Does E-E-A-T matter for being cited by AI?
Published on
Not as a box to tick. E-E-A-T, for experience, expertise, authoritativeness, and trustworthiness, is Google's vocabulary, and Google itself writes that it is not a ranking factor in itself. For its AI answers, Google states that there is no additional requirement. What comes closest to it in our measurements is authority in the sense of name recognition: ChatGPT first looks for sources whose name it already knows. An author page does not create that recognition; being cited elsewhere does.
For what type of question? It all depends on the AI and the question. For Google's AI answers, E-E-A-T is part of what its quality systems seek to recognize. For ChatGPT and Claude, the term has no meaning of its own: they have their own ways of choosing their sources, which we measure. This page covers both.
What Google writes about E-E-A-T
The acronym comes from Google. In its page on helpful, reliable content, Google explains that its systems seek to highlight content that demonstrates experience, expertise, authoritativeness, and trustworthiness, what it calls E-E-A-T, and that trustworthiness is the most important of these aspects. It immediately specifies that E-E-A-T is not in itself a ranking factor: it is a set of signals that its systems try to identify, with greater weight for topics that affect people's health, money, or safety.
A frequent confusion comes from quality raters, those outside people whom Google asks to judge results according to its guidelines, in which E-E-A-T is described at length. Google writes that these raters have no control over how pages rank and that their ratings are not used directly by its algorithms: they serve to check that the systems are working. On authors, Google encourages stating who wrote a piece of content, through a byline where one would expect it, and warns against invented author profiles.
What Google writes about its AI answers
For AI Overviews and AI Mode, Google states that there is no additional requirement to appear in them, nor any special optimization needed: a page must be indexed and able to be displayed with a snippet. Its guide on generative AI features adds that these features rely on its usual ranking and quality systems, and that there is no need to write in a particular way for AI search. Neither of these two pages mentions E-E-A-T or authors.
We can therefore say, without going beyond what Google writes, that what counts for quality in its classic search also counts for its AI answers, since they are the same systems. We cannot say that adding an author biography or Person markup makes a page appear in an AI Overview: Google does not write that, and we have found no measurement that shows it.
What comes closest to it in our measurements
We have not measured the effect of an author byline on citations. But two results of our measurements touch on what E-E-A-T seeks to describe. The first concerns authority, in the sense that a source is known: in 72 to 91% of its searches according to our corpora, ChatGPT writes the name of a specific publisher it already knows. A company that the AI has never seen named is not searched for. The authority that counts here is not declared on one's own page; it is built through mentions in the sources the AI consults.
The second concerns experience and expertise, in the sense of what only you know. With Claude, a page whose content overlaps with that of the other pages found is cited up to 1.5 times more often; once the page is retained, it is its own data, its figures, and its sources that set it apart. A page that reports a measurement, field experience, or a dated figure therefore brings something that a generic page does not, whether it is signed or not.
What to do, in practice
Stating who writes, who runs the company, and what its claims are based on remains good practice: Google recommends it, and a reader, human or not, judges a text better when they know where it comes from. It just needs to be put in its proper place, as one element of trust among others, and not as a proven citation lever. Invented profiles, which Google warns against, are to be ruled out.
To be cited by AIs, the most useful effort lies elsewhere: pages that precisely answer your customers' questions, data that only you can provide, and mentions in third-party sources in your field. It is this external presence that makes a company a source that an AI thinks of looking for.
What we do not know
Our measurements cover the API of GPT-5, Claude Sonnet 4.6, and Gemini 3.5 Flash with web search, not their applications nor Google's AI Overviews. We have not measured the effect of an author byline or of Person markup on citations.
Sources
- Google Search Central — “Creating helpful, reliable, people-first content”, .
- Google Search Central — “AI features and your website”, .
- Google Search Central — “Optimizing your website for generative AI features on Google Search”, .
- FaireDuBruit measurements, from September 4 to 16, 2026 — see “How we measured”.