EEAT, Google's shorthand for experience, expertise, authoritativeness and trustworthiness, just got its most concrete update in years. In early October 2026, Google rewrote its helpful content guidance to spell out how its quality raters judge the main content of a page: on effort, originality, talent or skill, and accuracy. For the first time, the abstract idea behind E-E-A-T comes with four tests you can run on your own pages.
In the same week, Google told publishers to manually fact-check AI-generated content before it goes live, and labelled fabricated author profiles a form of deception. Together, these changes describe what Google considers a high-quality page more plainly than anything it has published before.
This guide covers what EEAT means and what it does not, what changed in October, how each of the four main content tests works in practice, and a step-by-step audit you can run on your site this month.
What EEAT means, and what it is not
E-E-A-T stands for Experience, Expertise, Authoritativeness and Trustworthiness. It started as E-A-T in Google's Search Quality Rater Guidelines, the handbook for the people Google pays to rate search results, and gained the extra E for experience in December 2022.
- Experience: first-hand involvement with the topic, such as having used the product, visited the place or done the job.
- Expertise: the knowledge or skill the creator brings to the subject.
- Authoritativeness: how far the creator or website is a known, go-to source on the subject.
- Trustworthiness: whether the page is accurate, honest, safe and reliable.
Google is clear about which of the four matters most:
Of these aspects, trust is most important.
Google Search Central
Two other sentences on the same page explain why EEAT is so often misunderstood. Google writes that "E-E-A-T itself isn't a specific ranking factor", and adds: "Rater data is not used directly in our ranking algorithms." Raters do not move your rankings. Their scores tell Google whether its ranking systems surface the kind of pages people find helpful, and Google tunes those systems accordingly.
So you cannot add EEAT to a page the way you add a meta tag. What you can do is make the page better at the things raters are trained to look for, because those are the things Google's systems are built to reward. That is why the October update matters: it tells you, in Google's own words, what raters look at first.
What Google changed in October 2026
Google made three related changes to its documentation in the first week of October 2026, reported by Search Engine Roundtable between October 1 and October 6:
- Main content became a named quality factor. The helpful content guide now states that "one of the most critical factors for assessing page quality is the quality of the main content", and defines what main content is.
- Four attributes for raters. A new section lists the "fundamental attributes of the main content" that Search Quality raters are trained to evaluate: effort, originality, talent or skill, and accuracy.
- Deceptive authorship is spelled out. Fabricated creator profiles, AI-generated headshots and false credentials are named as deception, and deception as a signal of a low-quality page.
In parallel, Google's guidance on generative AI content now tells publishers to "manually factcheck and review all AI-generated content for accuracy and trustworthiness before publishing", because generative outputs "may contain inaccuracies (also known as hallucinations)".
None of this introduces a new ranking system, and Google has not tied it to a core update. What it does is close the gap between the vague advice to "write helpful content" and what a reviewer actually checks.
What Google counts as main content
Google defines main content as "any part of the webpage that directly helps the page achieve its purpose." That is broader than the article text most SEO advice focuses on. According to the guide, main content can include:
- The main text, articles, images, audio or video files on the page.
- Built-in functionality such as calculators, online tools, games or search.
- Forum discussions, customer reviews, comments and user-uploaded media.
- Information housed behind interactive tabs.
- The visible title and headings that summarize the page's topic.
Two points in that list deserve attention. First, content in tabs counts. The old fear that text behind tabs or accordions is ignored does not apply to how raters assess a page. Second, tools and user contributions are main content too. On a product page, the specs, photos, size guide and reviews are the main content. On a mortgage page, the calculator may be the most important part of it.
What is not main content: navigation, footers, ads, cookie banners, related-post widgets and boilerplate that repeats on every page. Those can still spoil the experience, but the quality judgment starts with the part of the page that does its job.
The four main content tests behind EEAT
Each attribute comes with a short definition in Google's guide. Here is what each one means, and a quick way to test your own pages against it.
Effort: how much human work went into it
Google defines effort as "the extent to which human work went into creating the content or the systems powering it." Original analysis, original photography and custom-built tools sit at the high end. The low end is described just as plainly:
Using generative AI to produce large amounts of text without manual oversight or curation represents little to no effort.
