google's ai overviews are the bigger ai traffic story nobody's chasing.
here's the number that should be getting more attention than any standalone-assistant traffic study: AI Overview prevalence is climbing fast and the estimates vary by source and date. Conductor's 2026 AEO/GEO benchmarks put it at roughly 25% of tracked queries. Ahrefs data from March 2026 puts it at 48% of all Google queries, up from 34.5% in December 2025. Different samples, different windows, same direction of travel, and the trend line matters more than any single figure. Every one of those is a potential zero-click answer sitting above the organic results your business used to rank for. Google's own AI surface, not ChatGPT, is very likely the larger AI-influenced traffic source overall, and almost nobody outside enterprise SEO teams is optimizing for it directly.
An AI Overview is defined as the AI-generated summary Google places above standard search results, citing a small set of sources. AI Mode is defined as Google's separate conversational search interface, which breaks a question into sub-queries and synthesizes across a broader source pool. They cite substantially different pages.
two different surfaces, two different games
AI Overviews and AI Mode are not the same product, and treating them as one thing is where most advice goes wrong.
AI Overviews are the automatic summary Google shows directly in standard search results for certain queries, similar to how a featured snippet used to work. AI Mode is a separate, more conversational interface that breaks a complex question into sub-queries and pulls from a broader mix of sources to build a fuller answer.
They don't even cite the same pages. Industry analysis from Stackmatix found only about 13.7% overlap between the sources cited in AI Overviews and the sources cited in AI Mode for the same topics. Optimizing for one doesn't automatically win you the other.
| surface | what it is | what tends to get cited |
|---|---|---|
| AI Overviews | automatic summary in standard search results | short, self-contained passages that directly answer one specific question |
| AI Mode | conversational, multi-query synthesis | broader mix of sources, including some that don't rank in the top 10 organically |
the click math is more nuanced than "AI killed my traffic"
Two things are true at once here, and both matter.
First, being cited pays off. Seer Interactive's 2026 analysis found that brands cited inside AI Overviews earn roughly 120% more organic clicks per impression than uncited brands on the same queries. Getting into the citation set reverses most of the click-through penalty that comes with an AI Overview appearing above you.
Second, a lot of that traffic never reaches a website at all. Seer's separate analysis of AI Mode sessions, 25.1 million impressions, found that 93% of AI Mode sessions end without a click to any external site. People are getting a complete answer inside the interface and stopping there. That's not a reason to give up on AI Overviews, it's a reason to stop measuring success by click volume alone and start measuring citation frequency as its own goal, the same way brand mentions matter even when someone doesn't click through immediately.
does ranking still matter?
Here the data gets genuinely mixed, and it's worth knowing both sides rather than picking the one that's more convenient.
Ahrefs, in a 730,000-response study, found that 76.1% of URLs cited in AI Overviews already rank in the top 10 organic results, which argues that traditional SEO is still the foundation. That share is falling fast, though: Search Engine Land's analysis of 863,000 keywords and 4 million AI Overview URLs found citations from top-10 pages dropped from 76% in mid-2025 to 38% by March 2026. Separately, analysis from iPullRank and LinkSurge found that 68% of pages cited in Google's AI Mode sit outside the top 10 of the classic search results, and BrightEdge puts a similar figure at roughly 80% for AI Mode specifically. Different studies, different surfaces, different methodologies, and they point in different directions.
The honest read: AI Overviews still lean heavily on pages that already rank well. AI Mode, because it fans a question out into many sub-queries and pulls a specific passage for each one, will cite a well-written, well-contextualized paragraph even on a page that doesn't crack the top 10 overall. Rank still helps. It's no longer sufficient on its own, especially for AI Mode.
what actually earns the citation
Google published its own AI Optimization Guide in May 2026, and reportedly avoided the terms AEO and GEO entirely, framing the goal as one thing: optimizing for the search experience, full stop, rather than a separate parallel discipline. A few consistent findings across the coverage of that guide and related 2026 analysis:
- Short, self-contained passages get extracted. A paragraph that clearly and completely answers one specific question is far more likely to be pulled into an AI Overview than information spread across a long, undifferentiated block.
- Structured data correlates with citation, but isn't a guarantee. Stackmatix's 2026 data found 65% of pages cited by AI Mode and 71% of pages cited by ChatGPT carry some form of structured data (Article, Organization, Product, Review, FAQ, or HowTo). Correlation, not causation, and not every schema type contributes equally.
- Freshness is measurable, not a vibe. Refreshing evergreen content on a quarterly cadence, updating dateModified schema, and adding at least one new data point with each refresh are consistent recommendations across multiple 2026 sources.
- First-hand experience and a specific point of view matter more here than in classic SEO. Google's own framing rewards non-commodity content, work that shows lived expertise rather than a repackaged summary of what ten other pages already say.
what to actually do this week
If you've been putting all your AI-visibility effort into ChatGPT specifically, that's the smaller and more crowded target. Google's AI surfaces reach far more queries and are less contested right now, precisely because most small operators assume "AI SEO" means optimizing for a chatbot. Before writing new content aimed at AI Overviews, make sure the basics from the crawler audit are already handled, none of this works if Google's own systems, or the AI layers built on top of them, can't reach your pages in the first place.
Part of the same series on Google's AI surface: you can now buy an ad inside google's ai answer. it will not get you cited.
want a straight read on where your site stands?
the free ai visibility scan checks your site across chatgpt and google's ai surfaces in about sixty seconds. results on the page, no email needed.
run the free scanfrequently asked questions
What is the difference between AI Overviews and AI Mode?
AI Overviews are automatic summaries shown above standard Google results. AI Mode is a separate conversational interface that fans a question into sub-queries and synthesizes across a broader source pool. Industry analysis found only about 13.7% overlap in the sources they cite.
Does ranking on Google get me into AI Overviews?
It helps substantially but does not guarantee it. Ahrefs found 76.1% of URLs cited in AI Overviews already rank in the organic top 10. AI Mode behaves differently, with separate analyses finding most of its citations come from pages outside the top 10.
Is it worth optimizing for AI Overviews if most sessions end without a click?
Yes, for citation rather than clicks. Seer Interactive found brands cited inside AI Overviews earn roughly 120% more organic clicks per impression than uncited brands on the same queries, which reverses most of the click penalty.
What content gets cited in Google's AI answers?
Short, self-contained passages that completely answer one specific question. A paragraph that stands alone is far more likely to be extracted than the same information spread across a longer block.
How often should I refresh content for AI citation?
Quarterly is a reasonable minimum for anything topical. Update the dateModified field and add at least one genuinely new data point with each refresh rather than only rewording.
written by elisabeth hitz, certified in anthropic's ai fluency program (framework & foundations, and ai capabilities & limitations), plus claude 101 and claude cowork. sources: conductor 2026 aeo/geo benchmarks; seer interactive 2026 ai overviews ctr and ai mode click analysis (25.1 million impressions); ahrefs 730,000-response ai overviews citation study; search engine land analysis of 863,000 keywords and 4 million ai overview urls (march 2026); ipullrank/linksurge ai mode citation analysis; brightedge citation data; stackmatix 2026 structured-data citation study; google search central's may 2026 ai optimization guide, as reported via digivate, e2m solutions, primeaicenter, orange monke, nexuro digital, and digital applied. published 26 july 2026.