How to Appear in Google AI Overviews (7 Steps)

Most advice on how to appear in Google AI Overviews starts with tricks.

Special files. Magic schema. "AI-friendly" rewrites of every page.

Google's own documentation says you need none of that.

What you need is simpler, and harder: a page Google can index and quote, and a passage that answers a specific question better than anyone else on the results page.

In this guide you will learn:

• What Google has confirmed about how AI Overviews pick their sources
• The seven steps I follow when a page should be cited and is not
• How to measure progress with Search Console's generative AI report and a query list

Every statement about how AI Overviews work below comes from Google's own documentation, which I checked in October 2026. Anything from my own client work is labelled as such.

What Google Says About AI Overviews

Google's page on AI features and your website is short, and it kills most of the myths.

Here's what it confirms:

Google also says that even a page meeting every requirement is not guaranteed to be shown.

That last point matters.

Nobody can promise you a citation.

What you can do is remove every reason Google has to skip you.

Step 1: Confirm the Page Is Indexed and Snippet-Eligible

Start here, because nothing else counts until this passes.

Open Search Console and run the URL through the URL Inspection tool.

You want two answers: the page is indexed, and Google can fetch it.

Then check the page for snippet controls.

Google's robots meta tag documentation says a nosnippet rule prevents the content being used as a direct input for AI Overviews and AI Mode.

A max-snippet value limits how much of it can be used.

So look for:

• A nosnippet or noindex robots meta tag left over from staging
• A tiny max-snippet value set by a plugin
• A data-nosnippet attribute wrapped around the section with your best answer
• Robots.txt rules or CDN bot protection that block Googlebot
• The Search generative AI setting in Search Console switched to exclude

That last one is newer. Under Settings > Search generative AI, a property is set to include by default. Switch it to exclude and Google hides your links and content from AI Overviews and AI Mode, without changing how you rank in regular Search. Check nobody flipped it.

I find at least one of these on a surprising number of audits.

In one project, a whole batch of new pages carried an instruction telling search engines to stay away.

You can read how that played out in the language services case study.

If the problem is wider than a few pages, it is a technical SEO job first.

Step 2: Find the Queries That Already Show AI Overviews

Not every query triggers an AI Overview.

Google only shows one when it decides it adds something to the normal results.

So before you rewrite anything, find out where Overviews actually appear in your topic.

Here's how I do it:

1. List 30 to 50 questions your buyers ask, in their words.
2. Search each one in a clean browser session, in the right country.
3. Note whether an AI Overview appears.
4. Record every source it links to.

You will see patterns fast.

In my own checks, definition and "how does it work" questions trigger Overviews often. Branded and transactional searches trigger them less often.

The sources tell you even more.

If the same three sites are cited across your questions, those are the pages you are competing with.

Answers change by user, location and session, so record the date and treat one run as a sample.

If you want this done at scale, my guide on tracking brand visibility in AI search covers the setup.

Step 3: Map the Sub-Questions Behind Each Query

This is the step most guides skip.

Google says AI Overviews and AI Mode may use "query fan-out". That means issuing several related searches across subtopics, then pulling supporting pages from each.

In other words: the page cited for a sub-question can end up in the answer, even if it does not rank for the main query.

So break each target query into the questions it implies.

For example:

I'm guessing at the exact sub-searches here.

Google does not publish them.

But the logic holds.

Each sub-question is a separate chance to be cited.

Understanding the intent behind each one helps you write the right answer.

My explainer on search intent goes deeper.

Step 4: Answer Each Sub-Question in a Self-Contained Passage

Google says AI Overviews pull supporting links from several related searches at once.

So each sub-question needs a section that makes sense when lifted out on its own.

The format I use:

• A heading that matches the question. "How long does planning permission take?" beats "Timescales".
• The answer in the first one or two sentences. No warm-up.
• The subject named, not implied. Write "A loft conversion usually takes...", not "It usually takes...".
• Detail after the answer. Conditions, exceptions, examples.

Here's the difference:

Weak: "Every project is different, and there are lots of factors to think about before anyone can talk about timing."

