AI Content, Thin Content, and the GEO Visibility Gap

There’s a quiet double-squeeze happening in search right now. Google is tightening its definition of what counts as thin content — and AI-generated pages are increasingly caught in that net — while simultaneously, a separate battle for citation authority inside LLM-powered answers is being lost by brands who don’t even know they’re competing.

These two pressures aren’t unrelated. They’re two faces of the same shift.

Google’s Thin Content Hammer Is Getting Wider

Search Engine Journal’s Roger Montti flagged something that deserves more attention than it’s getting: Google appears to be expanding its thin content classification to capture AI-generated pages that technically have word count but lack genuine informational depth. This isn’t a new algorithm — it’s an old one being applied more aggressively to a new content type.

The pattern is recognizable to anyone who lived through the Panda era. Content that exists primarily to occupy a SERP position, rather than to resolve a user’s specific information need, gets downweighted. The problem in 2026 is that a significant volume of AI-assisted content — particularly scaled programmatic pages and templated category content — falls squarely into this zone.

For Southeast Asian brands running Shopee or Lazada storefronts with AI-generated product descriptions across thousands of SKUs, this is an immediate operational risk. A product description that restates the spec sheet in different words is thin content. The fact that an LLM wrote it doesn’t change the classification — it may actually accelerate it, since Google’s systems are increasingly trained to recognize the structural signatures of low-effort AI output.

The tactical fix isn’t to abandon AI in content production. It’s to audit for information density: does each page answer something a real buyer actually needs to know that isn’t answered elsewhere on the site? If the honest answer is no, consolidation beats volume.

---

What ‘High Quality’ Actually Means for Algorithmic Trust

SEO.com’s breakdown of high-quality content in 2026 reinforces something that’s obvious in retrospect but routinely ignored in practice: quality signals are multi-layered, and the layers interact. Topical authority, source credibility, user engagement signals, and structured data coherence don’t operate independently — Google’s quality assessment appears to weight them in combination.

For marketers in Southeast Asia, this creates a specific challenge. Many brands operate in multilingual environments — Thai, Bahasa Indonesia, Vietnamese, English — and maintaining content depth across all language variants is expensive. The temptation is to produce high-depth content in one language and auto-translate the rest. Google’s quality signals treat these as separate entities. A low-depth Bahasa Indonesia page doesn’t inherit the authority of its English source.

The strategic implication: prioritize depth where your conversion audience actually reads. For most SEA e-commerce brands, that means mobile-first, vernacular-first, and short-form-first — not a 2,000-word English pillar page that 4% of your traffic will read past paragraph two.

One practical checkpoint that’s underused: internal linking coherence. Pages that are topically isolated — not linked to or from related content — are harder for Google to contextualize, which compounds thin content risk. A content audit that maps topical clusters alongside depth scores will surface more actionable fixes than a word count review alone.

The GEO Visibility Gap Most Brands Are Missing

This is the part that keeps me up slightly more than the others. Semrush’s recent feature for identifying AI visibility gaps — tracking which prompts trigger competitor citations in LLM-powered answers, and where your brand is absent — makes visible a competitive dynamic that’s been building for 18 months without most marketing teams noticing.

The mechanism matters: generative engines like ChatGPT Search, Perplexity, and Google’s AI Overviews don’t retrieve pages in the traditional sense. They synthesize responses from sources they’ve been trained to trust. That trust is built from entity authority — how consistently, accurately, and authoritatively a brand is mentioned across the web — not from keyword density or backlink volume alone.

Semrush’s gap analysis reveals the practical shape of this problem: a brand might rank on page one for a keyword but be entirely absent from the AI-generated answer to the same query. The searcher never scrolls to the blue links. The brand doesn’t exist in that interaction.

For Southeast Asian markets, this is compounded by the relative youth of local entity graphs. A Thai logistics brand or a Filipino fintech may have strong local SEO performance but thin entity coverage in the English-language web corpus that many LLMs weight heavily. Closing that gap means deliberately building brand mentions in authoritative English-language and regional trade publications — not as a link-building exercise, but as an entity-building one.

The audit starting point: run your core commercial queries through ChatGPT Search and Perplexity and note who gets cited. If your competitors appear and you don’t, you have a GEO gap. Semrush’s tooling now surfaces this systematically rather than requiring manual spot-checks.

Connecting the Threads: One Unified Content Strategy

What ties these three signals together is a single underlying shift: search quality assessment — whether by Google’s classifiers or an LLM’s citation logic — is increasingly about genuine epistemic value. Does this brand demonstrably know something? Is that knowledge expressed clearly, accurately, and with enough specificity to resolve a real question?

Brands that built volume-first content strategies over the last two years are now sitting on assets that fail both tests simultaneously: too thin for Google’s quality filters, and not authoritative enough to earn LLM citation. The rebuild isn’t a content calendar refresh — it’s a strategic repositioning of what the brand is willing to be known for, in enough depth to be trusted by both human readers and machine classifiers.

In Southeast Asia’s fragmented, mobile-first, multi-language search environment, that’s a harder brief than it sounds. But the brands that solve it first will own a disproportionate share of both traditional and AI-mediated search visibility.

The open question worth sitting with: if LLMs increasingly mediate the first answer a searcher receives, and that answer doesn’t cite you, does your SEO performance actually translate into brand consideration — or just into traffic that would have found you anyway?

---

At grzzly, we work with growth teams across Southeast Asia on exactly this intersection — diagnosing GEO visibility gaps, auditing content depth against both Google’s quality signals and LLM citation patterns, and building entity authority strategies that hold up across Thai, Bahasa, Vietnamese, and English-language search environments. If your brand is winning on traditional SEO but invisible inside AI-generated answers, that’s a conversation worth having. Let’s talk