Ranking in Google AI Overviews means getting your page cited as one of the links inside Google’s AI summary. Google says there are no extra requirements and no special optimizations for it. Your page must be indexed and eligible to show in Search with a snippet. After that, the work is ordinary SEO done well: crawlable pages, a direct answer near the top, real expertise, and coverage of the follow-up questions people ask next.
That’s the honest answer, and I know it’s less exciting than the “AI Overview hacks” circulating on LinkedIn. But it’s what Google documents, and the independent data backs it up. Below I’ll show you what Google actually says, how citations seem to get chosen, and the decision path I follow when a client asks me why a competitor gets cited and they don’t.
What Does Google Say It Takes to Appear in AI Overviews?
Google’s page on AI features and your website is blunt. It says “there are no additional requirements to appear in AI Overviews or AI Mode, nor other special optimizations necessary.” To be eligible, a page has to be indexed and eligible to show in Google Search with a snippet, meeting the normal Search technical requirements.
The same page lists the SEO basics that still apply:
- Allow crawling in robots.txt and in your CDN or hosting setup.
- Make the page findable through internal links.
- Give visitors a good page experience.
- Keep important content in text form, supported by images and video where they help.
- Make sure any structured data matches the visible text.
- Keep Merchant Center and Business Profile information current, if you have them.
In 2026 Google added a longer guide to optimizing for generative AI features, last updated July 10, 2026. Its mythbusting section matters more than its tips. It says you don’t need llms.txt or other AI text files, that there’s “no requirement to break your content into tiny pieces,” and that “you don’t need to write in a specific way just for generative AI search.”
So anyone selling a secret AI Overview format is selling something Google has explicitly said doesn’t exist. I’d rather you spend that budget on content nobody else can write.
How Does Google Choose Which Pages to Cite?
Google doesn’t publish a citation formula. What it does document is the mechanism. Both AI Overviews and AI Mode may use “query fan-out,” which Google describes as issuing multiple related searches across subtopics and data sources. Its optimization guide calls the retrieval step grounding, and says it relies on Google’s core Search ranking systems.
In plain terms: the model breaks your query into related sub-questions, runs searches for them, and cites pages that rank well for those pieces. That’s why a page ranking ninth for the main query, but first for a sub-question, can still earn the link.
What the Independent Data Shows
Two Ahrefs studies, both by Louise Linehan, show how much that fan-out has changed things. They measure correlation, not Google’s rules, so read them as evidence, not law.
| Study | Sample | Cited pages from the top 10 | From positions 11 to 100 | Outside the top 100 |
|---|---|---|---|---|
| Ahrefs, July 21, 2025 | 1.9 million citations, 1 million AI Overviews | 76.1% | 9.5% | 14.4% |
| Ahrefs, March 2, 2026 | 863,000 keyword SERPs, 4 million AI Overview URLs (January 2026 data) | 37.9% | 31.2% | 31.0% |
The March 2026 update found that just under 38% of cited URLs ranked in the top 10 for the original query. Ahrefs links the drop to query fan-out becoming more prominent. It also found YouTube URLs made up 5.6% of all cited URLs in the dataset.
One caveat: the 2026 study counted SERP “blocks,” which include ads, snippets and People Also Ask, not just blue links. Ahrefs says a blue-links-only check showed a similar pattern. My takeaway is simple. Ranking for the head term still helps, but ranking for the questions around it now matters almost as much.
How to Rank in Google AI Overviews: My Decision Path
When a page isn’t showing up in AI Overviews, I don’t start by rewriting it. I start with eligibility, because no amount of writing fixes a page Google can’t use. Here’s the order.
- Check whether an AI Overview even shows. Google says AI Overviews appear only when its systems decide they’re additive, so they “often don’t trigger.” No Overview, nothing to win.
- Confirm the page is indexed and snippet-eligible. Use URL Inspection in Search Console. Look for
noindex,nosnippetor a tightmax-snippetvalue. - Check the answer. Can someone find a direct, quotable answer to the query in the first 100 to 150 words?
- Check the follow-ups. Does the page cover the next three or four questions a searcher would ask?
- Check freshness and authorship. Are the facts current, dated and backed by someone who clearly knows the topic?
If step 2 fails, then stop and fix the technical problem first. My technical SEO for AI search checklist covers crawler access, rendering and canonicals in detail, so I won’t repeat it here.
If steps 1 and 2 pass but you’re not cited, then the gap is usually content: a buried answer or missing sub-topics. Work through the branches below.
If you’re already cited, then your job is defense. Keep the page fresh and watch whether the citation holds after each core update.
Branch 1: When the Answer Is Buried
This is the most common problem I see on small-business blogs. The post opens with three paragraphs of story before it answers anything. Readers skim past that, and so does any system looking for a clean passage to quote.
The fix is to put a two- to four-sentence answer right under the heading that matches the question. Use a clear definitional verb (“X is”, “X means”) and plain words. Then expand with the nuance.
This isn’t chunking for AI. Google says you don’t need tiny pieces, and I agree. It’s just writing the way a good editor would, which happens to make your best sentence easy to lift.
