citemark
Menu

Guide

How to get cited in Google AI Overviews

Published · By Shoaib Mustaque· 6-minute read

Google AI Overviews are built from Google's search index and rankings. Google says they are “built to only show information that is backed up by top web results” and that a page must be indexed and eligible for a snippet to be linked. The route in is SEO done well, then passages written to be lifted, then an unambiguous entity.

On the record

What Google has said about how AI Overviews work

Google's documentation makes five statements that matter. There are no additional requirements to appear in AI Overviews or AI Mode. No special structured data is needed. Both features may use “query fan-out”, issuing several related searches. Overviews run on a Gemini model customised for Search and tied into its ranking systems. And meeting every requirement still does not guarantee serving.

What Google says, and where
StatementWhere Google says it
“There are no additional requirements to appear in AI Overviews or AI Mode, nor other special optimizations necessary.”Search Central, “AI features and your website”
A page must be “indexed and eligible to be shown in Google Search with a snippet” to be a supporting link.Search Central, “AI features and your website”
“There's also no special schema.org structured data that you need to add.”Search Central, “AI features and your website”
AI Overviews and AI Mode “may use a ‘query fan-out’ technique — issuing multiple related searches across subtopics and data sources”.Search Central, “AI features and your website”
“AI Overviews are built to only show information that is backed up by top web results.”The Keyword, “AI Overviews: About last week”, May 2024
Google-Extended “does not impact a site's inclusion in Google Search nor is it used as a ranking signal”.Search Central, “Google's common crawlers”

The sentence to build a plan on is the second one. An Overview is a layer on top of ranking, not a separate index. Everything that follows is a consequence of that.

Step one

Rank, in Google, for the sub-questions

The first step is ordinary SEO aimed at the sub-questions an Overview assembles. Query fan-out means Google may search for several related topics to build one answer, so a page that ranks for a precise sub-question can be linked even when a broader page does not. Google's Search Essentials describe the technical requirements; nothing extra applies.

  1. 01

    Map the sub-questions

    For each prompt you care about, list the five to ten questions a thorough answer must cover: definition, cost, timeline, comparison, local availability. Those are the pages to rank.

  2. 02

    One page per sub-question

    A page that answers one question well beats a page that answers eight badly. Use the question as the H2 and open with the answer.

  3. 03

    Get indexed and snippet-eligible

    Check Search Console for coverage. Remove noindex and nosnippet where you want to be cited. Google's requirement is literal: indexed, and eligible for a snippet.

  4. 04

    Earn the top ten

    Answer features draw from pages already ranking well. Internal links, a specific title, and a handful of relevant external mentions do most of the work for long-tail questions.

Step two

Write passages an Overview can lift

An Overview lifts passages, not pages. The passage that gets lifted is the one that answers a sub-question completely in a few sentences, with a number or a date, under a heading that states the question. Tables and numbered lists are lifted for comparisons and processes. Structured data must describe the visible text, not substitute for it.

  • Question as heading, answer in 40–60 words. Every H2 on this site follows that rule, including this one. A model can lift the paragraph without the rest of the page.
  • One fact per sentence. Short sentences survive summarisation. A sentence with three clauses loses two.
  • Numbers, dates, names. “Google withdrew FAQ rich results on 7 May 2026” is quotable. “Google recently changed its FAQ policy” is not.
  • A table where the content compares, a numbered list where it sequences. Overviews render both.
  • A visible date and a named author. Google's guidance on helpful content asks for accurate bylines and warns against fabricated ones.
  • Schema that matches the text. Google's structured-data guidance says not to mark up anything the user cannot see.

Step three

Make the entity unambiguous

Google's Knowledge Graph holds, by Google's count, over 500 billion facts about five billion entities, drawn from the web, licensed databases and the markup site owners add. Organization structured data is how Google says it disambiguates your organisation in results. If Google is unsure which company you are, it has no reason to name you.

  1. 01

    One Organization block, on the home or About page

    Google recommends the most specific subtype that fits (ProfessionalService for an agency), with name, legalName, url, logo, sameAs, address and founder. Google says this markup is used to “disambiguate your organization in search results” and to pick the logo for a knowledge panel.

  2. 02

    The same facts everywhere

    Google Business Profile, LinkedIn, Crunchbase, Clutch and directory listings must say the same name, category, address and founding year. Contradictions are the usual reason a brand's knowledge panel never appears.

  3. 03

    A named person

    A founder with a Person entity, a LinkedIn profile and bylines on the site gives Google a second entity to connect to the first.

