Scoring Methodology

How Geol.ai measures AI visibility — every weight, every metric, fully transparent

Why We Publish Our Methodology

AI visibility scoring should not be a black box. Geol.ai publishes its complete scoring methodology so you can understand exactly how your score is calculated, what drives it up or down, and how to improve it.

Our scoring system combines two complementary engines: a Quality Score Engine that evaluates 6 weighted dimensions of your content's AI-readiness, and an AI Visibility Score that measures how well AI models can actually parse and understand your pages.

Design Principles

  • Conservative scoring — we cap mechanical scores to prevent false positives. A score of 75+ is genuinely excellent.
  • AI-validated quality — raw scores are calibrated by an AI quality validator to catch inflated metrics.
  • Actionable output — every score includes specific recommendations and critical issues to fix.
  • No vanity metrics — scores reflect real AI discoverability, not marketing-friendly numbers.

Quality Score Engine

The Quality Score Engine evaluates your content across 6 weighted dimensions, each measuring a distinct aspect of AI-readiness. The final score is the weighted average of all dimensions, normalized to a 0-100 scale.

Scoring Formula

// Weighted average across all 6 dimensions
Quality Score =
(Completeness × 0.20) +
(Richness × 0.15) +
(Freshness × 0.10) +
(Authority × 0.15) +
(Compatibility × 0.20) +
(AI Readability × 0.20)

Weight Distribution

Completeness
20%
Richness
15%
Freshness
10%
Authority
15%
Compatibility
20%
AI Readability
20%

The 6 Scoring Dimensions

Each dimension is scored independently from 0-100, then combined using the weights above. Here's exactly what each dimension measures and how to optimize for it.

Completeness

20% weight

Evaluates whether your structured data contains all required Schema.org fields for its type.

  • JSON-LD required fields (@type, name, description, url)
  • Type-specific fields (Article: headline, datePublished, author)
  • Metadata completeness (title, description, Open Graph, canonical)
  • Schema property coverage for your content type

Richness

15% weight

Measures data depth and content volume — how much useful information AI models can extract.

  • JSON-LD nesting depth (deeper = more relationships)
  • Entity count (10+ entities = maximum score)
  • Keyword density and variety (15+ keywords optimal)
  • Content length (2,000+ characters for full score)

Freshness

10% weight

Evaluates content recency using temporal signals from structured data and metadata.

  • Content age from dateModified or datePublished
  • Under 30 days: full score (100)
  • 30-90 days: recent (80), 90-180 days: consider updating (60)
  • Over 365 days: significant age penalty (20)

Authority

15% weight

Evaluates author and publisher credibility signals that AI models use to assess trustworthiness.

  • Author presence with structured details (@type, name, url)
  • Publisher information with logo and contact details
  • Contact information availability (email, phone)
  • Social media profiles and about/bio sections

Compatibility

20% weight

Measures cross-platform format coverage — how many AI-discoverable formats your site provides.

  • Core formats: JSON-LD, Open Graph, llms.txt (40 points)
  • Recommended formats: RSS, Schema JSON, Manifest (40 points)
  • Optional formats: robots.txt, sitemap, microdata (20 points)
  • 80%+ coverage = outstanding compatibility

AI Readability

20% weight

Evaluates how well your content is structured for AI parsing and comprehension.

  • Keyword coverage (15+ keywords = full score)
  • Entity diversity (count + 3+ distinct types = bonus)
  • Heading structure (single H1, H2 subheadings, hierarchy)
  • JSON-LD syntax validity, @context, @type, required fields

AI Visibility Score

Separate from the Quality Score, the AI Visibility Score is a mechanical measurement of how well AI models can actually parse your page. It's computed from three sub-components and capped at 80 to prevent score inflation.

Visibility Formula

// Mechanical score, capped at 80
AI Visibility = min(80,
(Content Quality × 0.40) +
(Structured Data × 0.30) +
(Metadata Completeness × 0.30)
)

Content Quality

40%
  • Word count via logistic curve (optimal at 800+ words)
  • Entity density (2-15 per 100 words is optimal)
  • Readability (16-22 words per sentence)

Structured Data

30%
  • Schema count: 1 schema = 45pts, 2-3 = 60-75pts
  • Diminishing returns above 4 schemas
  • Spam penalty for 6+ schemas (score decreases)

Metadata

30%
  • Title tag: optimal length 40-70 characters
  • Meta description: optimal 120-160 characters
  • Exactly 1 H1 heading (15pts vs 8pts for multiple)

Why cap at 80? The hard cap prevents unrealistically high scores from purely mechanical analysis. Scores above 80 are reserved for pages that also pass AI-powered quality validation, ensuring that a high score genuinely reflects real-world AI discoverability.

AI-Powered Calibration

Raw mechanical scores can be misleading — a page might check every box but still have poor quality content. Geol.ai uses an AI quality validator to calibrate scores against actual content quality.

How Calibration Works

  • AI evaluates content quality, structured fidelity, and metadata effectiveness
  • Produces a multiplier (0.0-1.0) applied to mechanical scores
  • Catches "optimized but empty" pages that game mechanical checks
  • Global maximum score capped at 95 — perfection is reserved for truly exceptional sites

Multiplier Ranges

0.95-1.0
Exceptional — scores are accurate as-is
0.80-0.94
Good quality with minor issues
0.65-0.79
Average — some score inflation detected
0.50-0.64
Below average — significant inflation
0.00-0.49
Poor quality — major score adjustment

Grade Scale

Final scores are converted to letter grades. Because of our conservative scoring approach, achieving a B or higher indicates genuinely strong AI optimization.

A
90-100

Exceptional. Top-tier AI discoverability across all dimensions.

B
80-89

Strong. Well-optimized with minor areas for improvement.

C
70-79

Average. Functional but missing optimization opportunities.

D
60-69

Below average. Significant gaps in AI readiness.

F
0-59

Poor. Needs immediate attention for AI visibility.

What Makes Scoring Conservative

Cap

Global maximum: 95

95-100 is reserved for truly exceptional sites only

Cap

Mechanical cap: 80

AI Visibility Score cannot exceed 80 without AI validation

Guard

Mismatch detection

High mechanical + low AI quality = capped at 70

Guard

Degraded fallback

If AI validator is down, multiplier defaults to 0.65

Transparency Commitments

We believe scoring methodology should be public, auditable, and continuously improved. Here's what we commit to.

What We Do

  • ✓Publish all scoring weights and formulas on this page
  • ✓Show per-dimension breakdowns in every scan report
  • ✓Provide specific, actionable recommendations (not vague suggestions)
  • ✓Flag when scores are generated in degraded mode
  • ✓Benchmark against known high-quality sites for accuracy

✕What We Don't Do

  • ✕Inflate scores to make results look better
  • ✕Use proprietary "secret sauce" without explanation
  • ✕Change scoring to favor paying customers
  • ✕Hide methodology behind paywalls
  • ✕Allow vanity metrics to override real quality signals

See Your Scores

Run a free scan to see exactly how your site performs across all 6 dimensions. Every report includes per-dimension breakdowns, specific recommendations, and critical issues to fix.

Ready to Boost Your AI Visibility?

Start optimizing and monitoring your AI presence today. Create your free account to get started.