How to Rank in Google AI Overviews

Short answerRanking in Google AI Overviews means being one of the linked sources in the AI summary above the results. Google says a page only needs to be indexed and eligible for a snippet, and that no special markup is required. So rank for the sub-questions Google's query fan-out searches, answer them directly, and stay snippet-eligible.

Key takeaways

Only 37.9% of URLs cited in AI Overviews ranked in Google's top 10 for the query in Ahrefs' March 2026 study of 863,000 SERPs, down from about 76% in its July 2025 study.
AI Overviews appeared on 57.9% of question queries and 46.4% of queries of seven or more words, against 20.5% of all 146 million SERPs Ahrefs checked in September 2025.
Being cited pays: Seer Interactive found informational queries had a 2.07% organic CTR for brands cited in the AI Overview, against 0.94% when not cited.
Since 31 August 2026, every site has a Search Console setting to opt out of AI Overviews and a Generative AI report that counts impressions in them.
Google says llms.txt and similar files are not used by Search, structured data isn't required, and FAQ rich results stopped showing on 7 May 2026.

What does it mean to rank in AI Overviews?

To rank in an AI Overview is to be one of the pages Google links as a source for the AI-written summary at the top of the results. There's no numbered position inside the summary, but there's a clear win and a clear loss: either your link is in the source list for the queries that matter to you, or a competitor's is.

It's worth doing because of what happens to the clicks. In Pew Research Center's study of 68,879 searches by 900 US adults in March 2025, people clicked a traditional result on 8% of visits with an AI summary against 15% without one, and clicked a link inside the summary on just 1%10. Ahrefs measured a 58% lower average click-through rate for the top result when an AI Overview was present, comparing December 2023 with December 202511. If the summary is going to sit above you anyway, you want to be in it. Seer Interactive's data from 53 brands, January 2025 to February 2026, shows why: on informational queries, organic CTR was 2.07% when the brand was cited in the AI Overview and 0.94% when it wasn't8.

AI Overviews share infrastructure with AI Mode and the Gemini app, but they're controlled and measured differently. For how Google's retrieval compares with ChatGPT, Perplexity and Claude, see how AI engines choose their sources.

How does Google choose sources for AI Overviews?

Google retrieves candidate pages from its normal Search index with its core ranking systems, runs extra related searches (query fan-out), and links the pages that support the summary. That's the whole pipeline Google describes, and each part tells you something.

Eligibility: indexed, snippet-eligible, opted in

Google's AI features documentation says a page "must be indexed and eligible to be shown in Google Search with a snippet," and that "there are no additional requirements to appear in AI Overviews or AI Mode, nor other special optimizations necessary"1. Its 2026 AI optimization guide adds a third condition: the site "must be included in Search generative AI features in Search Console"2. Inclusion is the default for every property, so this only bites if someone switched it off3.

Retrieval: core ranking plus query fan-out

The AI optimization guide says Google relies on "our core Search ranking systems to retrieve relevant, up-to-date web pages from our Search index"2. It also uses query fan-out, which it defines as "a set of concurrent, related queries generated by the model"2. A question about choosing payroll software might fan out into searches about pricing, integrations and compliance. A page can be cited because it ranks for one of those sub-queries even if it ranks nowhere for the original.

That explains the most important number on this page. In July 2025 Ahrefs found about 76% of cited URLs ranked in the top 10 for the query12. In its March 2026 update, with a larger sample, better parsing and Gemini 3 running AI Overviews, that share was 37.9%; 31.2% of cited URLs ranked 11 to 100 and 31.0% ranked beyond 10013. Ahrefs cautions the two studies aren't strictly comparable. BrightEdge reported a similar picture in September 2025: "Only 16.7% of citations come from top 10 results" in its nine-industry dataset9. The direction is still clear: ranking for the head term is no longer the main route in.

