How does GEO work?

How do Google's AI answers choose sources?

Google explains two techniques in its guide to generative AI features. The first is retrieval-augmented generation, or grounding: the AI relies "on our core Search ranking systems to retrieve relevant, up-to-date web pages from our Search index," then shows links to the pages that support its answer. The second is query fan-out: the AI runs several related searches at once. Google's example: a search for "how to fix a lawn that's full of weeds" might also fetch results for "best herbicides for lawns" and "how to prevent weeds in lawn."

Fan-out explains something that surprises many business owners. A page can be cited for a question it doesn't rank for, because it ranked for one of the related searches the AI ran behind the scenes.

Do you need to rank on page one to be cited?

No. Ahrefs analyzed 863,000 searches and 4 million AI Overview citations and found that 37.9% of cited pages also appeared in the first 10 positions on the results page for the same search, counting ads and other search features. The rest split almost evenly between positions 11 to 100 and beyond the top 100. Ahrefs' earlier version of this study had found a much higher overlap, so the number moves as Google's systems change.

The practical takeaway: covering a topic thoroughly, across the related questions people ask, gives an AI system more chances to find you. One strong page about a service, plus pages answering the specific questions around it, creates more entry points than a single page trying to rank for everything.

What makes content more likely to be cited?

The research paper that introduced the term GEO, by Aggarwal and colleagues and published at the KDD 2024 conference, tested nine ways of rewriting content across a benchmark of 10,000 queries. It found that "including citations, quotations from relevant sources, and statistics can significantly boost source visibility, with an increase of over 40% across various queries," and measured improvements of up to 37% on Perplexity, a live AI search engine. Keyword stuffing, the old SEO habit of repeating search phrases, didn't perform well.

Google's guidance points the same way from a different angle. It tells site owners to create "non-commodity content": a unique point of view or first-hand experience, not a summary of what's already online. Its example is a homebuyer's account of why they waived an inspection, set against a generic "7 Tips for First-Time Homebuyers." For a trade business, that means writing about the jobs you've actually done, with real numbers and named sources, instead of generic advice anyone could write.

What doesn't work?

Google's guide includes a list of popular GEO tactics it says you can ignore for Google Search:

• llms.txt and other "special" files. Google says creating them "will neither harm nor help your site's visibility or rankings in Google Search, as Google Search ignores them."
• Chunking content. "There's no requirement to break your content into tiny pieces for AI to better understand it," and "there's no ideal page length."
• Rewriting for AI. "You don't need to write in a specific way just for generative AI search," because AI systems understand synonyms and meaning.
• Special schema. "There's no special schema.org markup you need to add," though structured data still helps with rich results.
• Inauthentic mentions. Seeking fake "mentions" across the web "isn't as helpful as it might seem."

These points cover Google only. ChatGPT and Perplexity use their own systems, but the research above suggests the same basics, real sources and real substance, travel well across all of them.

How does GEO apply to a local business?

AI answers don't appear for every search. Google says AI Overviews are "only shown when our systems determine that it is additive to classic Search, and as such, often don't trigger." For someone who wants a plumber right now, the map pack still does the work. AI answers matter more for the research that happens earlier, like questions about cost, process, or which option to choose.

The new AI search category in Whitespark's 2026 Local Search Ranking Factors survey shows what local experts expect to matter. Their top factors were being on expert-curated "best of" lists, having a dedicated page for each service, prominence on key industry websites, and mentions in newspapers, blogs, government sites, and industry associations. Most of that is prominence: being known and written about, which no file or formatting trick can replace.