Google cites pages in AI Overviews that are indexed and snippet-eligible, with no special markup. Then it picks passages that answer a sub-question.
LLaunchScaler·Published ·9 min read
To get cited in Google AI Overviews, a page must be indexed and eligible to show in Google Search with a snippet; Google states there are no additional technical requirements and no special schema to add. Past that bar, Google links the pages whose passages answer the sub-questions it searched for, which is why a top-10 ranking on its own does not earn the citation.
So the work has two parts. First make sure nothing on the page or the server stops Google from indexing it or quoting it. Then write each answer as a short passage that can be lifted whole.
What does Google require for a page to appear in AI Overviews?
Google's "AI features and your website" guide sets one condition: to be shown as a supporting link in AI Overviews or AI Mode, a page must be indexed and eligible to be shown in Google Search with a snippet. It adds: "There are no additional technical requirements." Everything else is ordinary Search eligibility.
The same guide rules out the shortcuts people sell. It says "You don't need to create new machine readable files, AI text files, or markup to appear in these features," and that there is no special schema.org structured data you need to add. By Google's own account, an llms.txt file, an "AI sitemap" or a custom JSON feed is not needed for AI Overviews.
Two limits are worth knowing before you start measuring:
AI Overviews only appear when Google's systems decide they add something to classic results, and Google says they "often don't trigger." A query with no AI Overview cannot cite anyone.
Meeting every requirement does not oblige Google to crawl, index or serve the page. Google says so directly: "Indexing and serving isn't guaranteed."
Google's list of SEO basics that still apply is short and specific. Crawling has to be allowed in robots.txt "and by any CDN or hosting infrastructure," important content has to be available as text, internal links have to make the page findable, and structured data has to match the visible text on the page.
Why isn't ranking in the top 10 enough?
Ranking helps, but it does not decide which page gets the link. AI Overviews use a "query fan-out" technique, which Google describes as issuing multiple related searches across subtopics and data sources to build one response. A page can be cited for a sub-question it answers well even when another page holds position one for the head query.
Questions, answered
What people ask about this
01
Do I need special markup to appear in Google AI Overviews?
No. Google says a page only has to be indexed and eligible to show in Search with a snippet, and that there is no special schema.org structured data, AI text file or machine-readable file to add.
02
Will ranking first on Google get my page cited in the AI Overview?
Two large studies from July 2025 measured the overlap and reached figures that look contradictory until you see what each counted.
Study
What it measured
Finding
Semrush, 200,000 keywords (100,000 US mobile, 100,000 US desktop)
Overlap between all AI Overview links and the organic top 10
"Only around 20 to 26%" overlap. The top-ranked result appeared in 46% of desktop and 34% of mobile AI Overviews.
Ahrefs, 1.9 million citations from 1 million AI Overviews
Ranking of the top 3 most visible citations in each AI Overview
76.10% of those cited pages ranked in the top 10; about 86% ranked in the top 100.
Read together, they say the most prominent citations usually come from pages that already rank, while most of the links in an Overview do not come from the top 10 for that exact query. Semrush also found that over 50% of desktop and 60% of mobile AI Overviews did not link to the top organic result at all.
The practical reading: rank for the sub-questions, not only the head term. If your page targets "best invoicing tool for freelancers," the fan-out may also search "invoicing tool with recurring billing" and "invoice tool that handles VAT." A page with a direct, quotable answer to each of those has more ways in than a page tuned for one keyword. The guide to writing content AI answers cite goes deeper on that structure.
How do you check that a page is eligible?
Check eligibility page by page, starting with the pages you most want cited. Every check below reads what Googlebot receives, not what your browser shows, because a stray header or a CDN rule can hide a page from Google while it looks fine to you.
Open Search Console, paste the URL into URL Inspection and confirm it says the URL is on Google. If it is not indexed, nothing else in this article applies yet.
In the same tool, run Test live URL and read "Crawl allowed?" and "Indexing allowed?". A robots.txt block or a noindex shows up here.
View the page source and search for nosnippet, max-snippet and data-nosnippet. Then run curl -sI https://yoursite.com/page and look for an X-Robots-Tag header carrying the same rules.
Confirm your CDN, firewall or bot protection is not challenging Googlebot. Google lists "any CDN or hosting infrastructure" alongside robots.txt as something that must allow crawling.
Confirm the answer is in the HTML as text. An answer inside an image, a PDF or a tab that only loads on click is harder to quote.
If the page has structured data, check that every value in it also appears in the visible text.
How should you write a passage Google can cite?
Put the question in a heading and answer it completely in the first 40 to 60 words below it. Name the subject instead of writing "it," include the specific number or setting, and make sure the passage still makes sense if it is the only thing a reader sees. Detail and caveats go after that opening answer.
No search engine publishes a passage length. Forty to sixty words is a working target: long enough to hold a complete answer with its condition, short enough to be quoted without cutting. Microsoft gives the same advice for its own AI answers, in an October 2025 post on getting content into AI search answers: write "self-contained sentences that make sense when extracted," use question-and-answer structure that can be "lifted word for word," and avoid hiding answers in tabs or expandable menus.
