The AI Overviews click through rate is low. Pew Research Center found clicks on links inside AI summaries in 1% of visits to such result pages (2025), and a small October 2026 case study estimated about 0.9% for AI impressions. The regular top result loses clicks too.
This guide is part of the generative engine optimization hub. Click through rate (CTR) is clicks divided by impressions. An AI Overview is the generated summary Google shows above or between classic results, with links to its sources, and AI Mode is Google's conversational search mode that answers a whole question with a generated response. The main recommendation: the published figures disagree because they measure different things, so estimate the effect on your own pages in Search Console before you change anything, and treat the Search generative AI control, the Search Console setting that removes a site from AI features, as a last resort.
What do the 2026 studies say about the AI Overviews click through rate?
The 2026 studies say the AI Overviews click through rate is low in two ways: few people click links inside the summary, and the classic result below loses clicks too. Pew measured 1% for summary links in 2025, Ahrefs a 58% lower CTR for position 1 in 2026, and a case study about 0.9% for AI impressions in 2026.
Three data points carry most of the debate as of November 2026:
- Pew Research Center, 2025. In a panel of 900 US adults and 68,879 searches from March 2025, users clicked a traditional result in 8% of visits to pages with an AI summary and in 15% of visits without one. They clicked a link inside the summary in 1% of visits.
- Ahrefs, 2026. An Ahrefs analysis of 300,000 keywords in aggregated desktop Search Console data found position 1 CTR on AI Overview keywords fell from 7.3% in December 2023 to 1.6% in December 2025. Against a forecast of about 3.7% without AI Overviews, that is roughly 58% lower (Ahrefs, 2026). Ahrefs calls it a correlation.
- Cyrus Shepard, 2026. In a case study reported by Search Engine Roundtable on October 30, 2026, he opted one unnamed site out of AI features and estimated AI impressions at about 8% of impressions with roughly 0.9% CTR.
Why do AI Overviews CTR studies give such different numbers?
AI Overviews CTR studies give different numbers because each one measures a different click: a link inside the summary, the organic result below it, or all external clicks from a search. They also differ in method and year, so the 1% figure from 2025 and the 58% figure from 2026 can both be right.
| Study | Year | Method | Click it measures | Headline figure | Main caveat |
|---|---|---|---|---|---|
| Pew Research Center | 2025 | Browsing panel, 900 US adults | Links inside the AI summary, and classic results | 1% inside summary; 8% vs 15% classic | Data from March 2025, before AI Mode was widely used |
| Ahrefs | 2026 | 300,000 keywords, aggregated desktop Search Console data | Position 1 organic result | About 58% below forecast | Correlation, desktop only, vendor study |
| Cyrus Shepard case study | 2026 | One site opted out of AI features | Estimated CTR of AI impressions | About 0.9% | One site, estimate by subtraction |
| Wang and coauthors, arXiv | 2026 | Randomized field experiment, 1,100 US Chrome users | Any click to an external website | Minus 18.8 points in forced AI Mode | Younger, educated sample; forced adoption |
| Reddit Q3 letter | 2026 | Company traffic data | Search referrals to reddit.com | More than 1 million fewer per week in October | No split by cause or search engine |
A CTR near 1% inside the summary does not mean AI Overviews cost you only 1% of clicks. In the Pew and Ahrefs data, the larger loss sits on the classic result below the summary.
Does AI Mode have a lower click through rate than AI Overviews?
AI Mode has a lower click through rate than AI Overviews in the only controlled study so far. Forcing every search into AI Mode cut the share of searches that led to an external website by 18.8 percentage points, a larger effect than hiding AI Overviews had in the other direction, according to an experiment published in August 2026.
The study, by Stephanie T. Wang, Christo Wilson, Danaé Metaxa and coauthors, is a preregistered field experiment on arXiv (2026). Participants spent three baseline days and seven test days in one of three conditions. In the AI Mode condition, use of Bing, DuckDuckGo or Yahoo rose 11.2 points (2026) and trust in Google fell. Before the test, participants used AI Mode for only 0.6% of searches (arXiv, 2026), so the result shows forced adoption.
Reddit is the first large publisher to blame AI Mode on the record. In its Q3 2026 results, Reddit said search referrals fell by more than 1 million per week on average in October, and CEO Steve Huffman said AI Mode "does not drive traffic," Search Engine Journal reported on October 30, 2026. That is a reported opinion: Reddit also blamed machine translated content and gave no split between causes. For the link changes Google made in October 2026, see our Google AI Mode update explainer.
