How to Rank in Google AI Overviews: What Works in 2026

How to rank in Google AI Overviews comes down to two jobs: make the page eligible (indexed and allowed to show a snippet), then rank for the question and the related sub-questions Google's AI searches in the background. Google says no special markup, files or schema are needed, and that original, well-organized content matters most.

Key takeaways

• Eligibility is basic: the page must be indexed and snippet-eligible, and your site must be included in Search Console's Search generative AI control (the default).
• Only 37.9% of URLs cited in AI Overviews ranked in the top 10 for the exact query in Ahrefs' March 2026 study, so ranking for the sub-questions counts, not just the head term.
• Cover the real sub-questions of a topic in depth; a separate page for every keyword variation can breach Google's scaled content abuse policy.
• Skip llms.txt, "chunking", AI-only rewrites and special schema: Google says Search ignores them or doesn't need them.
• Measure with the Generative AI performance report in Search Console, rolled out to all sites as of August 31, 2026.

What are Google AI Overviews, and how is AI Mode different?

Google AI Overviews are AI-generated summaries at the top of some Google results that answer the question and link to the pages supporting the answer. Google says they help people "get to the gist of a complicated topic or question more quickly" and are shown only when its systems judge them additive to classic results, so they "often don't trigger" (Google Search Central: AI features and your website).

AI Mode is the conversational version, built for questions that need exploration, reasoning or comparisons, with follow-up questions. The two features may use different models, so the same question can produce different links in each, but Google applies the same eligibility rules and SEO fundamentals to both. As of October 2026, AI Overviews are available in a long list of countries and languages, including the US, the UK, Egypt, Saudi Arabia, the UAE and Arabic (Google Search Help).

How does Google choose which pages to cite in AI Overviews?

Google picks AI Overview sources with its normal ranking systems. Its May 2026 guide to optimizing for generative AI features says they are "rooted in our core Search ranking and quality systems" and rely on two techniques (Google's AI optimization guide):

• Retrieval-augmented generation (grounding): core ranking systems retrieve relevant, up-to-date pages from the Search index, the model reads the specific information on them, and the response links to the pages that support it.
• Query fan-out: the model issues "a set of concurrent, related queries" to fetch more results. Google's example: for "how to fix a lawn that's full of weeds", fan-out queries might include "best herbicides for lawns", "remove weeds without chemicals" and "how to prevent weeds in lawn".

Query fan-out has a practical consequence for AI Overviews SEO: a page can be cited for a query it doesn't rank for, because it ranks for one of the hidden sub-queries the AI ran while building the answer.

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

No. Most AI Overview citations now come from pages outside the top 10. Ahrefs analysed 863,000 keyword SERPs and 4 million AI Overview URLs and checked where each cited URL ranked for the same keyword (Ahrefs, March 2026):

Three details matter:

1. The share fell fast. Ahrefs writes that Google now selects far fewer pages straight from the original SERP: about 76% in July 2025 versus about 38% now. Ahrefs also improved its parsing in between, so the drop isn't like-for-like.
2. Fan-out is the likely cause, not a proven one. The author suggests AI Overviews now rely more on "the sources showing up in fan out query SERPs" and notes that AI Overviews are now powered by Gemini 3.
3. Video is part of the mix. Of the cited pages that didn't rank in the top 100, 18.2% were YouTube URLs, or 5.6% of all AI Overview citations in the dataset.

The practical reading: the top 10 is still the largest single group, but a newer site that can't reach page one for a competitive head term can still be cited by ranking for the narrower questions the AI asks. This is vendor data, not peer-reviewed, and the shares will drift as Google changes its models.

What we found: what the top-ranking pages on this topic cover

We ran "google ai overviews seo" through RankTera's KW Research on October 15, 2026, using US Google results. The five top-ranking pages it analysed (semrush.com, search.google, finch.com, seo.com and searchengineland.com) averaged 2,376 words, and RankTera extracted 83 keywords from the text under their headings. The most used were "AI Overviews" (29 times per page on average), "AI Overview" (7), "search results" (4) and "structured data" (3), with "Google Search" and "brand mentions" among the terms at 2.