  4. 04

    Mentions on pages Google already trusts

    Industry lists, local press, directories. These are also the pages Overviews cite for “best” and “top” prompts.

The schema guide shows how we generate the Organization, Service and Article blocks from one data file so they can never contradict each other.

Candidly

What does not help

Four things people pay for do not move AI Overviews. Special “AI schema”: Google says none is needed. FAQ markup for rich results: Google withdrew the feature on 7 May 2026. Blocking or allowing Google-Extended: it does not affect Search inclusion or ranking. And paid links, which Google's spam policies cover.

FAQ structure is a special case. The rich result is gone, but a real question with a real answer under it is still the cleanest shape for any answer model to parse, so we keep the structure and drop the expectation of stars.

A guarantee of inclusion is the fifth thing not to pay for. Google's own documentation says that meeting all requirements and best practices does not mean Google will crawl, index or serve a page. Anyone who promises otherwise is promising something Google says it cannot promise.

Timeline

Timeline, and how to measure it

AI Overview inclusion follows organic rankings, so it moves at the speed of your SEO: typically months two to four for low-competition questions as pages reach the top ten, and six to twelve months for competitive category prompts. Measure it per mapped question, monthly, with a screenshot of the Overview and its links.

AI Overview timeline
WhenWhat is usually true
Weeks 1–2Indexing and snippet eligibility fixed; sub-question map agreed; Organization schema live
Weeks 2–8Restructured pages re-crawled; featured snippets for long-tail questions where you already rank
Months 2–4First AI Overview links for low-competition questions as rankings rise
Months 6–12Links in Overviews for competitive prompts; brand facts correct in the Overview's description of you

Overviews do not appear for every query, and whether one appears for a given search is Google's call, not yours. Our reports record three states per question: no Overview shown, Overview shown without you, Overview shown with you. Only the third counts.

If you want to know your state today, the free AI visibility audit checks your category's questions on Google and the four assistants.

Sources

Sources

Every factual claim above about how an engine works, or about what a company publishes, links to the source it came from. Each URL was opened and checked on the date shown at the top of this guide. If a source later changes or disappears, the claim should be re-checked, not assumed.

  1. 01Google Search Central, “AI features and your website”https://developers.google.com/search/docs/appearance/ai-features
  2. 02Google, The Keyword, “AI Overviews: About last week”, May 2024https://blog.google/products/search/ai-overviews-update-may-2024/
  3. 03Google, The Keyword, “Generative AI in Search: Let Google do the searching for you”, 14 May 2024https://blog.google/products/search/generative-ai-google-search-may-2024/
  4. 04Google, The Keyword, “How Google's Knowledge Graph works”https://blog.google/products/search/about-knowledge-graph-and-knowledge-panels/
  5. 05Google Search Central, “Organization (Organization) structured data”https://developers.google.com/search/docs/appearance/structured-data/organization
  6. 06Google Search Central, “Introduction to structured data markup”https://developers.google.com/search/docs/appearance/structured-data/intro-structured-data
  7. 07Google Search Central, “Creating helpful, reliable, people-first content”https://developers.google.com/search/docs/fundamentals/creating-helpful-content
  8. 08Google Search Central, “Google Search Essentials”https://developers.google.com/search/docs/essentials
  9. 09Google Search Central, “Google's common crawlers” (Google-Extended)https://developers.google.com/search/docs/crawling-indexing/google-common-crawlers
  10. 10Google Search Central, documentation updates (FAQ rich result deprecation, 8 May 2026)https://developers.google.com/search/updates

Keep reading

Related guides

  • How ChatGPT chooses which sources to cite

    ChatGPT search retrieves pages through third-party search providers, with Bing as its default, then picks passages that answer the question. What OpenAI documents, what it doesn't, the three crawlers, and a checklist for getting indexed and quoted.
  • Schema markup for AI search: what actually matters

    Which schema.org types help AI search and which don't: Organization as ProfessionalService, Service, Article, BreadcrumbList and FAQPage after Google's May 2026 withdrawal. Why to generate schema from one source of truth, and how to check it.
  • GEO vs AEO vs SEO: what each one is for

    SEO ranks a page in a list of links. AEO wins the direct-answer slot inside Google. GEO earns a citation inside an answer written by ChatGPT, Gemini, Perplexity or Claude. A comparison table, a decision list for which to do first, and the timelines.

See which prompts you're missing from

The free audit runs your category's prompts on ChatGPT, Gemini, Perplexity, Claude and Google AI Overviews and shows who each engine cites today. Five working days, screenshots included.