs the drop in total clicks. Users arriving after seeing an AI mention often arrive further along in their decision. The more useful question is whether citation share, sentiment, and downstream branded search are moving in the right direction. Raw click volume matching last year's number matters far less. A UTM-only view of AI referral traffic will typically understate the actual commercial impact, since most citations point to a third party rather than the brand's own domain, and even those rarely convert to a tracked click.
Continuously, not on a monthly or quarterly cycle. The citation overlap figures above shifted from 76% to 38% inside roughly eight months. CTR moved from a 1.3% floor to a 2.4% recovery inside two months. Both moved faster than a typical quarterly reporting cadence would catch. A brand relying on periodic manual checks or a quarterly agency report is often reacting to a shift that already happened weeks or months earlier. Continuous, prompt-level tracking is what surfaces a citation or sentiment change while there's still time to respond to it.
.svg)
.svg)
