- A new SEO debate questions whether GEO rewards real expertise or simply repeats the content patterns that LLMs already recognize.
- The argument exposes a bigger problem for marketers: AI search citations are real, but the signals behind them are still far less predictable than classic rankings.
The debate around generative engine optimization is getting louder, but not necessarily clearer.
A post from AI SEO consultant David Quaid has pushed one of the industry’s biggest GEO assumptions back into focus: do AI search systems actually reward expertise, trust and authority, or do they mostly repeat patterns that already appear frequently across the web?
EEAT is not preferred by GEO – GEO is just recycling anything that resonates as confirmation bias
This is LLM content.
Anyone who promotes EEAT should kind of be ashamed of this. pic.twitter.com/HEppo9Z1Bf
— David Quaid – AI SEO (@DavidGQuaid) May 20, 2026
In a post on X, Quaid criticized the idea that E-E-A-T is naturally preferred by GEO, arguing that generative engine optimization can end up recycling content that resonates as confirmation bias. His post responded to a Reddit thread claiming that GEO is not just a rebrand of SEO, but a different discipline focused on citations rather than rankings. Quaid’s post reflects a larger tension inside the SEO industry: GEO advice is spreading quickly, but many of its claims are still difficult to prove.
The GEO promise is simple
The basic GEO argument is easy to understand.
Traditional SEO optimizes for ranking. GEO, at least in theory, optimizes for being mentioned, cited or used inside AI-generated answers.
That difference matters. In an AI Overview, ChatGPT answer, Gemini response or Perplexity result, users may never scroll through ten blue links. They may read the generated answer, accept the summary and move on. In that environment, being used as a source can be more valuable than sitting somewhere below the answer box.
That is why many marketers now argue that brand entities, author trust, original data, expert quotes and structured answers are becoming the new visibility signals.
There is some logic to that. Google has long used quality concepts such as experience, expertise, authoritativeness and trust in its search quality rater guidelines. In 2022, Google expanded E-A-T to E-E-A-T by adding “experience,” explaining that some content is more useful when it comes from people with first-hand involvement in the topic. Google Search Central said the update was meant to help evaluate whether results demonstrate useful qualities such as expertise, authoritativeness and trustworthiness.
But that does not automatically prove that AI-generated search answers consistently prefer E-E-A-T-rich content in a simple, measurable way.
That is where the GEO story starts to get messy.
AI citations are not just organic rankings with a new label
One reason GEO is hard to prove is that AI citations do not behave exactly like classic organic results.
A 2026 academic study on Google AI Overviews found that almost 30% of domains cited in AI Overviews did not appear in the co-displayed first-page organic results. The authors concluded that AI Overview source selection appears to be distinct from Google’s traditional ranking algorithm. The study also found that 11% of atomic claims in AI Overview responses were unsupported by the cited pages.
Another 2026 study comparing Google Search, Gemini and AI Overviews found that retrieved sources can differ substantially between systems. The researchers reported low source overlap, inconsistent AI Overview behavior across repeated runs and sensitivity to small query changes. Their findings suggest that generative search systems retrieve and present information differently from traditional search, which makes visibility harder to measure and optimize.
That is the problem with treating GEO as a clean playbook.
Many GEO recommendations assume a relatively stable system: improve authority, structure content clearly, publish original data, build brand mentions and win more AI citations. Those steps may help. Some are simply good content strategy. But they do not yet add up to a predictable formula.
Traditional SEO has imperfect but visible feedback loops: rankings, clicks, impressions, CTR and crawl data. GEO visibility is less stable. A brand can appear in one answer, disappear in another and be cited differently depending on the prompt, model, location, personalization and wording of the query.
The risk: confusing repetition with trust
This is where Quaid’s criticism becomes useful.
If GEO becomes a game of producing content that LLMs can easily repeat, marketers may start confusing “AI repeated this” with “AI trusted this.”
Those are not the same thing.
A brand can be cited because it is authoritative. But it can also be cited because its framing is simple, widely repeated, easy to summarize or already embedded in the material AI systems retrieve.
That does not make E-E-A-T irrelevant. It means E-E-A-T is not the whole story.
Strong expertise, clear authorship, original evidence and source-backed claims still matter. They make content more defensible for users and more useful for search systems. But AI systems may also be influenced by retrievability, repetition, entity recognition, source format, brand prominence and how information is distributed across the web.
That is an uncomfortable reality for marketers who want a clean answer.
The content that is most often repeated by AI is not always the content with the deepest expertise. Sometimes it is simply the content that fits the model’s expected pattern best.
