E-E-A-T Explained: What Google Means by Experience, Expertise, Authoritativeness and Trust

Published 14 September 2026 · 9 min read

E-E-A-T stands for Experience, Expertise, Authoritativeness and Trustworthiness. It comes from Google's Search Quality Rater Guidelines, not from the ranking algorithm, and Google says it is not a direct ranking factor. It is a description of the qualities Google's systems try to reward, and the practical work is making those four qualities visible on the page.

Where E-E-A-T comes from

Google pays external quality raters to evaluate search results against a public document, the Search Quality Rater Guidelines. The raters do not decide where any page ranks. Google says their ratings are used to judge whether a proposed change to its ranking systems is an improvement. The guidelines are therefore a statement of what Google wants its results to look like.

The original acronym was E-A-T, and it has been in the guidelines since 2014. On December 15, 2022, Google added a second E for Experience and announced the change on the Search Central blog. The addition recognizes that for many topics, having done the thing matters as much as having studied it: a shoe review by someone who ran in it is not a summary of the spec sheet.

The same update made one more thing explicit: Google describes Trust as the most important member of the family. The other three are ways a page earns trust, and an untrustworthy page has low E-E-A-T no matter how expert its author appears.

Not a ranking factor, but a target

Google's page Creating helpful, reliable, people-first content says that E-E-A-T itself is not a specific ranking factor, and that its systems instead use a mix of signals that can identify content with good E-E-A-T. In our reading this settles two questions. There is no E-E-A-T score to optimize, because Google does not compute one. And no single tag, badge or plugin raises it, because the systems look for a pattern across many signals, not a declaration.

It also means claims about effect size are unsupported. Google publishes no number for how much a byline or a contact page changes rankings, and we will not invent one. What can be said honestly is which on-page signals correspond to what raters are asked to look for.

The four parts and what demonstrates them

The guidelines phrase each part as a question about the content creator and the page. The table maps each question to the kind of evidence that answers it.

The second table lists the concrete page elements, which part each mainly supports, and what good looks like.

None of this is a trick. Each row is what a skeptical reader uses to decide whether to believe a page, and the guidelines ask raters to behave like that reader.

YMYL topics and why the bar is higher

The guidelines single out a category they call YMYL, short for Your Money or Your Life: topics that could significantly affect the health, financial stability, safety or welfare of people or society: medical, financial, legal, safety and civic-news topics. For these, raters are told to apply a higher standard of E-E-A-T, to weigh expertise and trust most heavily, and to give the lowest rating to pages that are inaccurate or potentially harmful.

The practical consequence is that the same byline can be adequate on one site and a weakness on another. “Written by the editorial team” on a page about standing desks is fine. On a page about drug interactions or mortgage refinancing it is a gap, because the reader cannot judge whether the writer is qualified to give that advice.

The guidelines are also clear that expertise does not always mean credentials. They describe everyday expertise: for many topics, the most useful author is someone who has lived with the condition or run the small business. The page should show that experience rather than borrow a credential it does not have.

Common mistakes

Most E-E-A-T problems are signals added for appearance that fall apart when anyone looks.

• Fake author personas. Invented names, generated headshots, bios that cannot be verified anywhere outside the site. Raters are instructed to look up who is behind the content. A persona that exists only on your domain is worse than an honest company byline, because it fails the check and signals intent to mislead.
• Stock “expert reviewed” badges. A badge that says “medically reviewed” with no named reviewer, no date and no credentials is decoration. If a review happened, name the reviewer, link their bio and say when.
• No contact page. Or a form with no company name, address or email. The guidelines treat not being able to identify who is responsible for a site as a reason for a low rating, especially on YMYL topics.
• Saying it instead of showing it. “We are the leading experts in…” is a claim. A dated screenshot of your own test is evidence.
• Dates that lie. Scripts that refresh the “updated” date on every load, or a modified date that moves without any change to the text. Once a reader notices, every other date on the site loses credibility.
• Citing the summary instead of the source. Linking a blog post that describes a Google announcement instead of the announcement itself.