Compare two answers to "does the plan include an API?". The weak one says "Absolutely! Our platform was built with developers in mind, and we think you'll love how flexible it is." The citable one says "The Team plan includes the REST API with 10,000 requests a day. The Starter plan has no API access; upgrading from Starter to Team turns it on immediately."
The second one answers the question, states the limit and names both plans, so an engine can quote it without adding anything. Tables work the same way for comparisons: one row per option, one column per attribute, values instead of adjectives. How to optimize for AI search covers the rest of the on-page work.
Which snippet controls stop AI Overviews from quoting you?
Google's snippet controls apply to AI Overviews too. The robots meta tag documentation says nosnippet also prevents content from being used "as a direct input for AI Overviews and AI Mode," and max-snippet limits how much content may be used that way. If a page you want cited carries either rule, remove it.
Control
Where it goes
Effect on AI Overviews
nosnippet
<meta name="robots"> or X-Robots-Tag
No text snippet, and not used as a direct input for AI Overviews or AI Mode
max-snippet:[number]
Meta tag or header
Caps the characters used as a snippet and as direct input
max-snippet:0
Meta tag or header
Same as nosnippet
max-snippet:-1
Meta tag or header
Google chooses the length; no limit
data-nosnippet
Attribute on span, div or section only
That part of the page is not used in snippets
noindex
Meta tag or header
The page leaves the index, so it cannot be cited
Google-Extended in robots.txt
robots.txt
No effect on AI Overviews; it governs Gemini Apps and Vertex AI
When rules conflict, Google applies the more restrictive one, so max-snippet:50 plus nosnippet means no snippet. Two mistakes are easy to make: an SEO plugin setting a small max-snippet value site-wide, and a template wrapping the main content column in data-nosnippet to keep a cookie banner or pricing widget out of snippets. If your goal is the opposite, to keep a page out, the guide to opting out of Google AI Overviews walks through each control.
Does structured data help you get into AI Overviews?
Structured data helps Google understand a page, but it does not buy inclusion. Google's AI features guide says there is no special schema to add, and its structured data guidelines say "Google does not guarantee that your structured data will show up in search results," even when the markup passes the Rich Results Test.
Use it for what it does well. Organization markup on the home page helps Google disambiguate your company. Product markup states a price and availability as typed values. FAQ-style content helps because the question-and-answer text is on the page, not because of the markup around it. The rule that matters for AI features is the one Google repeats: the markup must be "a true representation of the page content," so a price or a feature that exists only in JSON-LD is a liability.
How do you measure AI Overview citations?
Search Console counts AI Overview and AI Mode appearances inside the Performance report's Web search type, with no separate filter, so you cannot isolate them there. Track citations by hand: keep a list of 20 to 30 real questions, search each one monthly, and record whether an AI Overview appeared and which of your URLs it linked.
Record the date, the query, whether an Overview showed, and every cited URL, yours and others'. After three months you will see which pages earn links and which competitors hold the sub-questions you miss. Google also says clicks from results pages with AI Overviews are "higher quality," meaning people spend more time on the site, so pair the log with conversions in your analytics rather than raw sessions.
AI Mode is measured the same way but behaves differently; Google says the two "may use different models and techniques, so the set of responses and links they show will vary." The guide to ranking in Google AI Mode covers what changes there.
Check what Google can read before you rewrite anything
Most AI Overview problems start with a page Google cannot index or cannot quote, and those are cheaper to fix than a content rewrite. Run the free scan on LaunchScaler: it takes a URL and no account, and runs 156 checks across 6 of its 7 categories at no cost. Its search checks read the blockers that decide eligibility, including robots.txt rules that block a page you want ranked, noindex in the meta tag or the X-Robots-Tag header, canonicals pointing at redirected or 404 URLs, soft 404s and main content that only appears after JavaScript runs. It also checks whether the page has a concise answer passage an engine could lift. Fix what it flags, run Test live URL in Search Console, then request indexing.
Not reliably. Semrush's study of 200,000 keywords found the top-ranked result in 46% of desktop AI Overviews and 34% of mobile ones, and only about 20 to 26% overlap between AI Overview links and the organic top 10.
03
How do I stop Google from quoting my page in AI Overviews?
Use nosnippet, max-snippet, data-nosnippet or noindex. Google's documentation says nosnippet also stops the content being used as a direct input for AI Overviews and AI Mode. Blocking Google-Extended does not remove you from AI Overviews.
04
Can I see AI Overview clicks in Search Console?
Only as part of the total. Google counts AI Overviews and AI Mode in the Performance report under the Web search type, with no separate filter for them.
05
How long before a change shows up in AI Overviews?
Google has to recrawl and reprocess the page first. Its AI features guide says crawling can take anywhere from several days to several months, depending on how often Google decides a page needs refreshing.
