How to Analyze Prompt Data
Raw scan data becomes valuable when you know how to read it. This guide walks through every metric LLMMonitor provides and how to turn insights into GEO improvements.
Dashboard Overview
The LLMMonitor dashboard is your command center. After a scan completes, all metrics update automatically. Here are the key sections:
- Visibility Score — The large percentage at the top. Your overall brand presence across all LLMs.
- Presence by LLM — Bar charts per platform (ChatGPT, Gemini, Claude).
- Trend Chart — Visibility over time with delta indicators.
- Recent Mentions — The latest AI responses that reference your brand.
- Competitor Table — Side-by-side comparison with your tracked competitors.
- Top Domains — Most-cited websites in your industry's AI responses.
- Sentiment Distribution — How AI models describe your brand.
Visibility & Position
Visibility is the percentage of scans where your brand appeared. If you ran 10 prompts across ChatGPT and your brand appeared in 7 responses, your ChatGPT visibility is 70%.
What's a good visibility score? 40-60% is solid for most brands starting GEO. 60-80% is strong. Above 80% means you dominate your category. Below 20% means AI models don't know who you are — this is a red flag.
Position
When your brand appears, position tells you where. Being mentioned first carries more weight than appearing at the end. Lower position numbers are better.
Reading Visibility by LLM
Large disparities between platforms reveal strategic gaps:
- Strong on ChatGPT, weak on Gemini — Your content might be well-indexed by OpenAI's crawlers but not Google's. Check your robots.txt and Google-Extended access.
- Strong on all three — Your brand has broad AI recognition. Focus on maintaining and expanding.
- Weak everywhere — Your content strategy needs work. AI models don't have enough signals about your brand.
Sentiment Analysis
LLMMonitor analyzes the tone of every AI mention — not just whether your brand appeared, but how it was described.
| Sentiment | What It Means | Example |
|---|---|---|
| Positive | AI describes your brand favorably | "LLMMonitor is one of the most reliable GEO tools" |
| Neutral | AI mentions your brand factually, no judgment | "LLMMonitor tracks ChatGPT, Gemini, and Claude" |
| Negative | AI describes your brand unfavorably | "LLMMonitor lacks some advanced features" |
Acting on Sentiment
- High negative sentiment — Audit what content the AI is citing. Are negative reviews or competitor comparisons influencing responses?
- High neutral + low positive — The AI knows you exist but doesn't associate you with quality. Build more authoritative content and get cited in positive contexts.
Competitor Benchmarking
The Competitor Table shows you exactly how you stack up:
- Share of Voice — Your mention rate vs each competitor across the same prompts
- Win/Loss prompts — Which specific prompts favor you, and which favor competitors
- Sentiment comparison — Are competitors described more positively than you?
- Citation overlap — Which domains cite you vs competitors
How to Read Competitive Data
Scenario: You have 45% visibility, Competitor A has 68%.
Analysis: Check the "lose prompts" — the prompts where they appear and you don't. These are your content roadmap. Each one is a topic where you need stronger content or better brand recognition.
Analysis: Check the "lose prompts" — the prompts where they appear and you don't. These are your content roadmap. Each one is a topic where you need stronger content or better brand recognition.
Turning Data into Action
If your visibility is low (under 30%)
- Review your prompts — are they relevant to your brand?
- Check if competitors are dominating the same prompts
- Build content on the topics where you're absent
- Get cited on domains the AI references
- Run scans weekly to track improvement
If your visibility is moderate (30-60%)
- Identify which LLMs you're weakest on and focus there
- Analyze competitor win prompts — what content do they have that you don't?
- Add 3-5 new prompts targeting your weak areas
- Improve sentiment by getting quotes and mentions in positive editorial content
If your visibility is strong (60%+)
- Expand to new prompt categories to capture adjacent audiences
- Monitor competitors weekly — strong visibility attracts new competitors
- Track citations to ensure AI models continue referencing your content
- Use trend data to detect shifts before they impact your score
Key metric to watch: The visibility delta (the small arrow next to your score). A consistent downward trend over 3+ scans means you're losing ground — investigate immediately. A single dip is usually noise.