Google Search Central
One line in this section catches out a lot of content teams: "Attribution or giving credit to other sources doesn't replace the need for original effort." A well-sourced summary of other people's research is still a summary. Citing sources is good practice, but it does not count as work done.
The phrase "or the systems powering it" matters too. A comparison tool fed by your own product data can show a great deal of effort, even if the text around it is short.
Test: could a capable writer produce this page in an hour from the top five search results? If so, it scores low on effort, however polished it reads.
Originality: what does this page add?
Originality asks whether the content offers "unique, original information or perspectives that aren't already available on other websites." This is where most SEO content falls down. Pages written by studying what already ranks tend to repeat what already ranks, and a page that says what the top ten say gives a rater, or a search system, no reason to prefer it.
Every business has sources of originality, even without a research budget:
- Your own data: prices, lead times, return rates, project results, answers from customer surveys.
- Your own experience: what went wrong on a project, what customers ask before they buy, what you would do differently.
- Your own media: photos of the real product, the real team and the real work, instead of stock images.
- A clear position: a recommendation that says who something is not for, rather than a balanced list of pros and cons.
Test: delete every sentence that also appears, in substance, on the top three ranking pages. What remains is the page's originality.
Talent or skill: is it done well?
The third attribute looks at "whether the content shows the talent, skill, or expertise necessary to provide a satisfying experience for visitors." That covers clear writing, but also production: a readable layout, useful images, a tool that works on a phone. Google adds a fair caveat: "Keep in mind that not all content requires specialized expertise." A recipe blog needs a good cook, not a nutrition degree. A page about tax deductions needs someone who knows tax.
This is where experience and expertise become visible on the page itself. Skill shows in the detail a practitioner includes and a generalist leaves out, in the examples, and in knowing which question the reader will ask next.
Test: after two paragraphs, would a professional in this field recognize the author as one of their own?
Accuracy: is it correct?
For informational pages, Google says, "the content should be factually accurate." The bar rises for YMYL topics, short for Your Money or Your Life: health, money, safety and other subjects that "could significantly impact people's lives or well-being". There, the content "must be highly accurate" and in line with expert consensus.
Accuracy is also where AI-assisted content breaks most often. A draft can be fluent and well structured while getting a price, a date, a legal threshold or a product spec wrong. That is exactly why Google now asks for a manual fact-check of every AI-generated piece before publication.
Test: pick five factual claims on the page. Can you point to a source or a first-hand record for each one, and is each one still true today?
How the four tests map onto E-E-A-T
Google does not draw a one-to-one line between the acronym and the new attributes, so treat this as a working model rather than official doctrine. In practice, the four tests are how E-E-A-T becomes visible on a single page:
- Experience shows up as originality: the photos, data and stories only someone who did the thing could provide.
- Expertise shows up as talent or skill: the depth and precision of the content.
- Authoritativeness lives mostly off the page, in mentions, links, reviews and reputation. The main content tests do not replace it.
- Trust rests on accuracy, and on honesty about who made the page and how. That is where the authorship changes come in.
Fake authors are now officially deception
For years, the standard EEAT tactic was to add an author box. Some sites went further and invented the author: an AI-generated face, a made-up name, a bio with credentials nobody could check. Google's guide now names that practice directly:
Fabricating creator profiles (such as by using AI-generated headshots, made-up names, or false credentials to make content appear as if it was written by human experts) is a form of deception.
Google Search Central
It continues: "Any form of deception makes a page untrustworthy to both users and our automated quality systems, and is a signal of a low-quality page." Note the words automated quality systems. This is not only an instruction for raters.
The guide frames authorship as one of three questions to ask about every page: who, how and why. Who: readers should be able to see who created the content, with bylines where they would expect them and a link to background on the author. How: where readers might wonder how something was made, explain it, including how automation or AI was used and why it helped. Why: the answer should be that "you're creating content primarily to help people", not to attract search engine visits.
