Chinese consumer brands are splitting into two groups in AI search: those that own a concept, and those that are just mentioned. The gap is widening.
A Tale of Two Scores
We ran AI visibility diagnostics on a dozen major Chinese consumer brands. Two results crystallized the divergence happening across the domestic brand landscape:
- Yuanqi Senlin (元气森林): 83/100 — Excellent
- Perfect Diary (完美日记): 71/100 — Good, but not dominant
Both are nationally recognized brands. Both have extensive media coverage and e-commerce presence. But in AI-generated recommendations, they occupy very different positions.
Why Yuanqi Senlin Scores 83
Yuanqi Senlin did not just build a brand — it built a vocabulary.
Zero-sugar sparkling water and Yuanqi Senlin are nearly synonymous in Chinese digital content. When someone asks an AI assistant what is a healthy alternative to soda, the training data is saturated with content linking these concepts to Yuanqi Senlin specifically.
Three things drive this position:
- Category invention: Yuanqi Senlin largely created the mainstream Chinese market for zero-sugar sparkling water. When you invent a category, your brand and the category vocabulary become intertwined in training data.
- Comparative content density: Years of Yuanqi Senlin vs Coca-Cola articles have trained AI to recommend Yuanqi Senlin in specific scenarios.
- Cross-platform depth: From Xiaohongshu posts to Zhihu health discussions, the brand appears in the specific contexts where AI learns to make recommendations.
Why Perfect Diary Scores 71
Perfect Diary is genuinely well-known. Its 71 score means AI assistants are aware of it and mention it in relevant contexts.
But 71 is the score of a brand that AI knows about rather than one it actively recommends.
The challenge is structural. Beauty is a highly competitive category where AI training data is dominated by international brands with decades of comparative content. Domestic Chinese beauty brands are often mentioned as alternatives, not as category leaders.
The Pattern Across Chinese Brands
High GEO Scores (80+) — These brands define a concept, not just a product:
- Pop Mart (泡泡玛特): 89/100 — Blind box and designer toy are near-synonyms for Pop Mart internationally
- Yuanqi Senlin: 83/100 — Category inventor in zero-sugar sparkling water
Mid GEO Scores (65-79) — Known but not owned:
- Perfect Diary: 71/100 — Recognized but competing against deeper international content
The Critical Insight: AI Does Not Reward Fame
The most counterintuitive finding: fame and GEO score are weakly correlated.
A brand can be universally recognized and still score 60/100 on AI visibility. This happens when the brand is famous but does not own specific comparative vocabulary, or most mentions are neutral rather than comparative.
What the High Scorers Did Right
- Own one specific phrase — Be the definitive answer to one specific question
- Generate comparative content — Enable third-party comparison where your brand wins
- Build English-language presence — For global brands, English coverage in authority publications has outsized AI training weight
- Measure it — You cannot optimize what you cannot see
See the full Yuanqi Senlin report: anchor.agentese.ai/r/d65e9394fcb8
Run your own brand AI visibility report at anchor.agentese.ai
Top comments (0)