He is not wrong. In fact, he is entirely correct about the Google-centric, retrospective view of EEAT. But I fear he misses the larger, far more interesting picture.
If there is ever a so often overlooked resource and in so many cases, it is the differentiator, it is people. Yes, people, who would figure the most important marketing and advertising medium, in business, would be people. The inability for people to see outside the box, the fog, the pre sold, broken, programming that spawned the industry we know today as, SEO.
We are no longer just dealing with Google's Quality Rater Guidelines. We are dealing with an evolving LLM world where the entity doing the evaluating is a probabilistic language model. And in this world, EEAT is not a Google concept at all. It is a human concept that Google tried ...... and failed ...... to operationalise, but which LLMs are now perfectly positioned to reward.
What is EEAT — and What It Was Never Meant to Be
E.E.A.T stands for Experience, Expertise, Authoritativeness, and Trustworthiness. It is a vocabulary Google wrote for its human quality raters; the people employed to assess search result quality ...... not a control panel for SEOs.
The 'Experience' dimension was added in 2022, upgrading the original EAT framework to acknowledge that first-hand, lived experience carries a different weight to academic or professional credentials. A plumber writing about pipe corrosion does not need to tell you he is a plumber. You know within two sentences because of what he notices that a content writer never would.
But here is where the industry went catastrophically wrong with E-E-A-T in content writing. They treated the framework as a production checklist. Add an author bio. Mention your years of experience in paragraph one. Slap a trust badge in the footer. The result was worse content that fooled nobody; least of all the quality systems trained on the judgements of actual human readers.
The rater guidelines instruct evaluators to research a website's reputation externally, off the page, before judging quality. Reputation is an entity-level property. A recognisable brand pre-loads the trust evaluation before a single word of body copy is read. An unknown site with an immaculate author bio has pre-loaded nothing.
The Ghost of Authorship Past
To understand why the industry got EEAT so wrong, you have to look at 'history'. Google has been trying to algorithmically measure human credibility for the better part of fifteen years, and every single attempt has ended in a quiet, embarrassing shutdown.
Remember Google Authorship? Launched in 2011, it was the grand promise that rel=author markup would tie your content to your Google+ profile, rewarding verified experts with rich snippets and author photos in the SERPs. It was the ultimate checklist. What did the industry do? 'Remember the intesnse grfters on Google+ sprouting Authority Products'. They gamed it, they spammed it, and webmasters largely ignored the complex implementation. By June 2014, Google stripped the author photos. By August, John Mueller announced the complete shutdown of the program.
Then came Google+ itself, the forced social layer meant to map the world's relationships and expertise. Dead by 2019. Then the 2021 push for Author structured data; another attempt to get webmasters to self-declare their authority.
The pattern is undeniable. Google keeps trying to build a machine that measures trust, and the industry keeps feeding it manufactured signals. David is right that EEAT became a checklist, but he misses why. It became a checklist because that is the only way the 95% know how to operate. They want a 'round peg for a neatly fitting round hole'. The E-E-A-T framework is no different to every other framework Google has ever published ...... it gets weaponised by the people it was designed to filter out.
How Does E-E-A-T Fit in an LLM World?
This is the question David does not ask, and it is the most important question in search right now.
AI systems like ChatGPT, Perplexity, and Claude do not read your author bio. They do not care about your schema markup. They do not use the Quality Rater Guidelines. What they do is build a probabilistic model of who is worth citing, based entirely on the corpus they were trained on and what they can retrieve at inference time.
The E-E-A-T framework, in an LLM world, is operationalised through citation. If you are cited by others, referenced in industry discussions, mentioned in podcast transcripts, and quoted in conference write-ups, you exist as a node of authority in the model's internal representation of your topic. If you only exist on your own website, you are a ghost in the machine.
This is not a new idea. It is the oldest form of credibility in human communication. The academic who is cited by other academics. The practitioner whose name comes up in every serious conversation about their field. The author whose ideas are referenced even by people who disagree with them. EEAT was always describing this. Google just gave it a bad name and the industry turned it into a form to fill in.
And how do you get into the training data? How do you become the entity the model retrieves when asked a complex question? Well, that is not so easy.....
Layer 1 ...... Training data (closed, batch): Large brands, major publications, Wikipedia-level entities. You are not getting in here unless you are operating at serious scale. This is not a realistic target for most businesses.