Strong: "Most loft conversions take six to ten weeks on site, plus the time needed for planning approval if your project needs it."

The figures are illustrative, but the shape is the point. The strong version can stand alone.

The weak version says nothing quotable.

This is the core of answer engine optimisation, and it applies to featured snippets as much as to AI Overviews.

Use lists for steps and tables for comparisons.

Keep important content as text, which Google lists as a best practice.

Step 5: Add What Only Your Business Knows

Here's the thing:

If your answer is the same as ten other pages, Google has no reason to cite yours.

In May 2025 Google published advice on performing well in AI search.

Its first point was to focus on unique, valuable content for people.

That means detail a competitor cannot copy:

• Your real price ranges and what moves them
• The steps you actually follow, in your order
• Mistakes you see customers make
• Photos or video of your own work
• A named author with real credentials

Google's helpful content guidance also asks you to consider "Who, How and Why" for every page. Who wrote it, how it was made, and why it exists.

Generic copy fails all three.

If your content plan is thin to begin with, that is a content strategy problem to fix before you shape individual passages.

Step 6: Keep Structured Data Honest and Business Details Current

Google is clear: there is no special schema for AI Overviews.

But it does list two related best practices.

First, make sure any structured data matches the visible text on the page.

Markup that claims things the page does not show is a risk, not a shortcut.

Second, keep your Merchant Center and Business Profile information up to date.

For most service businesses that means:

• Organization and Article schema that describe the page accurately
• A Business Profile with the same name, address, phone and services as your site
• Opening hours and service areas that match everywhere

If you need to build markup from scratch, my schema markup generator produces valid JSON-LD you can check before it goes live.

Local businesses should also review their Business Profile and local signals, because those feed the facts Google shows about you.

Step 7: Measure Progress With the Generative AI Report

Search Console now has a Generative AI performance report, which Google says reached all websites worldwide by 31 August 2026.

It counts impressions: how many times links to your site were shown in AI Overviews and AI Mode. You can break that down by page, country, device and date.

Two limits matter.

It reports impressions, not clicks. And it groups AI Overviews and AI Mode together, so you cannot tell which feature showed your link.

The clicks themselves are still counted inside the normal Performance report, under the Web search type, mixed in with ordinary results.

If the report does not appear for your site, Google says that usually means you have not had enough impressions in these features yet. A property set to exclude shows no data either.

So I combine three signals:

1. The generative AI report. Which pages earn AI impressions, and whether that changes after you rework a page.
2. A fixed query list. The same questions from Step 2, rerun on a schedule, with cited sources logged each time. This is the only way to see which feature cited you and for what.
3. Engagement in GA4. Google claims clicks from results pages with AI Overviews are higher quality, meaning people spend longer on the site. Check whether engagement and enquiries from organic search move.

Be patient with changes.

Google notes that recrawling can take anywhere from several days to several months.

A fixed indexing block often shows results faster than a content rewrite.

If you want hundreds of prompts tracked with a screenshot per answer, that runs as AI search visibility tracking.

Mistakes That Keep Pages Out of AI Overviews

Most of these come straight from audits:

• Blocking snippets by accident. A leftover nosnippet tag removes you from AI Overviews and normal snippets alike.
• Burying the answer. A 200-word intro before the reply makes the passage hard to quote.
• Paying for an llms.txt "fix". Google says no AI text files are needed.
• Mass-producing pages. Hundreds of thin AI pages give Google no reason to trust any of them.
• Chasing every query. Some queries never show an Overview. Spend effort where they do.
• Judging from one screenshot. One search is one sample. Track a list over time.

None of these need a big budget to fix.

Get Your Pages Into the Answer

To recap: get indexed and snippet-eligible, find the queries that show Overviews, map the sub-questions, and answer each one in a passage only your business could write.

Start with Step 1 on your five most important pages today.

It takes an hour and often finds the real blocker.

If you want the whole programme handled, including Bing, ChatGPT and Perplexity as well as Google, my AI search optimisation service covers access, content, corroboration and measurement in one plan.

And if you are still deciding whether this differs from the SEO you already do, read AI search vs traditional SEO next, or see how I approach getting cited across AI answers in practice.