AI Visibility
AI citation
When a generative engine attributes part of its answer to a specific source, naming and usually linking the page it drew from.
By Shimon Carroll, Founder, SEO for AI Agents · Last updated
An AI citation is the moment a model points at a source. In Perplexity, ChatGPT Search, Gemini, and Google AI Overviews, the answer is accompanied by a short list of links the system used. Being one of those links is the AI-era equivalent of ranking on page one, except the competition is for a much smaller number of slots.
Citations are not distributed at random. Independent research from Vectara and others has found that AI answers lean heavily on a small set of trusted sources, with Wikipedia accounting for a large share of factual citations in measured samples. The implication is blunt: corroboration and entity recognition matter more than raw page-level optimization. A brand with a Wikipedia entry, a Wikidata Q-number, organic Reddit discussion, and consistent facts across review sites is cited more than a brand twice its size that exists only on its own domain.
The levers that move citation rate, in rough order of impact, are: being readable to non-JS crawlers, having the answer present as an extractable passage, being corroborated across the sources the engine trusts, having clear entity association via schema and sameAs links, and freshness. Tracking citations means querying the engines for the questions that matter to your brand and recording which of your pages, if any, get named.