How do you measure the AI Overviews traffic impact on your own site?
You measure the AI Overviews traffic impact on your own site by splitting pages into a high AI exposure group and a low exposure group, then comparing how each group's CTR moves over time. Search Console reports AI feature impressions per page but no AI clicks, so a direct AI click through rate is not available.
Google's Generative AI performance report shows impressions in AI Overviews and AI Mode by page, country, device and date, with no click or CTR metric as of November 2026. The Performance report counts a click on an external link in an AI Overview or AI Mode as a normal click, and gives every link in an AI Overview the same position. In GA4, the default channel group puts AI Overviews and AI Mode visits in Organic Search, not in the AI Assistant channel. Our guide to Google Search Console covers the exports, and an AI Overviews tracker adds which queries show an AI Overview.
The AI Exposure Split, step by step
The AI Exposure Split compares the CTR change of pages with a high share of AI feature impressions against pages with little or none over the same dates, so the extra drop in the high exposure group is the part you can attribute to AI.
- Export AI impressions per page. Take the last three months from the Generative AI performance report.
- Export clicks and impressions per page. Use the same dates and the Web search type in the Performance report.
- Compute each page's AI share. Divide AI impressions by total impressions.
- Split the pages. Put pages with a high AI share in group A and pages with little or none in group B.
- Compare CTR change, not CTR level. Calculate each group's CTR for the same period a year earlier and now.
- Check position. Keep only pages whose average position stayed roughly stable, so ranking changes do not pollute the result.
Hypothetical worked example, not real data. Group A pages had a 4.0% CTR a year ago and 2.6% now (a 35% drop). Group B pages went from 3.0% to 2.7% (a 10% drop). The difference of about 25 points of relative CTR loss is the part you can attribute to AI exposure with some confidence.
Does opting out of AI Overviews bring clicks back?
Opting out of AI Overviews rarely wins back clicks, because the AI Overview keeps appearing for the query with other sources cited. The Search generative AI control removes your links from AI Overviews, AI Mode and generative AI in Discover, and the one public test of it saw regular organic clicks fall, not rise.
The Search generative AI control sits under Settings in Search Console, with the options "Include my site's links and content in Search generative AI features" (the default) and "Exclude my site's links and content from Search generative AI features." Google says exclusion takes effect within one to two days, is not used as a ranking or inclusion signal for other parts of Search, and does not affect AI training. The control rolled out to all sites on August 31, 2026, Search Engine Journal reported.
In Shepard's October 2026 test, pages with the most AI visibility lost position while excluded and regular organic clicks dropped. He called it a possible "Search SERP Halo Effect." Search Engine Roundtable called the study "super limited."
When is opting out of AI Overviews worth it?
Opting out of AI Overviews is worth it only when your own data shows the high exposure pages losing far more CTR than the rest, rankings are stable, and page rewrites have not helped. Excluding a site removes its links from AI Overviews and AI Mode while the summary still shows, so keeping the site included is the default.
| What your data shows | Likely cause | Decision |
|---|---|---|
| High exposure group loses far more CTR, rankings stable | AI Overviews absorb the answer | Rewrite pages to earn the citation first; test exclusion on a URL-prefix child property only |
| Both groups lose CTR equally | Wider search or demand change | Keep the site included |
| AI impressions grow, clicks flat, conversions stable | Citation brings visibility, not visits | Keep the site included and track branded searches |
| Content is the product (answers, data) and clicks keep falling | Answers used without visits | Consider exclusion and log the result for eight weeks |
For page changes that earn a citation, see how to optimize for AI Overviews.
Common mistakes when reading AI Overviews click through rate data
The common mistakes when reading AI Overviews click through rate data come from mixing different metrics, trusting one figure as a benchmark and reading site totals as proof. Each mistake leads to the wrong fix, most often an opt-out that removes visibility without bringing clicks back.
- Treating 0.9% as a benchmark. The 2026 figure comes from one site and was estimated by subtraction. Measure your own page groups instead.
- Comparing a summary link CTR with an organic CTR. A link inside the summary and the classic result below it are different clicks. Name which one a number describes.
- Reading site-wide CTR. A site average hides pages that gained citations and pages that lost clicks. Compare page groups.
- Ignoring ranking changes. A position drop lowers CTR without any AI Overview. Remove pages whose position moved.
- Opting out the whole domain first. Exclusion removes every page from AI features. Test on a URL-prefix child property if you test at all.
- Mixing paid and organic. AI Mode also shows ads, covered in Google AI Mode ads. Keep paid clicks out of the organic CTR comparison.
