E-E-A-T in SEO: Google Says It Is Not a Ranking Factor, and It Still Decides Your Recovery
· Royking Niba
Google’s own wording is the place to start, because almost everything written about E-E-A-T ignores it. In its guidance on creating helpful, reliable, people-first content, Google says: “While E-E-A-T itself isn’t a specific ranking factor, using a mix of factors that can identify content with good E-E-A-T is useful.” And on the four letters, it says: “Of these aspects, trust is most important.” So E-E-A-T is a description of what Google’s systems are trying to detect. It is not a dial you turn. That distinction is the difference between work that moves a site after a core update and work that produces an author box nobody reads.
I raise it because E-E-A-T is the single most common thing a client brings me after a demotion, usually in the form of a vendor proposal to add author schema across four hundred pages. That proposal will not work, and the reason it will not work is in Google’s documentation, not in anyone’s theory.
Letter by letter, against what Google actually published
Each letter maps to something concrete you can put on a page, and to something else that merely looks like it. The middle column here is Google’s language or the self-assessment question it publishes. The right-hand column is what I actually build.
| Letter | What Google’s documentation asks | The artefact that satisfies it | The version that does not |
|---|---|---|---|
| Experience | Whether the content shows first-hand involvement with the subject | Specifics only someone who did the thing would have: the number, the date, the thing that went wrong, the screenshot of your own data | A sentence claiming years of experience |
| Expertise | “Is this content written or reviewed by an expert or enthusiast who demonstrably knows the topic well?” | Demonstration inside the body copy, plus a named reviewer where review genuinely happened | A reviewer byline on content nobody reviewed |
| Authoritativeness | Whether the site is a recognised source on the topic | Depth on a narrow subject, and citations from people who had no reason to link to you | Directory listings and badges |
| Trust | “Does the content present information in a way that makes you want to trust it, such as clear sourcing, evidence of the expertise involved” | Sourcing next to the claim, a dated last-review line, a corrections note, a real contact route, disclosure of commercial relationships | A trust seal image in the footer |
Notice that trust, the letter Google calls most important, is the one that is almost entirely made of things a developer can ship this week. Experience and expertise live in the writing and cannot be retrofitted with markup. That asymmetry is why most E-E-A-T projects fail: they buy the two letters you cannot buy and skip the one you can build.
The Who, How and Why test, used as an audit
Google frames its self-assessment around three questions. I run them literally, as a pass or fail against a sample of pages, because a vague sense that a site has a quality problem is not actionable and a list of failed checks is.
| Question | Google’s published wording | How I test it | Typical failure I find |
|---|---|---|---|
| Who | “Is it self-evident to your visitors who authored your content?” and “Do pages carry a byline, where one might be expected?” and “Do bylines lead to further information about the author or authors involved” | Open twenty pages as a logged-out visitor. Count how many name a person, and how many of those names lead anywhere | A byline that links to an archive page with no bio, or a category name where an author should be. This is a front-end requirement, not a markup one, so structured data alone does not clear it |
| How | “Is the use of automation, including AI-generation, self-evident to visitors through disclosures or in other ways?” | Check whether the production method is disclosed where a reader would want to know, and whether review claims are true | Undisclosed generated content at scale, which is a policy problem and not an authorship one. I go through where that line sits in the piece below |
| Why | Content should be “created primarily to help people, content that is useful to visitors if they come to your site directly” | For each page, ask who would want it if search did not exist. If the answer is nobody, it is search-engine-first content | Pages built to cover a keyword, which Google describes as content “primarily made to attract visits from search engines” |
The Why test is the one that finds real damage, and it is brutal on sites that scaled. Google explicitly warns against producing “lots of content on many different topics in hopes that some of it might perform well in search results.” If a section of your site fails Why on most pages, no amount of author markup on those pages will help, because the problem is the pages.
Where the raters fit, and why people misread them
Google describes quality raters as “people who give us insights on if our algorithms seem to be providing good results”, and says they are “trained to understand if content has strong E-E-A-T.” Raters do not assign your site a score that feeds a ranking. They evaluate whether changes to the ranking systems produce better results. That makes the rater guidelines a statement of intent, which is genuinely useful, and not a scoring rubric, which is how they are usually sold.
The practical consequence: you cannot argue your way to a better rating and there is nobody to appeal to. This is the same structural point as the difference between a human decision you can contest and a score nobody tells you about, which I set out in manual action penalty or algorithmic demotion.
What I actually do on an E-E-A-T brief
The order matters more than the list, because the cheap items are also the ones that do nothing on their own.
| Order | Work | Why it sits here |
|---|---|---|
| 1 | Segment the site by the Why test and decide what should not exist | Everything downstream is wasted on pages that fail Why. Removal and consolidation come before improvement |
| 2 | Fix trust at the template level: sourcing, dates, corrections policy, contact, commercial disclosure | Trust is the letter Google calls most important and the only one that ships as a template change |
| 3 | Put real bylines on pages that expect them, leading to a real bio | Google’s wording is about what is self-evident to visitors, so this is front-end work |
| 4 | Rewrite the surviving pages so experience is demonstrated in the body | The expensive part, and the part nothing substitutes for |
| 5 | Add structured data last | It describes what the page already says. It cannot create it |
Step one is where the traffic comes from, and it is the step clients resist hardest, because it means deleting things they paid for. The recovery sequence I ran on a sports property, where the surviving pages were a small fraction of the original, is written up in full in this spam update recovery case study. The same judgement about which weak pages are worth saving is in thin content, where there is no word count and no thin content penalty.
Three things E-E-A-T is not
It is not a score. Nothing in Search Console reports it and no third-party tool measures it, whatever the dashboard says. It is not a schema property, and there is no markup that asserts expertise. And it is not an authorship question in the way people assume: Google’s policy is about scaled content abuse and about purpose, not about who or what typed the words, which I set out in is AI content bad for SEO.
Where this bites hardest is commercial and money-adjacent content, because that is where trust signals carry the most weight and where the incentive to fake them is strongest. The version of this problem I see most often is in regulated affiliate markets, and I have written about what survives there in iGaming SEO after the spam updates.
If you are reading this because traffic dropped and somebody told you it was an E-E-A-T problem, get the diagnosis first. Most collapses are not what the first guess says they are, and the sequence I run before touching anything is in Google penalty recovery.
Leave a Reply