Our same-day SERP analysis of the six top Google results for "how to rank in google ai overviews" found lengths from 914 words (Google's own Search Central post, in first place) to 4,562 words. Two of the six had an FAQ section and one used tables.

Two points stand out. Structured data is one of the most repeated terms on the ranking pages, while Google's guide says no special schema is needed for its AI features. And the first-placed page is also the shortest, which fits Google's statement that "there's no ideal page length".

How to rank in Google AI Overviews: 8 steps

Step 1 is a gate; steps 2 to 8 decide whether your page gets picked.

1. Make sure the page is eligible

To be eligible 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; Google says there are "no additional technical requirements". Google's newer AI optimization guide adds one condition: the site must be included in the Search generative AI control in Search Console (Settings > Search generative AI). Inclusion is the default, so you only need to confirm nobody changed it (Search Console Help: Search generative AI control). Then check that:

• robots.txt, your CDN and your hosting allow Googlebot to crawl the page;
• URL Inspection in Search Console shows the page as indexed;
• the main content carries no nosnippet, data-nosnippet or restrictive max-snippet rules, which limit what AI features can show;
• important content is in text, not only in images or unrendered JavaScript.

2. Target questions where AI Overviews actually appear

AI Overviews appear where Google judges an AI summary adds something beyond classic results. In practice that usually means informational queries: how-to and "why" questions, comparisons and multi-part problems that need several sources. Before you invest in a page, search the query in the target country and note whether an AI Overview appears and what it cites. Treat one check as a snapshot; overviews change over time.

3. Map the fan-out sub-questions before you write

Because the AI searches related sub-questions, list them before you outline:

• the People Also Ask questions and related searches for the query;
• the headings of the pages that rank and of the pages the AI Overview already cites;
• the questions customers ask in sales calls and support tickets, which competitors can't copy.

Then group the list by intent. Questions a reader of your page would naturally ask next belong on the same page; a question with a different reader or task deserves its own page.

4. Answer each sub-question in its own clearly headed section

Write a heading that names the sub-question, answer it in the first one or two sentences, then add detail, steps and examples. This is ordinary good structure, not a special AI format: Google says readers appreciate pages "organized by paragraphs and sections, along with headings that provide a clear structure". Make each section hold up when quoted alone: repeat the subject noun instead of opening with "it", keep definitions and numbers in visible text, and use a table when you compare options.

5. Add something the other sources don't have

Google calls this non-commodity content and says it will likely influence your presence in generative AI search "in the long run more than any of the other suggestions" in its guide. Its example: "7 Tips for First-Time Homebuyers" is commodity content built on common knowledge, while "Why We Waived the Inspection & Saved Money: A Look Inside the Sewer Line" offers an experienced take. A page that only restates other sources gives the model nothing new to cite. Most businesses can add their own data (a survey, test results, real prices), a worked example from their work, a checklist built from experience, or original photos and video.

6. Build a topic cluster, not a page per keyword variation

Fan-out rewards depth on a topic, but Google draws a firm line. Its guide warns that creating separate content "for every possible variation of how people might search", including fan-out queries, primarily to manipulate rankings or AI responses, violates the scaled content abuse policy. The spam policies now name attempts to "manipulate generative AI responses in Google Search" as spam, and the guide adds that you don't need to capture every long-tail variation, because AI systems understand synonyms.

The safe model is one strong page per distinct intent, linked into a cluster: a pillar page covers the broad topic and links to cluster pages for sub-topics with their own intent, and each links back. Our guide to generative engine optimization shows the same cluster model across AI search engines.

7. Earn genuine mentions and publish video

Brands that people talk about tend to be more visible in AI Overviews. In Ahrefs' 2025 study of 75,000 brands, branded web mentions correlated with AI Overview brand visibility at 0.664 (Spearman), about double Domain Rating at 0.326 and well above backlinks at 0.218 (Ahrefs). Ahrefs notes that correlation isn't causation, and Google warns that seeking inauthentic mentions "isn't as helpful as it might seem". The version that works is real: publish data others want to reference and give expert quotes to industry writers.

