How to Appear in Google AI Overviews and AI Mode

Google documents the rules for its AI surfaces, and they are narrower than the advice being sold around them. There is no opt-in, no special markup, and one very common mistake that costs snippets for nothing.

Short answer

There is no opt-in for Google AI Overviews. Google states that a page must be indexed and eligible to be shown in Google Search with a snippet to appear as a supporting link in its AI experiences — the same foundations as classic SEO, with no extra technical requirement. So the work is ordinary and unglamorous: be indexable, be snippet-eligible, answer one question per page in its opening lines, and cover the sub-questions a reader would ask next on their own pages.

The eligibility rule, stated plainly

Most of the confusion around AI Overviews comes from treating it as a new channel with its own admission process. It is not. Google's own documentation on its AI features is direct about the requirement: a page must be indexed and eligible to be shown in Google Search with a snippet. That is the gate. There is no separate index, no submission form, and no markup that admits you.

This is genuinely good news, because it means the diagnostic is one you can already run. If a page is not indexed, nothing downstream matters. If it is indexed but you have suppressed snippets on it, you have quietly excluded yourself. If it is indexed and snippet-eligible and still never cited, the problem is the content, not a technical door you failed to unlock.

Anything sold to you as a dedicated AI Overviews optimization — a special file, a schema type, a submission service — is describing a mechanism Google does not document. Ask to see the source before you pay for it.

AI Overviews and AI Mode are not the same surface

Google now runs two distinct AI surfaces, and conflating them leads to work aimed at the wrong one.

  • AI Overviews — a generated summary above the classic blue links, triggered on a subset of queries. It leans on pages that already perform in Search for the query and its close variants.
  • AI Mode — a conversational surface where a single question is broken into several searches, run in parallel, and synthesized into one answer. It rewards breadth of coverage across related questions rather than one strong page.

The distinction has a practical consequence for how you plan content. To be present in AI Overviews you mostly need to be a strong, snippet-eligible result for the head question. To be present in AI Mode you need to hold territory across the cluster of sub-questions that surround it — which is a content planning decision, not a technical one.

The Google-Extended mistake

Here is the error we see most often, and it is expensive because it costs something real while achieving nothing.

A business decides it does not want its content feeding Google's AI. Someone adds Google-Extended to robots.txt, and the team believes it has withdrawn from AI Overviews. It has not. Google-Extended governs whether your content helps improve Gemini models and grounds their responses. AI Overviews is part of Search, and Search eligibility is governed by ordinary indexing rules.

The control that does apply to what Search displays is the snippet family — nosnippet, data-nosnippet and max-snippet. Those work, and they cut your ordinary search snippets at the same time. There is no lever that removes you from AI Overviews while leaving your normal result intact. Anyone who tells you otherwise has not read the documentation.

The same class of confusion affects the assistants, where training crawlers and retrieval crawlers are separate decisions with separate consequences. We took that apart in getting cited by ChatGPT and Perplexity.

Why one page rarely wins a decomposed question

When a conversational surface answers "which automation platform should a 40-person company choose", it does not run that string as a single search. It decomposes it: what platforms exist, what they cost to run, what mid-sized companies typically pick, what the migration burden looks like. Then it assembles.

A single page trying to answer all four does each one shallowly and gets beaten on every sub-question by a page dedicated to it. The counter-move is a cluster: one page per genuine sub-question, each answering its question outright near the top, linked to each other so a reader — and a crawler — can move between them.

This is also why comparison content earns its place. When the sub-question is "X versus Y", a page built to answer exactly that is hard to beat with a general overview. Our tool comparisons exist for that reason rather than for their own sake.

What happens to your traffic

Be careful with confident numbers here. Published estimates of click loss vary enormously depending on who measured, which queries they sampled and how they defined a click, and most of the alarming figures in circulation cannot be traced back to a primary source. What is not in dispute is the direction and the shape of the change.

Queries with a single factual answer lose their click, because the answer is now on the results page. Queries where the reader needs to compare options, check credibility or make a purchase still send visits, and those visits tend to arrive later in the decision — better qualified, fewer in number.

The strategic implication is uncomfortable but simple: content whose entire value was answering a one-sentence question is now a poor investment. Content that helps someone decide something is worth more than it was.

The work that actually moves this

Nothing on this list is exotic. That is the point.

  • Confirm the page is indexed, and that no directive is suppressing its snippet.
  • Serve the substance server-side. A passage that only exists after hydration may not be read.
  • Put the answer in the first two sentences under each heading, then justify it.
  • Build the cluster around the head question rather than one long page, and link the cluster together.
  • Keep specifics — numbers, dates, names — in the same sentence as the claim, so an extracted passage still makes sense on its own.
  • Keep publication dates honest. Bumping dates without changing content is detectable and counterproductive.

If you want a sense of how the same content earns citations from the assistants rather than from Google, the mechanics diverge sharply — that is covered in our piece on what generative engine optimization actually is.

Frequently asked questions

Is there a way to opt into AI Overviews?

There is no separate submission, no dedicated markup and no setting in Search Console. Google states that to be shown as a supporting link in its AI experiences, a page needs to be indexed and eligible to appear in Google Search with a snippet. Eligibility is inherited from ordinary Search, which is why the honest answer to "how do we get into AI Overviews" is usually "fix the same things you would have fixed anyway".

Does blocking Google-Extended remove me from AI Overviews?

No, and this is the most expensive misunderstanding in the area. Google-Extended controls whether your content helps improve Gemini models and grounds their responses. It does not govern AI Overviews. The lever that applies to what Search displays is snippet control — nosnippet, data-nosnippet and max-snippet — and those directives shrink your ordinary search snippets at the same time. There is no way to leave AI Overviews while keeping full snippets everywhere else.

Do AI Overviews reduce clicks to my site?

The click behavior changes shape rather than simply disappearing. Questions with a short factual answer get resolved in the overview and lose their click. Questions where the reader needs to compare, verify or buy still send traffic, and the visit tends to arrive further along in the decision. The practical response is to stop competing for questions that can be answered in one sentence, and to write the pages a reader opens after the overview.

Does structured data get me into AI Overviews?

It helps a machine understand what your page claims, which reduces ambiguity, but it is not a trigger. No schema type forces inclusion. Structured data is worth doing because it is cheap, standardized and useful in several places at once — not because it is a route into an AI answer.

If you would rather have someone run the audit and the content plan than assemble it yourself, that is what we do. Book a call and we will tell you where you currently stand.

LYVIA

LYVIA Team

AI automation and SEO/GEO visibility

LYVIA builds custom AI tools for companies of 10 to 100 people, and gets them found on Google and inside AI answers.

Free offer

Get your free AI audit
in 30 minutes

A LYVIA expert reviews your workflows, pinpoints the 3 highest-ROI AI opportunities, and hands you a concrete roadmap. No commitment, no jargon.

  • Full diagnostic of your business processes
  • Automatable quick wins, identified
  • A personalized roadmap you keep
Book my free audit

30 min · Free · No commitment