Authority Score
The Authority Score is a single 0–100 number that summarises how well a brand is positioned across all the signals the platform measures. A higher score means the brand appears more often, more consistently, and in more authoritative contexts in AI-generated answers and web search.
Important: This is GEO.actor's own proprietary composite metric. It is NOT an industry-standard "domain authority", "page authority", or any score produced by a third-party tool (Moz, Ahrefs, Semrush, etc.). Do not compare it directly to those metrics.
Formula
The score is a weighted average of component scores, each normalised to 0–100:
Authority Score = Σ (component_value_i × scaled_weight_i) / 100When a component cannot be computed (no data yet), it is excluded and the remaining weights are scaled proportionally. This guarantees the score always uses the full 0–100 range even during early data collection.
Components
| Component | Weight | Description |
|---|---|---|
| Share of Model | 35% | Fraction of unbranded prompt answers where the brand is mentioned |
| AI citation share | 15% | Fraction of all AI answers where the brand is cited as a source |
| Referring domain quality | 15% | Average quality of active referring domains |
| Entity consistency | 10% | Consistency of brand entity signals across the web (0–100 audit score) |
| Technical health | 10% | 100 minus penalties for open technical issues |
| AI Overview citation share | 10% | Fraction of Google AI Overview answers citing the brand |
| Co-occurrence share | 5% | How often the brand co-occurs with key topics in AI answers |
Weights sum to 100 when all components are available.
Versioning
Each score row stores a config_version field. When the formula or weights change, the version is bumped. Historical rows retain their original version so trends can be re-interpreted correctly.
How to improve your score
The highest-weight component is Share of Model (35%). Focus here first:
1. Topical authority — Create comprehensive, accurate content covering your subject area
2. Entity clarity — Ensure your brand, products, and key people are unambiguous across the web
3. Technical health — Fix JSON-LD errors, llms.txt gaps, and AI crawler blocks
4. Link quality — Earn citations from trusted, topically relevant sources
5. Co-occurrence — Appear alongside your key topics in credible third-party sources
Reading the trend chart
- Week-over-week score changes reflect changes in underlying components
- A score drop without a content change may indicate a competitor improved
- Hover over any data point to see the component breakdown for that week
Something missing? Let us know.