Video is the second channel: YouTube URLs made up 5.6% of AI Overview citations in the Ahrefs data, and Google says its AI features "can bring in relevant images and video". If a process is easier to show than describe, film it and embed it.

8. Keep the classic SEO basics in place

Google lists the fundamentals that "continue to be worthwhile" for AI features: crawl access, internal links, a good page experience, important content in text, quality images and video, structured data that matches the visible text, and current Merchant Center and Business Profile details for shops and local businesses. In Google's words, optimizing for generative AI search "is optimizing for the search experience, and thus still SEO". For where the disciplines overlap and differ, see GEO vs SEO vs AEO.

What doesn't work for AI Overviews

Several popular AI Overview tactics don't help in Google. Its guide's "Mythbusting" section lists what you don't need to do, and independent data backs several points.

One caveat: Ahrefs tested pages that were already heavily cited (100+ AI Overview citations each) over 30 days before and after the change, so the schema result may not transfer to pages AI doesn't cite yet. Google's own position is clearer: schema supports rich results, not entry into AI answers.

Worked example: getting a guide cited for an espresso question

The numbers below are an illustrative example, not client data.

A small online store selling home espresso equipment has a 900-word guide, "How to descale an espresso machine". It ranks 14th for the head query, which shows an AI Overview citing four other sites.

1. Eligibility. URL Inspection shows the page as indexed, with no snippet restrictions, and the Search generative AI control is on its default (include).
2. Fan-out map. From People Also Ask and the cited pages' headings, the team lists six sub-questions: how often to descale, whether vinegar or citric acid works, descaling versus backflushing, signs a machine needs descaling, how water hardness changes the schedule, and descaling a dual-boiler machine. The guide answers only the first and fourth.
3. Page or cluster. The first five questions are what the same reader asks next, so they become new sections. The dual-boiler question serves a different reader, so it gets its own page, linked from the guide.
4. Non-commodity value. Repair technicians log the water hardness of 40 customers whose machines came in with scale damage, publish a "descaling interval by water hardness" table, and add a two-minute video.
5. Measurement. The team tracks Google positions for the head and sub-question keywords and watches the Generative AI performance report.

Illustrative result after eight weeks:

In this example, the page goes from no AI feature impressions in the report to about 1,100 a month. It now ranks for several sub-questions an overview on this topic is likely to fan out to, and it carries a table no other page has. None of it needed special markup.

How to track AI Overview visibility

AI Overview tracking starts in Google Search Console, the only source of Google's own numbers. As of October 2026, use three views:

1. Generative AI performance report. It shows impressions (how many times links to your site were shown in AI Overviews and AI Mode) by page, country, date or device, with a filter for text-based versus multimodal searches such as Google Lens. Google rolled it out to all websites worldwide as of August 31, 2026; if you can't see it, your site usually hasn't had enough AI feature impressions yet, or it has been excluded with the Search generative AI control (Search Console Help: Generative AI performance report).
2. Performance report, "Web" search type. AI feature traffic is included here. A click on a link in an AI Overview counts as a click, and an AI Overview occupies a single position, with all its links assigned that position (how Search Console counts AI Overviews).
3. Your analytics tool for what happens after the click. Google says clicks from results pages with AI Overviews are "higher quality", meaning visitors are more likely to spend more time on the site.

The Generative AI performance report has no query dimension, so it doesn't show which questions earned the impressions. Fill that gap by tracking Google positions for every sub-question keyword in your map. Google also says no third-party tool has access to its ranking or AI systems, so "be wary" of tools claiming internal metrics. For a plan covering Google, ChatGPT and Gemini, see how to track AI search visibility.

How to do this in RankTera

RankTera covers the research and tracking around the overview, not the overview itself: it doesn't detect AI Overviews.

