> ## Documentation Index > Fetch the complete documentation index at: https://trakkr.mintlify.site/llms.txt > Use this file to discover all available pages before exploring further. # Understanding Scores > Learn how Trakkr calculates visibility and presence scores, how AI brand tracking works, and how to interpret your competitive landscape Trakkr measures your brand's performance across AI models using two key metrics: **Visibility Score** and **Presence Score**. This guide explains how these scores work, how we track your brand, and what your competitive landscape means. ## Visibility Score Your Visibility Score measures **how prominently** your brand appears when AI models mention you. It ranges from 0-100, with higher scores indicating better positioning. ### How Visibility is Calculated Each mention gets points based on ranking position: * **1st place**: 10 points * **2nd place**: 9 points * **3rd place**: 8 points * **10th place**: 1 point * **Not mentioned**: 0 points We sum points from all AI models (OpenAI, Anthropic, Gemini, Grok, DeepSeek, Meta) for each prompt Your total points ÷ maximum possible points × 100 = raw visibility percentage We apply a square root scaling to make score differences more meaningful and prevent artificial inflation ### Visibility Score Examples **Score 90-100**: Exceptional performance * Consistently ranked #1-2 across most AI models * Strong brand recognition in your industry **Score 70-89**: Strong performance * Usually in top 3-5 positions when mentioned * Good brand awareness with room for improvement **Score 50-69**: Moderate performance * Mixed rankings, often in positions 5-8 * Indicates opportunities for optimization **Score 0-49**: Needs improvement * Low rankings when mentioned, or infrequent mentions * Consider prompt strategy and brand positioning A visibility score of 88.1 (like Nike's example) means when AI models mention Nike, they typically rank it very highly - usually in the top 2-3 positions. ## Presence Score Your Presence Score shows **what percentage** of AI queries mention your brand at all. This represents your "share of voice" in AI conversations about your industry. ### How Presence is Calculated Presence Score = (Number of AI responses mentioning your brand ÷ Total AI responses) × 100 **Example**: If you run 5 prompts across 6 AI models (30 total queries) and your brand appears in 25 responses: * Presence Score = (25 ÷ 30) × 100 = 83.3% ### Presence Score Interpretation **80-100%**: Dominant presence * AI models consistently think of your brand for industry queries * Strong market awareness and top-of-mind positioning **60-79%**: Good presence * Regularly mentioned but some gaps in coverage * Solid brand recognition with growth opportunities **40-59%**: Moderate presence * Inconsistent mentions across different queries * Brand awareness varies by context **0-39%**: Limited presence * Rarely mentioned by AI models * Indicates low brand awareness or need for different prompts ## How AI Brand Tracking Works Understanding how we collect and analyze data helps interpret your scores: ### 1. Multi-Model Querying We track your brand across 6 leading AI platforms: GPT-4 models powering ChatGPT - most widely used AI platform Claude models - popular in enterprise and research Integrated across Google services and products Real-time information access, popular on X/Twitter Advanced reasoning capabilities for complex queries Open-source models widely adopted across platforms ### 2. Daily Report Generation * **Automated daily reports** query all AI models with your prompts * **Each prompt** is sent to all 6 AI models independently * **Results are analyzed** within 30 seconds to 2 minutes * **Scores are calculated** and historical trends updated ### 3. Name Mapping & Competitor Detection **Smart Name Recognition** * AI models might mention your brand using variations ("Apple Inc." vs "Apple") * We use AI to map different mentions to your canonical brand name * Handles typos, abbreviations, and alternate names **Automatic Competitor Discovery** * Brands mentioned alongside yours are identified as potential competitors * Must appear in **2+ reports** to be added to your competitor list * Helps you discover emerging competitors you might not know about ### 4. Scoring Methodology **Why We Use Scaling** * Raw percentage scores can be misleading (small differences appear huge) * Square root scaling makes meaningful differences more apparent * Prevents artificial score inflation while maintaining accuracy * Scores are capped at 100 to maintain the 0-100 scale ## Understanding Your Competitive Landscape ### Competitor Analysis The Competitors section shows other brands AI models mention when discussing your industry: **Competitor Presence Scores** * Shows how often each competitor appears compared to you * Higher percentages indicate stronger AI "share of voice" * Helps identify market leaders and emerging threats **Competitor Rankings** * While we don't show competitor visibility scores, you can see their presence * Use this to understand the competitive landscape * Identify brands that might not be on your radar ### Market Position Insights **Dominant Position**: High visibility + High presence * You're a market leader in AI conversations * Consistently mentioned and highly ranked **Niche Leader**: High visibility + Lower presence * Strong position in specific contexts * Opportunity to expand into broader industry conversations **Emerging Player**: Lower visibility + Higher presence * Frequently mentioned but not prominently positioned * Focus on improving ranking and brand strength **Specialist**: Variable scores across different prompts * Strong in some areas, weak in others * Tailor strategy to strengthen weak areas ## Factors That Influence Your Scores ### Content & Online Presence * **Quality content** that establishes authority and expertise * **Industry recognition** through awards, press, partnerships * **SEO optimization** helps AI models understand your positioning * **Active social media** and thought leadership ### Brand Recognition * **Market maturity** - established brands typically score higher * **Industry category** - some sectors have more AI awareness * **Geographic presence** - global brands often outperform local ones * **Marketing investment** in brand awareness campaigns ### Prompt Strategy * **Relevant prompts** that match your target market * **Balanced specificity** - not too broad, not too narrow * **Multiple contexts** to capture different use cases * **Regular updates** to reflect market changes Scores naturally fluctuate as AI models update their training data and as the competitive landscape evolves. Focus on long-term trends rather than daily variations. *** ## Improve Your Scores Understanding your scores is just the beginning. Here's how to actually improve them: **Start here**: Fix crawlability and technical issues affecting AI understanding Get page-specific recommendations to boost rankings and visibility Create better prompts that capture your true market position Discover 500+ relevant queries to expand your tracking ## Analyze Your Performance Master your dashboard and track performance trends See how your scores compare to competitor performance Understand which sources influence your visibility Track qualitative perceptions alongside quantitative scores