Why SEO Gurus Are Losing Ground to Entity Authority

There’s a quiet audit happening inside every large language model right now — and most SEO teams are failing it without knowing they’ve been assessed.

Google’s John Mueller recently told Search Engine Journal that people who self-identify as SEO gurus are, in his words, “clueless imposters.” It’s a pointed jab, but the real story isn’t Mueller’s candour. It’s why that kind of surface-level SEO posturing is becoming structurally obsolete — and what’s replacing it.

The Guru Model Breaks Down in Generative Search

Traditional SEO operated on a relatively forgiving substrate. You could rank for a keyword with the right density, backlink mix, and technical hygiene, even without deep topical authority. The algorithm had gaps you could exploit. Mueller’s frustration with self-proclaimed gurus points to a generation of practitioners who built careers on those gaps.

Generative search doesn’t have the same gaps. When a user asks Google’s AI Overview or a ChatGPT-powered search interface about the best e-commerce platform for SMBs in Vietnam, the model doesn’t scan a keyword match — it draws on its trained understanding of which entities are credible, cited consistently across trusted sources, and semantically associated with the right concepts. Keyword tricks don’t show up in that graph. Entity authority does.

The implication for Southeast Asian brands is sharper than it might seem elsewhere. Markets like Thailand, Indonesia, and the Philippines are experiencing rapid adoption of AI-assisted search — particularly on mobile — at the same time as local SEO practices are still catching up to basic structured data implementation. The gap between brands building genuine entity authority and those chasing short-term rankings is widening fast.

What Entity Authority Actually Requires

Entity authority isn’t a single metric. It’s the accumulated weight of how your brand, your people, and your ideas are represented across the structured and unstructured web — Wikipedia entries, Wikidata nodes, consistent NAP data, schema markup, authoritative third-party mentions, and the semantic coherence between what you say about yourself and what others say about you.

For a practical starting point: run your brand name through Google’s Knowledge Graph Search API. If your entity returns sparse or inconsistent data, you have a foundational problem that no amount of on-page optimisation will fix in a generative search context. From there, the work involves building citation depth — not link volume — by ensuring your brand is accurately and consistently represented on platforms that LLMs actively index: industry publications, credible directories, LinkedIn company pages with complete structured data, and press coverage that explicitly names and contextualises your expertise.

Shopee’s dominance in AI-generated shopping recommendations across Southeast Asia isn’t accidental. It’s the result of years of consistent entity signals: structured product data, localised platform presence, and high-frequency citation in regional commerce coverage. That’s entity authority at scale.

---

The AI Vulnerability Nobody Is Connecting to SEO

Google CEO Sundar Pichai recently acknowledged, in a discussion reported by Search Engine Journal, that AI models could expose more software vulnerabilities and agreed it was plausible that AI is affecting zero-day exploit markets. The comment was framed around cybersecurity — but there’s an underappreciated SEO dimension here.

As LLM-powered systems become more deeply integrated into search infrastructure, the attack surface for brand manipulation expands. Prompt injection, entity spoofing, and synthetic citation networks are emerging threats to brand discoverability that most SEO teams are not equipped to monitor. A competitor — or bad actor — could theoretically flood an LLM’s training context with misleading associations around your brand entity, subtly degrading how AI systems represent you in generated responses.

This isn’t hypothetical anxiety. It’s the logical extension of how generative models work: they synthesise consensus from their training data. If that data becomes polluted, your entity representation suffers. Brands that invest in authoritative, high-signal content now are building a defensive moat, not just an offensive SEO strategy.

For Southeast Asian teams managing multilingual brand presence across Thai, Bahasa, Vietnamese, and English simultaneously, this compounds quickly. Inconsistent entity representation across language variants is already one of the most common structural weaknesses we observe — and it’s exactly the kind of signal gap that generative models penalise.

What Good Looks Like in 2026

The shift from keyword-centric to entity-centric SEO requires a different operational model. Three practical moves that are underutilised in the region right now:

Structured data at the content layer. Every author, product, service, and location associated with your brand should have explicit schema markup — not just on homepages, but on every published asset. Google’s ability to resolve entities depends on consistent structured signals, and most SEO audits in Southeast Asia still treat schema as a technical checkbox rather than a strategic input.

Third-party citation campaigns, not link-building campaigns. The distinction matters. A link from a low-authority site adds almost nothing to entity authority. A mention — even an unlinked one — in a credible regional publication like KrASIA, Tech in Asia, or a local trade vertical builds the co-citation patterns that LLMs use to understand your brand’s domain relevance. Reframe your PR and content distribution strategy around earning those mentions deliberately.

Monitored AI representation. Run your brand name and core offerings through major LLM interfaces monthly. What does ChatGPT say about you? What does Gemini’s AI Overview surface? Discrepancies between your intended positioning and your AI-generated representation are the new ranking gap — and unlike traditional SERP positions, they’re harder to diagnose without deliberate monitoring.

The SEO gurus Mueller dismissed were optimising for a search environment that is, bluntly, receding. The brands that will own generative search discoverability in Southeast Asia over the next three years are the ones building entity authority quietly, systematically, and before most of their competitors have noticed the rules changed.

The open question worth sitting with: if your brand disappeared from every AI training dataset tomorrow, how confident are you that it would be reconstructed accurately from the signals you’ve left behind?

---

At grzzly, we work with growth teams and marketing directors across Southeast Asia to audit entity authority gaps and build GEO strategies that hold up as generative search becomes the default. If your brand isn’t showing up the way you’d expect in AI-generated answers, that’s a solvable problem — and the window to solve it ahead of the curve is still open. Let’s talk