A practical checklist

1. Every article has a byline that links to an author page with a real photo, links to profiles elsewhere, and a bio that says why this person is qualified for this topic.
2. Any claim a reader could check is linked to a primary source: documentation, the study, the filing, the standard.
3. Where you tested something, show the test: screenshots, dates, versions, setup, results.
4. A contact page with the legal business name and a way to reach a human. Legal pages that match where you operate.
5. HTTPS on every URL, with HTTP redirecting to it and no mixed content.
6. A short editorial policy: how content is produced, reviewed and updated, and whether AI tools are involved.
7. A corrections practice: when something material changes, note what and when, and let the modified date move only then.
8. Organization schema on the homepage and Person schema on author pages, each with sameAs links to profiles elsewhere.
9. For YMYL topics, a named, qualified reviewer with a linked bio, or a clear statement that the page is not professional advice.

Why the same signals matter to AI systems

AI answer engines retrieve candidate pages and decide which ones to draw on and cite. They read the same HTML a rater reads. In our reading, the things that let a human decide a page is trustworthy — a named author, a date, primary-source citations, an identifiable organization — are also what a retrieval system can use to prefer one candidate over another and attribute it correctly. No vendor publishes its weighting, and we do not claim to know it. What we can say is that a page with no author, no date and no owner gives a system nothing to attribute.

Structured data is where the two worlds overlap most directly. The schema markup guide covers Organization on the homepage; the E-E-A-T counterpart is Person on each author page, with sameAs pointing to profiles beyond your domain and worksFor tying the person to the organization.

<script type="application/ld+json">
{
  "@context": "https://schema.org",
  "@type": "Person",
  "name": "Jane Example",
  "url": "https://yourdomain.com/authors/jane-example",
  "jobTitle": "Technical SEO Lead",
  "description": "Ten years of technical SEO work; writes about crawling and indexing.",
  "worksFor": { "@type": "Organization", "name": "Your Company", "url": "https://yourdomain.com" },
  "sameAs": [
    "https://www.linkedin.com/in/jane-example",
    "https://github.com/jane-example"
  ]
}
</script>

Then reference that person as the author of each Article, so the markup and the visible byline agree. Markup that names an author the page does not is the structured-data version of a fake badge.

Two cautions. First, do not try to measure the effect of E-E-A-T changes on AI citations by running a handful of prompts before and after; the article on honest versus fake AI citation metrics explains why that produces noise, not evidence. Second, automated checks cover only the machine-readable slice. Aura's free scan tests a page for HTTPS, valid JSON-LD, Organization and WebSite schema, Article and FAQPage schema and a canonical tag. It does not evaluate author bios, source quality or backlinks. Those remain a human review, which is fitting for a concept that started as one.

Frequently asked questions

Is E-E-A-T a Google ranking factor?
No, not in the literal sense. Google says on its Creating helpful, reliable, people-first content page that E-E-A-T itself is not a specific ranking factor, but that its ranking systems use a mix of signals to identify content that demonstrates good E-E-A-T. The concept comes from the Search Quality Rater Guidelines, and rater scores do not directly change the ranking of any individual page.

What is the difference between E-A-T and E-E-A-T?
The extra E stands for Experience. Google added it to the Search Quality Rater Guidelines in December 2022. It asks whether the content creator has first-hand experience with the topic — has used the product, visited the place, or followed the process — which is separate from formal expertise or credentials.

What are YMYL topics?
YMYL stands for Your Money or Your Life. The rater guidelines use it for topics that could significantly affect the health, financial stability, safety or welfare of people or society, such as medical, financial, legal and civic-news topics. Raters are asked to hold pages on these topics to a higher E-E-A-T standard, with trust and expertise weighed most heavily.

Does E-E-A-T matter for AI search and citations?
The on-page signals that demonstrate E-E-A-T — a named author with a bio, dates, citations to primary sources, an identifiable organization with contact details, and consistent Organization and Person schema with sameAs links — are also what makes a page easy for an AI system to attribute and cite correctly. No AI vendor publishes how it weights these signals, so treat them as good practice rather than a measured effect.

Run the free 18-check scan