• Map the sub-topics: run your keyword through KW Research with the target country. RankTera reads the top Google pages (you pick up to five in Select SERP Results), extracts the phrases they use under their H2–H6 headings with their average count (AVG COUNT), and shows the pages' average word count and heading structure: a checklist of sub-topics to answer.
• Track the fan-out keywords: add the head query and each sub-question keyword to the Rank Tracker, group them in a collection, and RankTera checks their Google positions (top 50, desktop) daily or weekly, depending on your plan.
• Check other AI engines: for every tracked keyword, RankTera asks ChatGPT and Gemini once a week to search the web and records whether your domain is among the sources they cite. That is a separate signal from Google AI Overviews, built on the same keyword list.

Plans and keyword limits are on the RankTera pricing page.

How to opt out of or limit AI Overviews

Site owners can opt out entirely: set the Search generative AI control in Search Console to "Exclude my site's links and content from Search generative AI features". Your links and content leave AI Overviews, AI Mode and generative AI features in Discover, along with any traffic or impressions from them. Google says the control isn't a ranking or inclusion signal for the rest of Search, doesn't affect AI training (Google-Extended handles that) and generally takes a few days to apply. Weigh the trade first: content from other sites will still appear in those features, and Google notes "it may appear similar to yours".

For narrower limits, use the preview controls: nosnippet, data-nosnippet, max-snippet and noindex. They also apply to your regular search results (noindex removes the page from Search entirely), so prefer data-nosnippet on specific passages over restricting whole pages. Blocking Google-Extended in robots.txt won't keep your site out of AI Overviews: Google says the token "does not impact a site's inclusion in Google Search nor is it used as a ranking signal" (Google's crawler documentation).

Searchers can't turn AI Overviews off, but they can select the Web filter after a search, which shows only text-based links (Google Search Help).

Frequently asked questions

Can Google rank AI-generated content?

Yes, if it is accurate and adds value. Google's guidance says generative AI can help research a topic and add structure to original content, but using it "to generate many pages without adding value for users" may violate the scaled content abuse policy. Google also asks you to fact-check AI output, including titles and meta descriptions, because models can produce inaccuracies (Google's guidance on AI-generated content).

How do you rank higher on AI searches like ChatGPT and Gemini?

The fundamentals carry over: be crawlable by each engine's search crawler, rank for the question and its sub-questions, and publish content with original value. OpenAI, for example, recommends allowing OAI-SearchBot in robots.txt, because sites that block it are not shown in ChatGPT search answers (OpenAI crawler documentation). Our ChatGPT SEO guide covers the rest.

How long does it take to appear in an AI Overview?

There is no fixed timeline. A page has to be crawled and indexed, then rank for the query or a fan-out sub-query. Google says crawling can take "anywhere from several days to several months", depending on how often its systems decide a page needs refreshing. After a major update, request a recrawl in Search Console and watch the Generative AI performance report.

Can RankTera track AI Overviews?

No. RankTera doesn't detect AI Overviews or other SERP features. It tracks your Google positions (top 50, desktop) for every keyword and, once a week, checks whether ChatGPT and Gemini cite your domain among their sources for that keyword. For AI Overview impressions, use Search Console's Generative AI performance report.

Is optimizing for AI Mode different from AI Overviews?

Barely. Google applies the same eligibility rules and SEO best practices to both, and both may use query fan-out. They can use different models, so the links they cite often differ. One practical difference: Google treats a follow-up question in AI Mode as a new query with its own response and links, so pages that answer the likely next question get another chance to be cited.

Track the queries behind the overview

Learning how to rank in Google AI Overviews doesn't require a separate playbook: make the page eligible, rank for the sub-questions, add something original and measure in Search Console. RankTera handles the work around it: sub-topic research from the top-ranking pages, Google positions for every sub-question keyword, and weekly ChatGPT and Gemini citation checks in the same table. Start your 14-day free tier — no card needed; nothing is charged when it ends.