When an AI Overview appears at all

Google says AI Overviews are shown only when its systems determine they're "additive to classic Search"1. In practice that means longer, question-shaped searches. Ahrefs' trigger study of 146 million desktop SERPs from September 2025 found14:

Query type	Share showing an AI Overview
All keywords	20.5%
Question queries	57.9%
Seven or more words	46.4%

The overall rate swings. Semrush tracked AI Overviews on 6.49% of its 10 million keywords in January 2025, a peak of 24.61% in July 2025 and 15.69% in November 202515. Since January 2026 the summaries run on Gemini 3 by default16, and Google said in July 2026 that it had brought AI Overviews and AI Mode together into a single Search experience17.

Which sources get cited

Video and community sites lead. In Ahrefs' Brand Radar data for US AI Overviews in September 2026, YouTube took 22.9% and Reddit 18.5% of the citations going to the 50 most-cited sources, with Wikipedia at 4.0%18. Google's own guide notes that its AI features "can bring in relevant images and video," and that Merchant Center and Business Profile data can help products and services show up in AI responses2.

Our takeStop treating AI Overviews as a top-10 bonus. With well over half of cited URLs now ranking outside the top 10 for the query, the winning unit is the sub-question: a section that cleanly answers one thing the fan-out searches for. Google's own guidance is unusually blunt that no trick files, special schema or chunking are needed, and we believe it, because the citation data shows the same thing. Your edge is coverage of the follow-up questions and content that isn't a rewrite of what already ranks.

The 8-step plan to get cited in AI Overviews

Steps 1 to 3 are a week of setup. Steps 4 to 7 are where rankings move. Each step ends with a test.

1. Find the queries where AI Overviews appear for you

In Search Console, open the Performance report and filter queries with a custom regex for question words, for example ^(how|what|why|which|when|can|does|is|should)\b, then add queries of seven or more words. Those are the shapes most likely to trigger an AI Overview. Check the top 50 by impressions by hand in a clean browser, or use a tracker from our AI Overview tracker comparison, and record whether an AI Overview showed and whether you were cited.

Done when: you have a list of 30 to 50 queries that show an AI Overview, each marked cited or not cited, with your organic position.

2. Confirm you're eligible

Run each money page through URL Inspection and look for "URL is on Google." Search your templates for nosnippet, data-nosnippet and max-snippet; Google lists them as the controls that limit what it shows from a page1. Then open Settings, Search generative AI, in Search Console and confirm the property is included3.

Done when: every money page is indexed, carries no snippet restriction on its main content, and the Search Console setting shows the site as included.

3. Map the fan-out for each query

For each target query, list the sub-questions Google is likely to search: the People Also Ask questions, related searches, and the sub-topics covered by the pages cited today. Open each cited page and note the passage that was linked. Google calls the fan-out "concurrent, related queries"2; you're reconstructing them from the outside.

Done when: each target query has five to ten sub-questions, and you know which of them your page answers.

4. Answer each sub-question in its own section

Give each sub-question an H2 or H3 phrased the way people ask it, answer it in the first sentence, and support it with a specific number or example and its source. Keep the page as long as the reader needs. Google says there's "no requirement to break your content into tiny pieces"2, so this is about covering the questions, not chopping the page. Our citability grader checks answer-first structure.

Done when: every sub-question on your list has a heading and a one-sentence answer on the page that should own it.

5. Add what a summary can't copy from everyone else

Google's guide separates commodity content, which "typically adds little unique insight," from content with "unique expert or experienced takes that go beyond common knowledge"2. Add original data, a worked example, a tested result, a real price, a screenshot of the actual setting. If your section says what the five cited pages already say, there's no reason to cite you sixth.

Done when: each target page has at least one element (data, example, test or first-hand detail) that the currently cited pages don't.

6. Cover the formats Google pulls in

YouTube is the single most-cited source among the top 50 in Ahrefs' data18. For how-to queries, a short video answering the exact question, embedded on the page, gives you two chances. If you sell products, keep Merchant Center current; if you serve a location, complete your Business Profile. Google names both as helping products and services appear in AI responses2, and its 2025 guidance on AI search asked for high-quality images and video alongside current Merchant Center and Business Profile data7.

