Google AI Overviews: How Pages Get Cited

Short answer: Google cites a page in an AI Overview when the page is already indexed, eligible for a snippet, and ranks for one of the several related searches Google runs behind the answer. There is no AI-specific markup, file or schema to add, and Google says so in its own documentation. The work is ordinary search visibility plus pages that state facts plainly enough to be quoted.

Sources are Google Search Central documentation, Google's I/O 2026 announcement, Pew Research and Ahrefs, checked 17 October 2026.

What is a Google AI Overview?

An AI Overview is the generated summary with links that appears above the standard results for some searches. AI Mode is the conversational version, a tab where the whole result is an answer with follow-up questions.

• Google said AI Overviews were available in more than 200 countries and territories and more than 40 languages as of 20 May 2025 (Google).
• AI Mode reached Singapore users in English on 25 August 2025 (TechCoffeeHouse).
• At I/O on 19 May 2026, Google said AI Mode had passed one billion monthly users, with queries more than doubling every quarter, and made Gemini 3.5 Flash the default model (Google).
• Google describes the system as retrieval-augmented generation. It relies "on our core Search ranking systems to retrieve relevant, up-to-date web pages" and then shows "prominent, clickable links to relevant web pages that support the information in the response" (Google Search Central, updated 10 July 2026).

How does Google choose which pages to cite?

Three conditions, all from Google's documentation.

• Indexed and snippet-eligible. To appear as a link in AI Overviews or AI Mode, a page "must be indexed and eligible to be shown in Google Search with a snippet" (Google Search Central).
• Retrieved by query fan-out. Google runs several related searches at once for one question and draws links from all of them. Its documentation says this shows "a wider and more diverse set of helpful links" than a single results page. A page that ranks for a specific sub-question can be cited even if it never ranks for the head term.
• Supports a specific statement. Links attach to claims in the answer. A page that contains the claim in plain text, with a number or a date, is easier to attach than one that implies it.

Google's listed best practices for AI features are the same as for Search:

• Allow Googlebot in robots.txt.
• Link to the page from elsewhere on the site.
• Keep the important content as text, not inside images or video.
• Make structured data match the visible text.
• Keep Business Profile and Merchant Center data current.

For a Singapore service firm this means one page per service, the price range and timeline in the first paragraph, named locations, and the date the page was last checked. The ChatGPT side of the same work is in how to get your business mentioned by ChatGPT.

Does schema help a page appear in AI Overviews?

Not as a switch. Google's guide states: "Structured data isn't required for generative AI search, and there's no special schema.org markup you need to add." The same guide says you do not need "new machine readable files, AI text files, markup, or Markdown", and that there is "no requirement to break your content into tiny pieces for AI to better understand it".

What schema still does:

• Organization, LocalBusiness and Service markup make name, address, phone, hours and price range unambiguous for every Google system.
• FAQ and Article markup tie question and answer pairs, authors and dates to the page.
• Markup that contradicts the visible page is a spam risk, not a help.

Why is my content misquoted or skipped?

Common causes, in the order they turn up on SME sites:

• The fact is not on the page in words. Prices in a PDF or a JPEG, hours in an image banner, services in a slider. Google asks for important content in textual form.
• Two pages say two things. An old post with a 2023 price and a service page with a 2026 price. The engine may pick the old one. Update or redirect the old page.
• Snippet controls are on. Google states that nosnippet, data-nosnippet, max-snippet and noindex limit what appears in Search, including AI features. A theme or plugin that set a low max-snippet cuts the quotable text.
• The page is thin. A 120-word service page gives nothing to support a claim.
• No date and no author. Google's helpful-content guidance asks who wrote the page, how, and why (Google Search Central). An undated page with no named firm reads as lower trust.

Fix in this order: put facts in text, remove contradictions, check robots and meta tags, then expand the page.

Do AI Overviews reduce clicks to my site?

Two dated studies, both US-based.

View: informational pages lose clicks. Pages that answer "how much" and "who" for a named service in Singapore still bring enquiries, because the searcher needs a supplier, not a definition. What changed is that the page must earn the citation to be seen at all. The split between the two kinds of work is in GEO vs SEO.

How do I track AI Overview appearances?

Google added a Generative AI performance report to Search Console, announced 3 June 2026 (Google Search Central blog). Its help page describes what it shows (Search Console Help):

• Impressions: the times links to your site were shown in a generative AI feature on Search. Several links from one site in one answer count once.
• Pages, countries, devices and dates.
• AI Overviews and AI Mode are reported together, not split.
• Clicks are not a metric in this report. Traffic from AI features stays inside the Web search type in the main Performance report.

A monthly routine:

1. Export the report's page list and note which URLs appear.
2. Run your 10 most valuable searches in AI Mode and record whether you are cited and for which sentence.
3. Fix the pages that appear in the report but not for the searches you want.

What cense does

cense builds Next.js sites with every service page in plain text, dated, with structured data that matches the page and no plugins to override snippet settings. Pricing and process are on AI search optimisation in Singapore.