How We Calculate AI Visibility Scores
Transparency is at the core of the AI Visibility Index. Here's a detailed look at how we measure, score, and rank brand visibility across AI platforms.
1. Data Collection
Our data collection process involves systematically querying major AI platforms with carefully crafted prompts designed to elicit brand mentions and recommendations.
AI Platforms Monitored
- ChatGPT (GPT-4 and GPT-4o)
- Perplexity AI
- Google Gemini
- Google AI Overviews (SGE)
- Microsoft Copilot
For each industry, we use 100+ unique prompts covering different user intents: general recommendations, brand comparisons, feature-specific queries, best-of lists, and problem-solution queries.
2. Scoring Components
The AI Visibility Score is a composite metric derived from four key components, each measuring a different aspect of brand presence in AI responses.
Mention Frequency
How often the brand is mentioned across all monitored queries. Brands appearing in more AI responses score higher. We normalize for total response length to ensure fair comparison.
Position & Prominence
Where in the response the brand appears matters. Brands mentioned first, featured as top recommendations, or highlighted as examples receive higher scores than those mentioned in passing.
Citation Quality
When AI platforms cite sources, we analyze whether the brand is cited from authoritative sources. Citations from industry publications, official sources, and trusted review sites contribute more.
Query Coverage
We measure across how many different query types the brand appears. Brands visible across diverse queries (general, comparative, feature-specific) score higher than those appearing only for specific queries.
3. Score Calculation
AI Visibility Score Formula
Score = (MF × 0.40) + (PP × 0.25) + (CQ × 0.20) + (QC × 0.15)Mention Frequency
Position & Prominence
Citation Quality
Query Coverage
Each component is normalized to a 0-100 scale within the industry before the weighted average is calculated. The final score is rounded to the nearest integer.
4. Share of Voice
Share of Voice (SoV) represents the percentage of total AI mentions a brand receives compared to all competitors in its industry. It's calculated as:
SoV = (Brand Mentions / Total Industry Mentions) × 100A brand with 12.5% Share of Voice in the banking industry means that out of all bank mentions across all AI queries, this brand accounts for 12.5% of them. Higher SoV indicates stronger competitive positioning.
5. Data Quality & Validation
Cross-Platform Validation
We verify brand mentions across multiple AI platforms to ensure consistency and filter out platform-specific anomalies.
Hallucination Filtering
Our systems detect and filter AI hallucinations by cross-referencing with verified brand databases and industry records.
Statistical Normalization
Scores are normalized using statistical methods to ensure fair comparison across different industry sizes and query volumes.
Regular Audits
We conduct weekly audits of our data collection and scoring systems to maintain accuracy and reliability.
6. Update Frequency
The AI Visibility Index is updated daily to reflect the latest AI response patterns. This allows brands to:
- Track visibility trends over time
- Identify sudden changes in AI visibility
- Measure impact of visibility optimization efforts
- Monitor competitive movements in real-time
Glossary of Terms
Frequently Asked Questions
How do you collect data from AI platforms?
We systematically query major AI platforms (ChatGPT, Perplexity, Gemini, AI Overviews) with hundreds of industry-relevant prompts. These prompts cover various user intents including brand comparisons, product recommendations, industry overviews, and specific feature queries. We analyze the responses to identify brand mentions, citations, and recommendations.
How often is the data updated?
Our AI Visibility Index is updated daily. We continuously monitor AI responses to track changes in brand visibility over time. This allows brands to see trends, identify sudden changes, and measure the impact of their visibility optimization efforts.
What makes a brand score high on AI visibility?
High-scoring brands typically have: (1) Strong online presence with authoritative content, (2) Frequent mentions and citations across trusted sources, (3) Consistent brand information across the web, (4) Good coverage in industry publications and reviews, (5) Positive sentiment in customer feedback and reviews.
How do you handle AI hallucinations or errors?
We employ multiple validation techniques to ensure accuracy. This includes cross-referencing mentions across multiple AI platforms, filtering out obvious hallucinations, and using statistical methods to identify anomalies. We also regularly audit our data against known brand information.
Can brands influence their AI visibility score artificially?
The AI Visibility Score reflects genuine brand presence across the web. While brands can improve their scores through legitimate means (quality content, building citations, improving online presence), artificial manipulation attempts are unlikely to be effective as AI models draw from diverse, authoritative sources.
How do you ensure fairness across industries?
Scores are calculated relative to competitors within each industry. This means a bank with a score of 80 is compared against other banks, not against mobile phone brands. We use industry-specific query sets to ensure relevant comparisons.
See the Methodology in Action
Explore the AI Visibility rankings for your industry and see how our methodology translates into actionable insights.