Done when: your top how-to queries have a matching video, and product or local data is complete in Merchant Center or Business Profile where it applies.

7. Get mentioned on pages that rank for the sub-questions

In Ahrefs' study of 75,000 brands, branded web mentions correlated with AI Overview visibility at 0.664, against 0.218 for backlinks19. Correlation isn't cause, and Google warns that "seeking inauthentic 'mentions' across the web isn't as helpful as it might seem"2. So aim at real coverage on pages that already rank for your sub-questions: comparison articles, trade press, review sites, and the Reddit threads Google cites.

Done when: you have a list of the pages cited for your target queries, marked named or not named, with outreach under way on the gaps.

8. Measure in Search Console every month

Use the Generative AI performance report for impressions and the standard Performance report for clicks, as described below. Re-check your query list by hand or with a tracker on the same day each month.

Done when: a monthly report shows AI impressions by page, citation status for your query list and organic clicks, next to last month.

Worked example: an HR software blog post

This is a hypothetical company, "Crewcount," which sells HR analytics software. We haven't run these queries; the example shows the method and the decisions, not results.

The query. Crewcount's post "How to calculate employee turnover rate" ranks ninth. The AI Overview cites three other pages. Crewcount's fan-out list for the query looks like this:

how to calculate employee turnover rate
  what is a good employee turnover rate
  monthly vs annual turnover rate formula
  voluntary vs involuntary turnover
  how to calculate turnover rate in Excel
  what counts as an employee for turnover (contractors? interns?)

Its post answers the first and third, and buries the formula in paragraph four.

The before and after. The section under "How do you calculate turnover rate?" opened like this:

Before: Employee turnover is one of the most important metrics for any HR
team. Understanding it can help you retain talent and reduce costs. In this
section, we'll walk through the different ways to think about turnover.

After: Employee turnover rate = employees who left during the period, divided
by average headcount for the period, times 100. If 6 people left in March and
average headcount was 120, March turnover is 5%. Use average headcount, not
headcount on the last day, or growing teams will look better than they are.

The rewrite answers in the first sentence, gives a worked number, and adds a judgment call the cited pages don't make.

The decisions.

"What is a good employee turnover rate" has no section. Crewcount adds one, using a benchmark from a named survey with its date range, and links the source.
"How to calculate turnover rate in Excel" is cited to a YouTube video. Crewcount records a 90-second screen capture with the formula and embeds it.
If the post still isn't cited after a recrawl, Crewcount checks the snippet: a data-nosnippet wrapper on its "key formula" box would keep the best passage out.

What doesn't get you into AI Overviews

llms.txt and AI-only files

Google's guide says you don't need "new machine readable files, AI text files, markup, or Markdown," and that Google Search itself doesn't use them2. Add llms.txt for other reasons if you like; it won't change AI Overviews.

Special schema, and FAQ markup for rich results

"Structured data isn't required for generative AI search," per the same guide2, and Google says to make sure any structured data matches the visible text1. FAQ rich results are gone too: Google's Search Central updates log says the feature stopped showing on 7 May 202620. Keep schema for the rich results that still exist. Don't expect it to earn citations.

Scaled pages for every variant

Making many near-duplicate pages "primarily to manipulate rankings or generative AI responses" breaks Google's scaled content abuse policy2, and since May 2026 Google has said its spam policies also apply to generative AI responses20. One strong page per topic that covers the sub-questions beats fifty thin ones.

Blocking Google-Extended

Google-Extended controls Gemini training and Gemini app grounding. Google says it "does not impact a site's inclusion in Google Search"4. Blocking it won't remove you from AI Overviews, and allowing it won't get you in.

Can you opt out of AI Overviews?

