Google AI Overview SEO impact: what published studies found about clicks

The Google AI Overview SEO impact, in short

The SEO impact of Google AI Overviews is fewer clicks to ordinary results on the searches where an Overview appears. Every independent study below agrees on the direction. They disagree on the size, because each measured something different.

An AI Overview is the written summary Google shows above some search results, with links to the pages it drew on. When it answers the question on the page, some people stop there. The question for a site owner is how many, on which searches, and whether the clicks that remain are worth more.

• Pew Research Center watched real people search and found they clicked an ordinary result about half as often when an AI summary appeared.
• Ahrefs compared click-through rates for the top result across hundreds of thousands of keywords and found drops of 34.5% in one study and 58% in a later one.
• Seer Interactive tracked its own clients’ informational queries and found organic click-through rate down 61% where an Overview appeared.
• Google says total organic clicks have been relatively stable and that the clicks are of higher quality, without publishing figures.

The studies side by side

The studies used different people, different searches and different periods, so their figures cannot be averaged. Read each row on its own terms.

Pew Research Center: what real people did

Pew found that people clicked an ordinary search result on 8% of visits when an AI summary appeared, against 15% when one did not. It is the only study here that watched real people rather than aggregate data.

Pew’s study, published on 22 July 2025, used the browsing data of 900 US adults who had agreed to have it recorded, across 68,879 Google searches in March 2025. 18% of those searches produced an AI summary. People clicked a link inside the summary itself on 1% of visits. They ended their browsing session on 26% of pages with a summary, against 16% without one.

Pew states its own limits. It covered Google only, and it says a summary for a given search may change over time, so the results reflect what appeared when the data was collected. It measured how people behaved, not why.

Ahrefs: the click-through rate of the top result

Ahrefs found that the top-ranking page got a lower share of clicks when an AI Overview appeared, by 34.5% in its first study and 58% in its second. The second figure is larger partly because the method changed.

The first study, by Ryan Law, was published on 17 April 2025. It took 150,000 keywords that showed an AI Overview and 150,000 informational keywords that did not, and compared aggregated Search Console data from March 2024, before the US rollout, with March 2025. For keywords with an Overview, the top position’s click-through rate fell from 0.073 to 0.026. Ahrefs attributes a 34.5% reduction to the Overview after allowing for the fall seen on keywords without one.

The update, by Ryan Law and Xibeijia Guan, was published on 4 February 2026. It used the same size of sample but compared desktop data from December 2023 with December 2025. The top position’s click-through rate on keywords with an Overview fell from 0.073 to 0.016, and the study reports a 58% lower rate. The authors say they used a two-year gap, against one year before, because December 2024 was already after the rollout. A longer gap leaves more room for other changes in search to creep into the figure.

Seer Interactive: being cited helps, but the cause is unproven

Seer found organic click-through rate down 61% on queries with an AI Overview, and found that brands cited in the Overview did better than brands that were not. Seer itself says it cannot prove the citation caused the difference.

Seer’s September 2025 update, by Tracy McDonald and published on 4 November 2025, covered 3,119 informational queries across 42 client organisations, with 25.1 million organic impressions and 1.1 million paid impressions from June 2024 to September 2025. Organic click-through rate on queries with an Overview fell from 1.76% in June 2024 to 0.61% in September 2025. Paid click-through rate fell from 19.7% to 6.34% over the same period. Seer says the study covers only informational and educational queries. Queries without an Overview fell too, from 2.72% in June 2024 to 1.62% in September 2025, a fall of 41%.

Two things in that table matter. Being cited did not protect a brand from a fall, it only made the fall smaller. And queries without an Overview also fell sharply, which means the Overview is not the only thing pulling clicks away from ordinary results.

What Google says, and why it is hard to check

Google says total organic clicks have been relatively stable and that clicks from AI Overviews are higher quality. It has not published the data behind either claim.

On 6 August 2025 Liz Reid, the head of Google Search, wrote that “total organic click volume from Google Search to websites has been relatively stable year-over-year” and that Google is “sending slightly more quality clicks to websites than a year ago”. She defines a quality click as one where the person does not quickly click back. She also wrote that outside studies are “often based on flawed methodologies, isolated examples, or traffic changes that occurred prior to the roll out of AI features in Search”.

Google made the same point to site owners in a May 2025 post, saying that clicks from results pages with AI Overviews “are higher quality, where users are more likely to spend more time on the site”. Neither post gives a sample, a period or a figure. Google has the only full data set, so its claims cannot be tested from outside, and the studies above cannot see what Google sees.

The studies agree on the direction and disagree on the size. Your own Search Console data is the only figure that describes your site.

Where the studies disagree, and why

The figures differ mostly because the studies measured different things over different periods. None of them is simply wrong.

There is one point every source shares, Google included. Some people who see an AI Overview get their answer and do not click. Where they part is on whether the clicks that remain make up for it.

What none of the studies can tell you

None of the studies measures what a lost click was worth, and none covers every kind of search. Keep those gaps in mind before you apply any figure to your own site.

• Value is missing. The studies count clicks, not sales, sign-ups or leads, so they cannot say whether the clicks that remain are worth more, as Google claims.
• Mentions are missing. An Overview can name a brand without a click, and no study here counts that as a gain or a loss.
• Most of the data is American. Pew studied US adults, and the other studies do not claim to describe every country or language.
• The ground keeps moving. Google changes when Overviews appear and how they show links, so a figure from 2025 describes 2025.
• AI Mode is outside most of them. It is a separate, conversational view of Search, and these studies were built around AI Overviews.

The gaps cut both ways. They are a reason not to panic over the largest figure, and also a reason not to relax because of Google’s reassurance. Only your own data, read over time, settles it for your site.

What to do with this

Measure your own site rather than borrowing a figure from a study. The studies tell you what to look for, not what you will find. Our guide to reporting AI visibility honestly sets out rules for the numbers.

1. Open the Generative AI performance report in Search Console to see which of your pages appear in AI Overviews. Google says it shows impressions only, not clicks.
2. In the ordinary Search performance report, compare click-through rate on your informational queries this year with the same months last year.
3. Split the comparison by query type. A fall on “what is” questions and no fall on product or brand searches is the pattern the studies would predict.
4. Judge visits by what they do, not only by how many arrive. Google itself suggests looking at sales, sign-ups and engagement rather than clicks alone.
5. For questions where you lose clicks, ask whether your page could be the one cited. Seer’s data suggests a cited brand loses less, though it cannot prove why.

Our guide to Google AI Overviews SEO covers what Google says makes a page eligible to be cited. Agentcites does not measure AI Overviews, because there is no public way to ask them a question. It measures ChatGPT, Perplexity, Gemini and DeepSeek through their APIs, as the method page explains.