Layer 2 ...... Retrieval at inference time (open, ongoing): This is where the real opportunity is. Systems like Perplexity, ChatGPT with web search, Google AI Overviews, and Bing Copilot all perform live retrieval (fan-out queries) at the point of answering. They pull fresh, indexed content. If you have enough consistent mentions with freshness across authoritative sources, you can surface in these fan-out queries regularly. This is the realistic mechanism.
The New Game, The Old Penalties
Of course, the unenlightened are already trying to game this new system. We are seeing the rise of 'GEO' tactics; the grift version of SEO, as David rightly calls it; where the goal is to manipulate AI citations through recommendation poisoning, biased best-of listicles, and Reddit spam.
It is the same old trick in a newer costume. And the hammer is already falling.
Bing has been issuing manual actions for scaled content abuse and thin listicle manipulation throughout 2025 and 2026. Google's June 2026 Spam Update explicitly updated its policies to target attempts to manipulate generative AI responses in Search; the first time Google's written rules have named AI-answer manipulation as spam. And as Glenn Gabe documented in July 2026, when Reddit's massive scale of AI-translated content finally got hit by the May 2026 core update, it did not just drop in Google; it dropped heavily in ChatGPT citations downstream. The cascade is real, and it flows both ways.
When you drop in organic search, you drop in AI search. When you are penalised for scaled content abuse, you lose your citations. The ecosystem is more connected than the GEO vendors would like you to believe.
EEAT Principles in Practice: What Actually Works
So, what does applying EEAT principles actually look like in a world where the evaluator is a language model?
It looks like building a genuine external presence. Writing guest articles for publications your audience actually reads. Appearing on podcasts in your field. Producing YouTube content that gets referenced by others. Speaking at events that generate write-ups. Being quoted in industry journalism. Doing the work so well that other people feel compelled to mention you.
It looks like content written from the 'bones of your ass', from real, hard-won experience that a content writer producing twelve articles a week across five industries could never replicate. The calm specificity of a practitioner. The edge case that only someone who has actually done the thing would notice.
It does not look like listicles engineered to game AI citations. It does not look like Reddit spam campaigns. It does not look like recommendation poisoning or biased best-of pages built to manipulate AI Overviews. These are the new version of the old EEAT checkbox game, and they are already being penalised.
The Framework Outlives the Search Engine
EEAT is just a vocabulary Google wrote for human raters. But the concepts Experience, Expertise, Authoritativeness, Trustworthiness are ancient social mechanisms that predate Google by several thousand years.
People trust voices with track records. They trust people who have done the work, who speak with the calm specificity of a practitioner, and who are validated by their peers. LLMs are, at their core, trying to simulate this same human trust evaluation. They are imperfect at it, they can be gamed in the short term, and they will keep getting better at detecting the manufactured version.
If your company is genuinely known, genuinely competent, and your content reads like it was written by someone who has actually done the work, congratulations, you have all the EEAT you are ever going to get. And if it is not and it does not, no checker, no schema, no author bio, no listicle campaign, and no Reddit thread is going to manufacture it for you.
The framework outlives the search engine. It always did.
References
[1] Search Engine Land. (2014, August 28). The Rise & Fall Of Google Authorship. https://searchengineland.com/goodbye-google-authorship-201975
[2] SERoundTable. (2014, August 29 ). Google Completely Drops Authorship Support. https://www.seroundtable.com/google-authorship-dead-19077.html
[3] Sistrix. (2021, August 12 ). Who wrote that content? Google recommends declaring authors in markup. https://www.sistrix.com/blog/who-wrote-that-content-google-recommends-declaring-authors-in-markup/
[4] Google Search Central. (2026, May 15 ). Spam policies for Google web search. https://developers.google.com/search/docs/essentials/spam-policies
[5] Digital Applied. (2026, June 25 ). Google's June 2026 Spam Update: What Site Owners Do Now. https://www.digitalapplied.com/blog/google-june-2026-spam-update-rollout-site-owner-guide
[6] GSQi / Glenn Gabe. (2026, July 9 ). Reddit's AI Translations Finally Drop Heavily In Google And AI Search. https://www.gsqi.com/marketing-blog/reddit-ai-translations-drop/
[7] David Quaid - SEO Professional based in NY USA. LinkedIn @davidquaid
