Back to Help Center
Getting Started8 min readUpdated 2024-12-01

How the AI Perception Score Works

A detailed breakdown of our scoring methodology.

The AI Perception Score

Your AI Perception Score is a 0-100 metric that quantifies how well AI systems understand and represent your brand. Higher scores indicate better AI visibility and more accurate representation.

Score Components

1. Visibility (25%) Measures how often your brand appears in AI responses for relevant queries.

- High visibility: Brand mentioned in 70%+ of relevant queries - Medium visibility: Brand mentioned in 40-70% of relevant queries - Low visibility: Brand mentioned in less than 40% of relevant queries

2. Accuracy (25%) Evaluates whether AI-provided information about your brand is correct.

- Company facts (founding date, location, size) - Product/service descriptions - Pricing information - Leadership and key people

3. Sentiment (20%) Analyzes the tone and context of how AI describes your brand.

- Positive: Recommended, praised, highlighted as leader - Neutral: Mentioned factually without strong opinions - Negative: Associated with problems, complaints, or warnings

4. Authority (15%) Measures trust signals that AI models recognize.

- Citation in authoritative sources - E-E-A-T signals (Experience, Expertise, Authority, Trust) - Industry awards and certifications - Expert endorsements

5. Competitive Position (15%) Compares your brand perception against competitors.

- Share of voice in category queries - Recommendation likelihood vs. competitors - Feature comparison accuracy

Score Grades

Score Range Grade Meaning |-------------|-------|---------| 90-100 Excellent Industry-leading AI presence 75-89 Good Strong visibility with minor gaps 60-74 Average Room for significant improvement 40-59 Poor Urgent optimization needed 0-39 Critical Serious AI perception issues

Provider-Specific Scores

We analyze your brand across multiple AI platforms:

- OpenAI GPT-4: The most widely used AI assistant - Anthropic Claude: Growing rapidly in enterprise - Google Gemini: Integrated with Google Search - Perplexity AI: Popular for research queries

Each provider may have different perceptions based on their training data and methods.

How Scores Are Calculated

1. Query Generation: We identify relevant queries for your brand 2. Response Collection: Gather AI responses across providers 3. Analysis: NLP analysis of mentions, sentiment, accuracy 4. Normalization: Scores adjusted for industry benchmarks 5. Aggregation: Component scores combined with weights

Improving Your Score

Focus on areas with the lowest sub-scores first. Common improvement strategies:

- Low Visibility: Increase authoritative mentions and citations - Low Accuracy: Update official sources with correct information - Low Sentiment: Address negative reviews and build positive proof - Low Authority: Build E-E-A-T signals and expert content - Low Competitive: Create differentiation content

Was this article helpful?

How the AI Perception Score Works - Help Center