E-E-A-T is a quality framework, not an AI citation guarantee
E-E-A-T is still useful. But it should not be oversold as a direct GEO switch.
Google itself has repeatedly framed E-E-A-T as part of its quality rater guidelines, not as a simple ranking factor that can be optimized in isolation. Google’s explanation of E-E-A-T says the guidelines are used by search raters to evaluate the performance of ranking systems and do not directly influence ranking.
That distinction matters even more in AI search.
When marketers say “build E-E-A-T for GEO,” they may be directionally right from an editorial standpoint. But they are often skipping over the harder question: how exactly does an AI system retrieve, weigh and cite sources for a given answer?
In many cases, the honest answer is still unclear.
That tension also showed up in our analysis of Google’s AI search advice. Google’s public message is that AI search does not require a separate optimization playbook. The GEO debate shows why many SEOs are still skeptical: AI citations may overlap with traditional SEO signals, but they do not behave exactly like classic rankings.
The industry is turning GEO into a product before the evidence is mature
The SEO industry moves fast when a new acronym appears.
GEO is already being packaged into frameworks, audits, tools, dashboards and consulting offers. Common recommendations include writing direct answer blocks, adding original research, improving author bios, strengthening entity signals, using structured headings and making content easier for AI systems to extract.
Some of that advice is sensible. A page that clearly states the answer, supports claims with evidence and identifies the expert or brand behind it is useful for both users and machines.
The problem starts when these tactics are presented as guaranteed levers for AI visibility.
Google’s own AI Overviews documentation warns that generative AI can make mistakes and advises users to check important information in more than one place. Google says AI Overviews can provide a starting point with links to explore further, but also notes that generative AI is experimental and can be inaccurate.
That should make marketers more careful about treating AI answers as perfectly rational citation engines.
They are not simply ranking pages by trust. They are producing answers through a mix of retrieval, model behavior, source selection and generation. That system can change quickly and may not behave consistently across prompts.
GEO is real, but the measurement layer is weak
AI search visibility is real. Brands are being mentioned inside AI Overviews, ChatGPT, Gemini, Perplexity and other answer engines. Those mentions can shape perception even when they do not produce a direct click.
But the measurement layer is still immature.
Unlike SEO, where Google Search Console shows query impressions, clicks and average position, GEO lacks a standard source of truth. Tools are emerging, but they often test different prompts, locations, models and answer formats.
That creates a dangerous gap between confidence and evidence.
A brand may improve its AI visibility after publishing more structured content. But what caused the change? Was it the structure? Brand mentions? Backlinks? Freshness? Third-party citations? Model updates? Prompt variation? In many cases, it is difficult to isolate the cause.
This is why GEO reporting should be treated as directional, not definitive.
What marketers should actually do
The smart response is not to reject GEO. It is to stop treating it like a solved discipline.
Brands should keep building content that is useful, specific, well-sourced and easy to understand. They should also structure that content so AI systems can parse it without guessing.
Practical steps still make sense:
- Use clear headings that describe the question being answered.
- Put a direct answer near the top of important pages.
- Support important claims with original data, examples or credible sources.
- Make the author, brand and topical expertise easy to identify.
- Build pages around real user questions, not just keyword variations.
- Track AI mentions across multiple tools and prompts, but treat the data as directional.
- Compare AI visibility with classic SEO data instead of replacing one with the other.
But marketers should be careful with one claim: that E-E-A-T alone will win GEO.
Trust matters. Expertise matters. Originality matters. But AI citation behavior is still too opaque and unstable to reduce to a simple formula.
The Query Post view
The strongest takeaway from the latest GEO debate is not that E-E-A-T is useless or that GEO is fake.
The real takeaway is that the industry is ahead of the evidence.
AI search is changing how information is surfaced, summarized and cited. But the mechanisms behind those citations remain much harder to observe than traditional rankings. That makes the space vulnerable to overconfident playbooks, recycled advice and claims that sound more certain than they are.
Quaid’s criticism lands because it challenges the industry’s current comfort zone. If GEO is sold as “build E-E-A-T and get cited,” it risks becoming another simplified framework before it is properly understood.
The more honest view is less convenient.
GEO is not just SEO with a new name. But it is also not a solved discipline. It sits somewhere between classic search optimization, brand authority, content structure, entity recognition and the unpredictable behavior of generative systems.
For now, the best strategy is not to chase every GEO slogan.
It is to build content that deserves to be trusted, structure it so machines can understand it and stay skeptical of anyone claiming they already know exactly how AI citations work.