Branch 2: When You Answer the Main Question but Miss the Follow-Ups
Fan-out means Google is searching for the questions behind the question. If someone asks “how to rank in Google AI Overviews,” the sub-queries probably include things like “do I need schema,” “does ranking matter” and “how do I measure it.” A page that only answers the head query competes on one search out of many.
My method is low-tech. I read the People Also Ask boxes and the “AI Mode” follow-up suggestions for the query, then list the questions my page doesn’t answer yet. Each one gets a short section or an FAQ answer, as long as it genuinely belongs on the page.
The boundary matters, though. If a follow-up deserves its own article, I give it one sentence here and link out. Cramming everything into one URL creates keyword overlap across your own site, and that’s a different problem.
Branch 3: When Your Content Is Commodity
Google’s optimization guide uses the example “7 Tips for First-Time Homebuyers” for what it calls commodity content, based on common knowledge. What it rewards instead is “unique expert or experienced takes.”
That’s the hardest branch, because it can’t be fixed with formatting. It takes original data, real examples, a clear opinion, or a process you’ve actually run. When I audit content, this is where I usually tell clients the truth: the page is correct, but it says nothing a model couldn’t already produce on its own.
Edge Cases That Change the Advice
A few situations bend the usual path, and I’ve learned to check for them early.
- Local businesses: Google’s AI features page tells you to keep Business Profile information current. For “near me” style queries, your profile can matter as much as your website.
- Ecommerce: the same page points to Merchant Center, and Google’s optimization guide names it as a way to increase visibility in AI responses. Keep prices and stock accurate in your feed.
- Video-friendly topics: with YouTube at 5.6% of cited URLs in the Ahrefs 2026 data, a short explainer video can earn a citation your article doesn’t.
- Pages you don’t want quoted: Google names
nosnippet,data-nosnippet,max-snippetandnoindexas the controls. They also shrink your normal snippets, so use them surgically.
What about schema? It helps Google understand pages and qualify for rich results, but Google says it’s not required for AI features. I cover that in my guide to schema markup for AI search.
How Is AI Mode Different From AI Overviews?
They’re related but not identical. Google says AI Mode suits queries “where further exploration, reasoning, or complex comparisons are needed,” while AI Overviews give a quick gist on the regular results page.
The bigger point for site owners: Google says the two “use different models and techniques, so the set of responses and links they show will vary.” So being cited in one doesn’t guarantee the other. The eligibility rules are the same, though, and so is the advice above.
If you also care about ChatGPT or Perplexity, the mechanics differ enough that I wrote separate guides: how to get cited by ChatGPT and Perplexity SEO. For the bigger strategy question, see GEO vs SEO.
How Do You Measure AI Overview Visibility?
Start with Search Console. Google’s help page for the Generative AI performance report says it rolled out to all websites worldwide as of August 31, 2026.
The report counts impressions, meaning how many times links to your site were shown in a generative AI feature on Google Search. You can break that down by page, country, device and date. Several results from your site in one feature count as a single impression.
Know its limits before you report on it:
- It shows impressions only, with no clicks and no queries.
- Low-traffic sites may not see data at all.
- Clicks from AI features are mixed into the normal “Web” search type of the Performance report.
I check it page by page after each update. If a page’s AI impressions climb while its classic rankings stay flat, the fan-out effect is probably working in your favor. Whether those impressions turn into visits is a separate question, which I cover in AI Overviews traffic impact.
Why Did My AI Overview Citation Disappear?
Citations move a lot. Google’s systems regenerate answers, fan-out queries shift, and core updates reshuffle the rankings that grounding relies on. A lost citation right after a core update is often a ranking change underneath, not an AI-specific penalty.
So I compare the timing against Google’s update history first. My breakdown of Google algorithm updates has every confirmed date, which makes that check quick.
Can Anyone Guarantee an AI Overview Ranking?
No. Google doesn’t sell placement, doesn’t publish the citation formula, and doesn’t show an Overview for every query. Anyone promising a guaranteed spot is guessing, or worse.
What I can do is remove the obstacles and strengthen the content. That’s the core of our AI search optimization service. If you’d like a second pair of eyes first, request a free SEO audit and I’ll tell you which branch your pages are stuck on.
Frequently Asked Questions
Do I Need to Rank in the Top 10 to Appear in AI Overviews?
It helps, but it’s not required. Ahrefs found 76.1% of cited pages ranked in the top 10 in its July 2025 study. In its March 2026 update, that fell to 37.9%, with about 31% coming from beyond the top 100. Ranking for related sub-questions now matters too.
Does llms.txt Help With AI Overviews?
No. Google’s optimization guide says you don’t need new machine-readable files, AI text files, markup or Markdown to appear in its generative AI features. Google Search ignores them, so spend that time on content and crawlability instead.
How Long Does It Take to Show Up in an AI Overview?
There’s no set timeline. Once a page is indexed and snippet-eligible, it can be cited whenever an Overview triggers for a related query. In practice, it moves with your rankings for the main query and its sub-questions.
Can I Stop Google From Using My Content in AI Overviews?
Yes, partly. Google lists nosnippet, data-nosnippet, max-snippet and noindex as controls. They also limit your regular search snippets, so I only recommend them for sensitive sections, not whole sites.
Last updated: October 2026 by Mizanur Rahman