Yes, two ways. Since 31 August 2026 every site has a Search Console setting under Settings, Search generative AI, that excludes it from AI Overviews, AI Mode and generative features in Discover; Google says opted-out sites "will not receive traffic or impressions" from those features, that the control isn't a ranking signal for the rest of Search, and that it doesn't affect AI training35. The older, finer-grained route is snippet controls: nosnippet or data-nosnippet on the content you don't want quoted1. In the UK, the competition regulator required Google in June 2026 to give publishers effective tools to keep their content out of AI features in Search21. For most businesses that want customers, opting out is the wrong call.

How do you measure AI Overview rankings?

Search Console now has a Generative AI performance report. It counts impressions, defined as "how many times links to your site were shown to a user in a generative AI feature on Google Search," for AI Overviews and AI Mode together, broken down by page, country, date and device6. It doesn't report clicks. Those stay in the standard Performance report under the Web search type, mixed with everything else1.

So report three things monthly: AI impressions by page from the new report, organic clicks for the same pages, and citation status for your query list from manual checks or a tracker. Rising AI impressions with flat clicks is normal; falling AI impressions on a page that used to be cited is the signal to act on. Our guide to measuring AI visibility has a reporting template, and the AI Mode tracker comparison covers the other half of Google's AI search. For a free first read on whether AI engines name your brand, try the Arobis AI visibility checker.

How long does it take to get cited?

Eligibility fixes work as fast as Google recrawls: a removed nosnippet or a re-enabled Search Console setting can show up within days. Content changes need a recrawl and then enough ranking strength on the sub-questions, which is weeks for pages Google already trusts and longer for new ones. Mentions on other sites move on their owners' schedules. Plan a quarter, and judge it on the query list, not on one screenshot.

The same pages often do well elsewhere. See our guides to ranking in ChatGPT, Perplexity and Claude, which retrieve from different indexes.

Frequently asked questions

How do I get my website into Google AI Overviews?

Make sure the page is indexed, eligible for a snippet, and that your site is included in the Search generative AI setting in Search Console, which is on by default. Then answer the follow-up questions Google's query fan-out searches for, each in its own section, with something the already-cited pages lack, such as original data or a worked example.

Do you need to rank in the top 10 to appear in AI Overviews?

No. In Ahrefs' March 2026 study of 863,000 SERPs, only 37.9% of cited URLs ranked in the top 10 for the query, while 31% ranked beyond position 100. Query fan-out means Google also cites pages that rank for related sub-questions. Ranking well still helps, but covering the follow-up questions matters as much.

Does schema markup help with AI Overviews?

Google says structured data isn't required for its generative AI features and there's no special schema to add. It still asks that any markup match the visible text. FAQ rich results stopped showing on 7 May 2026, so FAQ markup no longer earns a rich result either. Use schema for the rich results that remain.

How do I track AI Overview traffic in Search Console?

Open the Generative AI performance report, which counts impressions where your links appeared in AI Overviews and AI Mode, by page, country, date and device. It doesn't show clicks. Those are included in the normal Performance report under the Web search type and can't be separated from regular organic clicks there.

Can I stop Google from using my site in AI Overviews?

Yes. Search Console's Search generative AI setting, available to all sites since 31 August 2026, excludes a site from AI Overviews, AI Mode and Discover AI features within a few days, without affecting the rest of Search. For specific passages, use nosnippet or data-nosnippet. Blocking Google-Extended does not remove you from AI Overviews.

Do AI Overviews reduce clicks to websites?

On average, yes. Pew found users clicked a traditional result on 8% of visits with an AI summary against 15% without, and Ahrefs measured a 58% lower click-through rate for the top result. Being cited softens it: Seer found 2.07% organic CTR for cited brands on informational queries against 0.94% for brands not cited.

Sources

Cite this page

AI Ranked Editorial. "How to rank in Google AI Overviews and get cited." AI Ranked, October 11, 2026. https://airanked.ai/guides/how-to-rank-in-google-ai-overviews