Google's AI Overviews are changing search results. DXDasia breaks down how to keep your brand visible and clicked as AI answers take over.
Quick answer: AI Overviews cut clicks to top organic results by roughly 30% to 60% in most independent studies, though Google says total clicks are stable. Measure your own impact by pairing Search Console's generative AI report, which shows AI impressions by page, with Performance report clicks, then judge results on conversions rather than sessions.
If your organic clicks have fallen while impressions have held up or risen, AI Overviews are a likely cause, but not the only one. The difficulty is proving it. Google blends AI Overview and AI Mode data into its normal reports, the newer AI-specific report shows impressions without clicks, and analytics tools cannot see an AI Overview at all. This guide summarises what the independent CTR studies found, what Google's tools can and cannot show you, and a step-by-step method we use to measure the impact on a real site before deciding what to do about it.
Most independent studies find a large drop in click-through rate when an AI Overview appears, commonly between about 30% and 60% for the top organic result. The figures differ because the studies measure different things: user panels, aggregated Search Console data and keyword trackers.
| Study | Data and period | Finding |
|---|---|---|
| Pew Research Center | 68,879 searches by 900 US adults, March 2025 | Users clicked a result in 8% of visits with an AI summary versus 15% without; 1% clicked a link inside the summary |
| Ahrefs (Feb 2026) | 300,000 keywords, Search Console data, Dec 2023 vs Dec 2025 | Position-one CTR down 58% on AI Overview keywords (0.073 to 0.016); position two down 50.8% |
| BrightEdge (May 2025) | BrightEdge client data, first year of AI Overviews | Impressions up over 49%; click-throughs down nearly 30% since May 2024 |
| Semrush (Dec 2025) | Same keywords before and after an AI Overview appeared | Zero-click rate fell slightly, from 33.75% to 31.53% |
One finding in the table complicates the simple story: Semrush's before-and-after comparison found no rise in zero-click behaviour for the same keywords. Pew's panel also found users were more likely to end their browsing session after a page with an AI summary, 26% of visits versus 16%.
Possibly, but the evidence is thin and correlational: brands that get cited may already be the stronger brands. Being the cited source is still a better outcome than being absent, so it is worth pursuing even when overall clicks fall.
Google disputes the size of the decline. In August 2025, Google's head of Search wrote that total organic click volume from Google Search to websites has been relatively stable year-over-year and that Google was sending slightly more quality clicks, meaning clicks where users do not quickly return. Google's developer documentation makes the same claim: clicks from results pages with AI Overviews are higher quality, with users more likely to spend more time on the site. In July 2026, it said it was sending billions of clicks to websites every week through AI features in Search.
These positions are not necessarily contradictory. Google reports totals across the whole web; the independent studies measure CTR on specific queries. The same Google post concedes that user trends are resulting in decreased traffic to some sites and increased traffic to others. Your job is to find out which group your site is in, and for which pages.
Partly. Clicks and impressions from AI Overviews and AI Mode are included in the standard Performance report, inside the "Web" search type, but they are blended with everything else. Since 2026 there is also a separate report that shows AI impressions only.
Google announced Search Generative AI performance reports on 3 June 2026 and rolled them out to all websites by 31 August 2026. They show impressions, pages, countries, devices and dates for generative AI features in Search and Discover. They do not show clicks or queries; Search Engine Roundtable confirmed in August 2026 that the report still does not include any click data or query data. Google's John Mueller has also said position for these features is tracked as a block and not separated out in the report.
In practice, this means the new report tells you which pages are exposed to AI features, and the Performance report tells you what happened to their clicks. You need both.
Segment your pages into AI-exposed and not exposed, then compare their click and CTR trends at similar positions. This is the method we use:
Keep the scale in proportion. In Ahrefs' February 2025 study of 3,000 sites, 0.17% of the average website's traffic came from AI chatbots. Assistant referrals are growing, but for most businesses the bigger number is the Google organic CTR change measured above.
Protect the pages that earn revenue, aim for citation on the queries you cannot win clicks for, and stop treating informational traffic as the goal in itself.
Before blaming AI Overviews for a weak CTR, check your own titles. On our site, 36 titles were over 60 characters, many cut off by a long " | Blog" suffix, and 48 descriptions were too long or too short. On a Singapore exhibition-display client, we rewrote titles on 72 pages and descriptions on 73 pages to fit Google's limits.
The most expensive measurement mistake we see is not about AI at all. Across several client accounts, we found the same pattern: a website redesign or relaunch removed the Google Ads conversion tag, and the account kept spending while reporting zero or wrong conversions. The same thing happens to analytics. If a relaunch coincides with a rise in AI Overviews, it is very easy to blame Google for a drop that is really a missing tag. That is why step one of our method is verifying tracking, and why we re-check conversion tags after every rebuild before judging any performance trend, paid or organic.
The second lesson is about the numbers people quote. When we audited our own blog, two statistics did not match any source, including an "AI Overviews appear on 40-50% of searches" claim. We corrected both to sourced figures. In a topic where studies range widely by sample, date and country, a confident unsourced figure can send a whole strategy the wrong way. Every number in a report to a client or a board should link to where it came from, and say when it was measured.
Both checks are standard in our SEO reporting: tracking integrity first, sourced benchmarks second, and only then a conclusion about AI Overviews.
Accept that CTR on informational queries is lower for good, and measure the right thing. Google's counter-claim about stable totals and better-quality clicks may well be true in aggregate, and the only way to know whether it is true for you is the segmented comparison above, judged on conversions rather than sessions. We would not cut content budgets in a panic, and we would not ignore the shift either. Move effort from generic explainers to pages that convert, chase citation on the questions you cannot win clicks for, and report AI exposure separately from organic clicks so the business sees both. For where Google is heading next with AI Mode and Gemini, see our Google to Gemini roadmap.
Not yet. AI Overview and AI Mode clicks are counted inside the Web search type of the Performance report, blended with regular results. The generative AI performance report shows impressions by page, country and device but, as of its full rollout on 31 August 2026, no clicks or queries. Combine the two reports to estimate the click impact on AI-exposed pages.
Because Google assigns every link inside an AI Overview the same position as the overview itself. If the overview sits at the top of the page, any link in it reports as position 1, even if it is the last of several sources and rarely clicked. Read position alongside CTR, and treat a high position with very low CTR as a sign of AI Overview exposure.
No. Visits from links in AI Overviews and AI Mode arrive from Google and are recorded as Google organic traffic. Only standalone assistants such as ChatGPT, Perplexity, Gemini, Copilot and Claude can be separated, using a custom channel group that matches their referrer domains. Some assistant traffic still lands in direct because referrer data is not always passed.
Usually not. Google offers nosnippet, data-nosnippet and max-snippet controls, but they also limit your normal search snippets, which can reduce clicks further. For most businesses, being the cited source is a better outcome than being absent.
Compare at least a full quarter against the same quarter a year earlier to remove seasonality, and check for other changes first, such as a site relaunch, tracking errors or a core update. Search Console keeps a limited window of history, so start exporting the data regularly if you want a longer baseline. A pattern of exposed pages losing more CTR than unexposed ones at similar positions is the signal to look for.