What Are AI Overviews?
AI Overviews are AI-generated summaries that appear above the traditional results for many Google queries, synthesising an answer from multiple pages and linking the sources it drew on. They are built on the ordinary Google Search index — not a separate AI index — which is why the levers that control them are the ones you already know.
What the Data Actually Says About Clicks
"Zero-click search" gets asserted far more often than it gets measured. The most rigorous public measurement to date comes from Pew Research Center, which tracked real browsing behaviour rather than surveying opinions:
"Users who encountered an AI summary clicked on a traditional search result link in 8% of all visits. Those who did not encounter an AI summary clicked on a search result nearly twice as often (15% of visits)."
— Pew Research Center, from the browsing data of 900 U.S. adults across 68,879 Google searches in March 2025, of which 12,593 produced an AI summary
There is a second number in that study that deserves more attention than the headline one. Being cited inside an AI Overview is not a reliable traffic channel either:
"Google users who encountered an AI summary also rarely clicked on a link in the summary itself. This occurred in just 1% of all visits to pages with such a summary."
— Pew Research Center, July 2025
Read those two figures together and the strategy changes. If citation inside an AI Overview converted to clicks, "get cited" would be a traffic tactic. At a 1% click rate it is mostly a brand visibility outcome: your name and position appear in front of someone forming an opinion, and the commercial return shows up later and elsewhere. Plan for it as brand impression, and measure it as brand impression — not as a channel you can attribute sessions to in Analytics.
What Google Tells Publishers to Do
There is a large industry of "GEO" advice built on the premise that AI Overviews need special treatment. Google's own documentation says the opposite, unusually plainly:
"There are no additional requirements to appear in AI Overviews or AI Mode, nor other special optimizations necessary." … "You don't need to create new machine readable files, AI text files, or markup to appear in these features. There's also no special schema.org structured data that you need to add."
— Google Search Central, AI features and your website
Google's stated prerequisite is the ordinary one: the page has to be indexed and eligible to be shown in Google Search with a snippet, meeting Search's technical requirements. Beyond that, it says to "apply the same foundational SEO best practices for AI features as you do for Google Search overall." Google is equally clear that meeting the requirements is not a guarantee of being crawled, indexed, or served.
Read the scope carefully. That guidance governs Google's AI features and nothing else — Google's own wording is "to appear in these features." OpenAI, Anthropic and Perplexity have published no equivalent statement. Their documentation describes what each crawler is for; none of it says what earns a citation. Treating Google's minimum as an industry-wide rulebook is the most common mistake in this field. The sound working position: meet Google's baseline, then do the extra work that plausibly helps engines which have told you nothing — it is cheap, it does not hurt your Google performance, and it is the only hedge you have.
What Actually Moves the Needle
Stripping out the steps that have no documented basis leaves a shorter and more honest list:
| Lever | Why it holds up |
|---|---|
| Rank organically first | AI Overviews draw from the Search index. Being eligible to appear in Search with a snippet is Google's stated prerequisite — everything else is downstream of it. |
| Keep Googlebot unblocked | Ordinary Googlebot is the crawler behind AI Overviews. Google-Extended is a different token and blocking it does not affect Search inclusion — but note it governs Gemini training AND grounding, so blocking it costs you Gemini citations even though AI Overviews are unaffected. |
| Answer the question in one sentence | A summariser needs a clean, attributable statement. Burying the conclusion three paragraphs down makes your page harder to quote than a competitor's. |
| Serve content server-side | Content that only exists after JavaScript hydration may never be seen. This is a Search requirement first, and an AI-feature requirement by inheritance. |
| Source your claims | Factual precision is what makes a page safe for a system to repeat under Google's name. |
| Keep it current | Stale pages lose eligibility the ordinary way — through ranking decline, not a special AI penalty. |
Two tactics on almost every "optimise for AI Overviews" checklist deserve a caveat rather than a place on the list. Adding FAQ or HowTo schema specifically to win Google citations does not work — Google says no special schema is needed — but structured data still earns rich results on its own merits, and whether it helps a non-Google engine quote you accurately is simply undocumented. Publishing an llms.txt file will not help you in Google, which has said it does not use the file; its audience is the agent ecosystem outside Google. Do both for the reasons that hold, not for AI Overviews.
Measuring Whether It's Working
This is the awkward part: Search Console does not break out AI Overview appearances, and at a 1% in-summary click rate you cannot infer citation from traffic either. Impressions from AI Overviews are folded into ordinary Search performance data, so a page can be cited daily and show nothing distinguishable in your reports.
That gap is what SearchVitals AI Visibility exists to close — it measures your brand's mention volume across ChatGPT and Google AI Overviews, the keywords triggering those mentions, and how your coverage compares against named competitors.
For the crawler-level mechanics across every AI platform — which agents are involved in a citation, and why the popular "block the training bots" configuration is a bet rather than a guarantee — see our guide to getting cited by AI answer engines.