Google AI Overviews, AI Mode and Gemini: How Visibility Works

Key facts

• Google states that pages must be indexed and eligible to be shown with a snippet to be supporting links in AI Overviews or AI Mode.
• Google says no special optimizations, AI text files or special schema.org structured data are required for AI Overviews or AI Mode.
• Traffic from AI Overviews and AI Mode is included in the Search Console Performance report under the Web search type, not as a separate report.
• Google-Extended manages use of content for training future Gemini models and for grounding; Google says it does not affect inclusion or ranking in Google Search.
• Google notes that indexing and serving are not guaranteed, even when a page meets every requirement.

What does Google require for a page to appear in AI Overviews or AI Mode?

Google requires a page to be indexed and eligible to be shown in Google Search with a snippet before it can be a supporting link in AI Overviews or AI Mode. Google's Search Central documentation says there are no additional technical requirements beyond that. Meeting the requirements makes a page eligible; Google also states that indexing and serving are never guaranteed.

This is a deliberately simple rule, and it removes a lot of noise. AI Overviews and AI Mode draw on pages that Google Search can already index and show. If a page is not indexed, or has been restricted from showing a snippet, it is outside the pool these features use.

Eligibility is the floor, not the target. Many indexed pages are eligible and only some are shown. Google describes AI Overviews and AI Mode as using query fan-out, issuing multiple related searches across subtopics and data sources, so they can surface a wider and more varied set of links than a single classic search. Pages that clearly answer specific subtopics have more opportunities to be relevant to one of those related searches.

Google also notes that AI Overviews appear only when Google determines they add value beyond classic results, so many queries never show one. And because AI Overviews and AI Mode may use different models and techniques, the links each one shows can differ for the same topic.

Does a site need special markup or AI files for Google AI Overviews?

A site does not need special markup or AI files for Google AI Overviews. Google states that there are no additional requirements or special optimizations for AI Overviews or AI Mode, no new machine-readable files or AI text files to create, and no special schema.org structured data to add. Standard SEO best practices remain what Google recommends.

This statement is useful when evaluating proposals. If a provider claims that a particular file or markup type is required to appear in AI Overviews, compare the claim with Google's own documentation. Structured data is still valuable for rich results and for stating facts clearly, and Google recommends that structured data match the visible text on the page, but it is not a special ticket into AI features.

The practices Google lists are familiar SEO work: allowing crawling in robots.txt, making content findable through internal links, providing a good page experience, putting important content in text form, and keeping structured data consistent with visible content. These are the foundation for both classic results and AI features.

• Allow Googlebot to crawl the pages you want shown.
• Link important pages internally so they can be found.
• Put key information in text, not only in images or video.
• Keep structured data accurate and consistent with what the page shows.

How can a site limit what Google shows in AI Overviews and AI Mode?

A site can limit what Google shows in AI Overviews and AI Mode with the same preview controls used for Google Search: nosnippet, data-nosnippet, max-snippet and noindex. Google's documentation points to these controls for AI features. Because AI features require snippet eligibility, a nosnippet rule also removes a page from being shown as a supporting link.

Each control has a different scope. noindex removes the page from Search entirely. nosnippet prevents any text snippet. max-snippet limits snippet length. data-nosnippet excludes specific parts of a page, which is useful for protecting a section such as pricing tables while leaving the rest of the page eligible.

These controls involve a trade-off: the stricter the restriction, the less material Google can show. A business should decide what it wants kept out of previews before applying them site-wide.

If content still appears after controls are added, Google recommends using the URL Inspection tool to confirm that Googlebot receives the correct HTML, and allowing time for recrawling. Google notes that crawling can take anywhere from several days to several months, and a recrawl can be requested.

What does Google-Extended control, and what does it not control?

Google-Extended is a standalone robots.txt product token that controls whether content Google crawls may be used to train future Gemini models and for grounding in Gemini Apps and Grounding with Google Search on Vertex AI. Google states that Google-Extended does not affect a site's inclusion in Google Search and is not used as a ranking signal in Google Search.

Google-Extended is often misunderstood as an AI Overviews switch. Google's crawler documentation describes its scope as Gemini model training and grounding, and Google's AI features guidance separately says that Googlebot directives remain the control for how a site is crawled for Search. The AI features guidance points to preview controls such as nosnippet for limiting what appears in Search, including AI features.

