claim-density

What this datapoint measures

Frequency of factual claims per unit of content. Whether the brand’s content makes specific assertions that AI systems can extract as citable claims, or whether the content is dominated by description, opinion, or narrative without specific extractable claims.

This datapoint is closely related to content-depth but measures something different. Content-depth measures substance; claim-density measures the frequency of specific assertions. Deep content can have low claim-density (essay-style depth without specific claims); shallow content can have high claim-density (a list of claims without supporting depth).

What high looks like

• Specific factual claims appearing regularly throughout content
• Claims that are extractable as standalone statements
• Claims supported by data, examples, or attribution
• Mix of claim types (descriptive facts, comparative facts, causal claims)
• Claims that are specific enough to be verifiable (named entities, specific numbers, defined timeframes)

What low looks like

• Content dominated by general descriptions without specific claims
• Vague assertions (“many users prefer,” “industry experts agree”) without specifics
• Opinion-heavy content without factual grounding
• Repetitive content where the same general claim is restated rather than supplemented with new claims

What at floor looks like

A brand at floor on claim-density has content that reads as descriptive narrative without specific extractable claims. AI systems retrieving from this brand find content that establishes the brand exists and has positions but cannot extract specific facts to cite.

This pattern is common in marketing-heavy content (which prioritizes emotional appeal over claim-density), in brands with thin original research, and in editorial cultures that emphasize narrative over factual precision.

What affects this datapoint

• Frequency of factual claims per content section
• Specificity of claims (named entities, numbers, dates)
• Extractability of claims as standalone statements
• Mix of claim types
• Supporting evidence or attribution for claims

OMG actions that influence this datapoint

Action	Influence
M-6 Evidence-Based Content & Citation Architecture	Direct, primary. M-6 work emphasizes specific claims with attribution.
G-8 Original Research & Proprietary Data	Substantial. Original research provides specific claims that lift density across content.
G-3 Comprehensive Long-Form Content	Substantial. Long-form content with proper structure naturally produces more claims per unit.

Multilingual considerations

Per-language claim-density is assessed in the page’s content language. Conventions differ:

• English editorial culture varies between marketing-heavy (low claim-density) and journalistic (high claim-density) traditions
• Japanese and Korean business writing tends toward more cautious claim-making; what reads as low claim-density in English may be appropriate in Japanese
• Indonesian and Traditional Chinese have varying conventions

Per-language assessment should account for these differences while maintaining the underlying principle: AI systems benefit from specific extractable claims regardless of the prose style around them.

Common failure modes

• Marketing copy with generic claims (“the leader in,” “the most trusted”) that are not extractable as factual claims
• Content that describes positions without making them specific
• Vague comparative claims (“better than,” “more efficient”) without specifying the comparison
• Repetition of the same claim throughout instead of progression of claims

Diagnostic interpretation

Claim-density at floor with content-depth also low indicates broad M-pillar work needed. The remedy is editorial discipline targeting both depth and specific claims.

Claim-density at floor with content-depth at high indicates a brand with deep but vague content. M-6 work specifically lifts claim-density without requiring more content production.

Claim-density at high with source-attribution-quality at low indicates a brand making many specific claims without attributing them. The claims are extractable but their reliability is unverified. M-6 attribution discipline is the remedy.

---