The fix is simple, if sometimes uncomfortable: real names, real photos and verifiable credentials, or an honest team byline. A brand byline that tells the truth beats an invented expert every time.
What this means for AI-assisted content
Google still does not ban AI content. The guidance targets two specific failures: volume without oversight, which scores as little to no effort, and publishing without a fact-check, which puts accuracy at risk. We looked at the traffic evidence in what Google actually penalizes in AI-generated content, and the October changes confirm that pattern rather than change it.
A workable rule for content teams: AI can draft, structure and speed up research, but the effort and the originality have to come from people. The data, the photos, the opinion, the edit and the fact-check are what a rater looks for, and they are what a model cannot supply on its own.
Google's John Mueller summed up the attitude on October 29, when asked about tricks for AI visibility. His advice was to "focus on that end goal of having something that search engines and AI systems will want to show to people", adding that "none of this is new." Main content that passes Google's four tests is also what AI search tools prefer to cite, which is why we treat EEAT work and answer engine optimization as one job.
An EEAT audit for your main content
You do not need a quality rater to run these tests. This is the order we work in on client sites, from highest impact down:
- 1Start with the templates that earn money. Product, category, service and location pages usually matter more than the blog. For each one, write down the page's purpose and which parts directly serve it. That is your main content.
- 2Run the originality strip test. Remove what the top three results already say. If little is left, add your own data, photos, examples or a clear recommendation.
- 3Look for proof of effort. Original images, measurements, test results, calculators and comparison tables built from real data all count. Stock photos and rewritten manufacturer copy do not.
- 4Fix authorship. Every article needs a real person or an honest team byline, linked to a page that explains who they are. Remove any author you cannot stand behind.
- 5Fact-check, YMYL pages first. Check prices, figures, legal and medical claims, and dates, and show an updated date when you revise a page.
- 6Check the headline. Google's self-assessment asks whether the page title avoids "exaggerating or being shocking in nature." A clickbait title now works against the page it sits on.
- 7Prioritize with data. Start with pages that lost traffic after an update, and with pages stuck in crawled, currently not indexed, a status that often points to pages Google sees as adding too little.
On ecommerce sites the biggest wins are usually in the product and category templates, where a few changes, such as unique descriptions, real photos and visible reviews, roll out across thousands of pages at once. That is where our ecommerce SEO work tends to begin.
Frequently asked questions
What is EEAT?
EEAT, also written E-E-A-T, stands for Experience, Expertise, Authoritativeness and Trustworthiness. It is the framework Google's quality raters use to judge whether a page and its creator can be trusted on a topic. Google calls trust the most important of the four.
Is EEAT a ranking factor?
Not directly. Google states that E-E-A-T itself isn't a specific ranking factor and that rater data is not used directly in its ranking algorithms. Its systems do use many signals that line up with what raters reward, so improving the qualities EEAT describes still tends to help rankings.
What is main content according to Google?
Main content is any part of a page that directly helps it achieve its purpose. That includes text, images and video, but also tools such as calculators, customer reviews and comments, content inside tabs, and the page's visible title and headings. Navigation, ads and footers are not main content.
What are Google's four main content attributes?
Effort, originality, talent or skill, and accuracy. Google added them to its helpful content guide in October 2026 as the fundamental attributes Search Quality raters are trained to evaluate in a page's main content.
How do I improve EEAT on my website?
Add what only you can provide: original data, real photos, first-hand experience and a clear point of view. Show real authors with verifiable backgrounds, fact-check every page before it goes live, and build your reputation off-site through mentions, reviews and links.
Does AI content hurt EEAT?
Not by itself. Google rates large amounts of AI text without human oversight as little to no effort, and asks publishers to fact-check all AI-generated content manually before publishing. AI-assisted pages that people edit, verify and enrich with original material can meet the same standard as any other page.
Get your main content checked
The October update turns EEAT from a vague idea into four questions you can ask of every page, and most sites fail at least one of them on their most important templates. Request a free SEO audit and we will show you which pages fall short on effort, originality, skill or accuracy, and what to fix first. If the content itself needs rebuilding, that is what our content strategy work is for.