This gives a business three separate decisions. Googlebot governs crawling for Search. Preview controls govern what Search may show. Google-Extended governs the Gemini training and grounding uses Google documents for it. Blocking Google-Extended is a legitimate choice for a publisher with concerns about model training, but it should not be expected to change classic Search rankings.

How does Gemini relate to Google Search content?

Gemini relates to Google Search content through grounding. Google's crawler documentation describes grounding as providing content from the Google Search index to the model at prompt time, in Gemini Apps and in Grounding with Google Search on Vertex AI. A business that is well indexed in Google Search is therefore better placed to be available when Gemini grounds an answer, subject to its Google-Extended choice.

Google does not publish how Gemini decides which grounded sources to mention, and Gemini's behaviour changes over time. What is documented is the connection to the Search index and the Google-Extended control. That makes the Search foundation doubly important: it serves classic results, AI Overviews and AI Mode, and grounding in Gemini Apps.

Visibility in Gemini should be tested directly, using the same disciplined prompt method used for other AI assistants, because there is no Gemini-specific report in Search Console.

Why is SEO still the foundation for Google AI visibility?

SEO is still the foundation for Google AI visibility because AI Overviews and AI Mode draw their supporting links from pages that Google Search has indexed and can show with a snippet. Crawlability, indexing, clear content, internal linking and page experience decide whether a page is even eligible. Generative Engine Optimization builds on that foundation rather than replacing it.

In practice this means the work that improves classic Search, such as fixing crawl errors, consolidating duplicates, improving thin pages and clarifying internal linking, also widens the set of pages that AI features can draw on. The GEO layer then focuses on what makes a page useful as a supporting source: direct answers to specific subtopic questions, precise facts and clear identification of the business.

JavaScript-heavy sites need particular attention. If key content only appears after client-side rendering and Google does not render it as expected, the indexed version of the page may be thinner than what visitors see. The URL Inspection tool shows what Google actually received.

What should Arabic-language businesses in Saudi Arabia and Jordan consider?

Arabic-language businesses in Saudi Arabia and Jordan should make sure their Arabic pages are indexed in their own right, with separate URLs, native Arabic content and correct hreflang annotations linking the Arabic and English versions. Google uses hreflang to understand localized versions, so a missing or one-way annotation can mean the wrong language version is shown or the Arabic page is overlooked.

Google's documentation on localized versions says each language version must list itself and every other version, and that if two pages do not point to each other the annotations are ignored. It also notes that Google does not use hreflang or the lang attribute to detect a page's language, so the content itself must genuinely be in Arabic.

Market context matters too. A Jordanian business serving Saudi clients remotely should say so plainly rather than implying a local office it does not have. Clear statements of where a business operates help both buyers and Google's systems match the business to the right queries.

• Give Arabic and English versions separate, indexable URLs.
• Use reciprocal hreflang annotations, including a self-reference.
• Write Arabic content natively, not as a machine-literal translation.
• State service areas accurately, including remote service.

How can a business measure its visibility in Google AI features?

A business can measure Google AI visibility by combining Search Console data with direct, dated prompt testing. Google includes AI Overviews and AI Mode traffic in the Search Console Performance report under the Web search type, without a separate breakdown. To see whether the business is actually mentioned or linked, test a fixed set of queries in Arabic and English and record the results.

Because Search Console does not separate AI feature traffic, a change in Web search performance cannot be attributed to AI Overviews alone. Read Search Console as the overall Search signal, and use structured observation to understand AI features specifically.

For observation, keep a stable query list per country and language, record the date, whether an AI Overview appeared, whether the business was linked, which other domains were linked and how prominent the link was. Repeat on a schedule. For Gemini, use the same method with prompts instead of queries.

Can any agency guarantee inclusion in Google AI Overviews?

No agency can guarantee inclusion in Google AI Overviews, AI Mode or Gemini answers. Google states that indexing and serving are not guaranteed even when a page meets every requirement, AI Overviews only appear for some queries, and supporting links vary. A business can meet the documented requirements, improve its content and measure honestly, but no one controls Google's selection.

Google's documentation describes no paid or partner route into organic AI Overviews. Any claim of privileged access to Google's AI features should be treated with scepticism.

A credible plan sets out what will be checked, what will be changed and how change will be measured against a dated baseline. That is the honest alternative to a guarantee.

Frequently asked questions