Ranking in AI Overviews has five conditions: a top-ten organic position for the query, a page Googlebot can retrieve without error, a passage that answers the question and can be quoted whole, specifics rather than adjectives, and a business Google can resolve to one entity. Ahrefs found 76 percent of pages cited in AI Overviews already rank in the top ten. The rest is writing and plumbing.
What gets cited, what does not, the six-step method, and how to measure whether it worked.
To rank in Google AI Overviews, a page has to rank in the top ten for the query first, since Ahrefs found 76 percent of AI Overview-cited pages also hold a top-ten organic position (1.9 million citations, July 2025). Then it has to be retrievable by Googlebot without error, answer the question directly in a passage the model can quote whole, state specifics rather than adjectives, and belong to a business Google can resolve to one entity in one place. There is no separate submission and no trick.
Each condition can be checked. Ranking is visible in Search Console. Retrievability is a fetch: if the server returns 406, 429 or a bot challenge to Googlebot, the Overview cannot read the page even when ChatGPT, drawing on cached and third-party sources, still names the company. We measured exactly that in August 2026: 44 ChatGPT mentions and 2 AI Overview appearances for the same company in the same week, caused by one line in a server configuration.
Quotability is a writing discipline. The model lifts passages that stand alone. A paragraph that opens with the answer, under a heading phrased as the question, survives being lifted. A paragraph that builds up to the point does not.
AI Overviews cite passages that answer the query directly, carry a specific figure or definition, and come from pages Google already trusts for the topic. In practice that means a question-form heading, a bold one-paragraph answer beneath it, evidence after, and nothing promotional in the quoted span. Pages that rank for the query and are written this way get cited; pages that rank and are written as brochures do not.
The reason is mechanical. The Overview is composed from retrieved passages, and the model selects passages that resolve the question with the least surrounding context. "We are a leading provider of kitting services" resolves nothing. An illustrative rewrite, "Kitting is the assembly of multiple SKUs into one ready-to-ship unit, priced per kit," resolves the question and can be quoted with attribution. Every page on this site is written to the second pattern for that reason.
Google's own guidance says AI features use the same ranking and quality systems as Search and that no special markup is required. That is true and also incomplete: the pages that get cited are the ones that made themselves easy to quote.
Semrush measured AI Overviews on 6.49 percent of US desktop Google queries in January 2025 and 13.14 percent by March 2025, roughly doubling in two months. Informational and question-form queries trigger them most. The level varies by query set and method; every published study shows the same direction.
For B2B categories the practical reading is that a buyer's first search now often returns an answer before it returns a list. If the company is in the answer, it is on the shortlist. If it is not, position four on the list underneath receives a fraction of the clicks it did in 2024; Seer Interactive measured organic click-through on informational queries with an Overview at 0.61 percent in September 2025, against 1.62 percent on queries without one.
Six steps, in order, which put the five conditions above into practice: confirm the page ranks top ten or fix that first; fetch the page as Googlebot and confirm a 200; rewrite the opening of each section as a question heading with a direct answer; replace adjectives with figures, names and dates; add Organization, WebPage and FAQ structured data; resolve the business to one name, address and phone everywhere. Then query the Overview for the category and record whether the page is cited.
Then measure. Ask Google the category question with the market attached, record whether the Overview appears, whether the page is cited, and which sources are. Repeat monthly. Compare with AI Mode, ChatGPT and Gemini, because they disagree and the gaps are diagnostic.
Keyword stuffing, blocking AI crawlers while expecting Google to cite you, a blended "AI visibility score" that hides which engine is failing, and buying links. Also content written for the model rather than the buyer: pages of question headings with nothing underneath are read as thin and ranked accordingly.
Figures described as measured are from Search Standing's own analyses using Semrush, Google Search Console and direct queries to each AI engine, on the dates stated. Published ranges move quarter to quarter; the patterns are what we rely on.
Send your company and your market. We will run the analysis and send back a summary naming your competitors, the terms they are winning, and where you sit against them.
No call required and no obligation. Read it, check it yourself, and decide. If the findings are worth discussing after that, we can talk then.
Prepared for your company against the competitors you name. Written by the person who would run the paid work.
Or email contact@searchstanding.com directly.