The AI Visibility Score: What It Measures and Why Each Component Matters
SiteNexis's AI Visibility Score is a composite of six sub-scores, each measuring a distinct dimension of how AI systems perceive and retrieve web content. This article explains each component, the formula, and why the weights are set as they are.
The AI Visibility Score is a SiteNexis-derived composite used to describe selected retrieval and citation-readiness signals. It is not a Google metric, an industry standard, an independently validated predictor, or a guarantee that a provider will cite a page. This article documents the current implementation and its limits.
The Six Components and Their Weights
The current repository implementation calculates the score as: Machine Readability × 0.15 + Entity Confidence × 0.20 + Retrieval Readiness × 0.20 + Citation Probability × 0.20 + Semantic Trust × 0.15 + Schema Completeness × 0.10. These weights are SiteNexis design choices for a derived analytical score; they do not establish causal impact on Google or other providers.
Machine Readability Score (15%)
Measures extraction fidelity: how much meaning survives the AI extraction pipeline from raw HTML to a usable semantic chunk. The 15% weight is a SiteNexis scoring choice. It should be interpreted alongside the underlying extraction evidence, not as a validated threshold for citation outcomes.
Entity Confidence Score (20%) — Highest Weight Among Semantic Dimensions
Measures entity clarity, consistency, coverage, and disambiguation within the SiteNexis model. The 20% weight reflects the current score design; the implementation does not establish that a ten-point change causes a predictable change in provider citation rates.
Retrieval Readiness and Citation Probability (20% each)
Retrieval Readiness and Citation Probability are separate SiteNexis diagnostics. Their stored inputs and interpretations can evolve with the audit implementation, so the scores should not be read as provider-private selection rules or as guaranteed outcomes from a content change.
Why Schema Completeness Is Weighted Lowest (10%)
Schema Completeness is weighted at 10% not because it is unimportant, but because most sites with reasonable technical SEO practice have sufficient schema coverage for the score to not be the primary limiting factor. The score weight reflects the marginal impact of schema improvements once a baseline level is achieved. For sites with no schema at all, improving schema completeness produces outsized results relative to the 10% weight — because the baseline was so low that even reaching a moderate level produces significant score improvement across multiple dimensions simultaneously.