AI Overviews SEO: What Google Says You Need to Do

    ·
    AI Overviews SEO: What Google Says You Need to Do

    TL;DR

    • Google says there are no special requirements or tricks to appear in AI Overviews.
    • Your page must be indexed and eligible to show with a snippet. That's the entry ticket.
    • AI features can run many related searches at once, so cover the subtopics around your question.
    • Measure it in Search Console. Skip the AI text files and special schema.

    AI Overviews SEO is the work of making your pages eligible and useful for the AI summaries at the top of Google results. Google says there are no extra requirements and no special optimisations for it. Your page needs to be indexed and able to show with a snippet. After that, it's normal good SEO.

    This post is one part of our AI search guide. It covers the Google side only.

    What does Google say you need to do?

    Google's page on AI features in Search is blunt. It says there are no additional requirements to appear in AI Overviews or AI Mode, and no special optimisations. It also says you don't need new machine-readable files, AI text files or special schema.

    To be a supporting link, a page must be indexed and eligible to show with a snippet. That's the bar.

    So the question changes. It's not "how do I trick the summary?" It's "is my page one Google can use, and is it worth using?"

    What is query fan-out?

    Google says AI Overviews and AI Mode may use "query fan-out". That means the system runs multiple related searches across subtopics to build an answer.

    In plain words: one question in, several searches out. The answer pulls from pages that cover the parts of the question, not just the headline.

    What that means for you:

    • Cover the subtopics around your main question on one page, or link them tightly.
    • Answer the obvious follow-up questions.
    • Use clear headings that match real questions.

    A page that only repeats the headline has less to offer than a page that answers what comes next.

    Example (made up to show the pattern)

    Say your page answers "how do I choose a CRM for a five-person agency?". A thin page gives a list of tools. A page that fits query fan-out also answers the searches around it:

    • What does a CRM cost per user?
    • What should a small agency track first?
    • How long does it take to set up?
    • When is a spreadsheet enough?

    Each of those can be an H2 with a short, direct answer. The headline question gets answered, and so do the questions next to it.

    Do this for your own topics. Open each cited page and write down what they share: a direct answer near the top, clear headings, a named source. Your own list will beat any generic checklist.

    Do people still click when an AI summary appears?

    Less often. Pew Research Center studied 68,879 Google searches made by US adults who shared their browsing data (March 2025). Here's what it found:

    • 18% of searches produced an AI summary.
    • People clicked a traditional result in 8% of visits that had a summary. That compares with 15% for visits without one.
    • People clicked a link inside the summary in 1% of visits.

    That's one study of US users at one point in time. It doesn't tell you what will happen on your queries. But it does say the click you used to get may be harder to win. Plan for being useful even when you're not clicked.

    What should you check on your pages?

    Google's advice on succeeding in AI search gives you the list. Work through it page by page.

    Access

    • Google can reach the page. It isn't blocked, it returns HTTP 200 and it's indexable.

    Content

    • The page has unique, non-commodity content. Something other pages don't say.
    • It's easy to read and has a good page experience.

    Previews

    • Your snippet settings don't stop Google showing the page. Google's snippet controls include nosnippet, max-snippet and data-nosnippet. A tight setting can limit what's shown.

    Structured data

    • Any structured data matches the visible text on the page.

    Our spoke on schema markup in the AI era covers the structured data side in more depth.

    What should you skip?

    Google is clear on this one. It says you don't need special AI files or new schema. So don't spend a week on those for Google's sake.

    That includes llms.txt. Other tools may read it, and we cover that in our llms.txt guide. But for Google's AI Overviews, Google says no AI text files are needed.

    Spend that week on the thing Google does reward: a page with something real to say.

    How do you measure AI Overviews traffic?

    Google says clicks from AI features are counted in Search Console's Performance report, under the Web search type. Use the Performance report and its search type filter.

    A simple routine:

    1. Pick 5 to 10 queries you care about.
    2. Check their impressions and clicks each month.
    3. Search them yourself and note whether an AI summary shows, and which pages it links to.
    4. Write it down. Watch for changes.

    For the bigger picture, read how SEO and GEO fit together.

    Frequently Asked Questions

    Is there a special way to optimise for AI Overviews?

    Google says no. There are no extra requirements. Your page must be indexed and eligible to show with a snippet.

    Do I need schema or an llms.txt file?

    Google says it doesn't need special schema or AI text files for these features. If you use structured data, make it match what's on the page.

    Can I opt out of showing in AI summaries?

    Google's snippet controls (nosnippet, max-snippet and data-nosnippet) let you limit how your content is previewed. Read what each one does before you change anything.

    Where do I see the traffic?

    In Search Console's Performance report, under the Web search type.

    Your next step

    Pick five queries and search them today. Note which pages the AI summary links to and what those pages cover that yours doesn't. Then read our AI search content strategy to plan the gaps.