AI Visibility
Prompt Clusters
See your AI visibility for each prompt category, find where competitors lead, and spot the categories that need work.
3 min read Updated
What is it?
The Prompt Clusters screen (page title Prompt Category Coverage) groups your prompts by category. AI Visibility is the LLM Tool in your Prism sidebar. The screen describes itself as "AI-based intent bucketing…". For each category it shows your coverage and the leading brand.
Why use it?
- To see which topics you own in AI answers.
- To find the topics where a competitor leads by a wide gap.
- To decide which content to write next.
Before you start
Your prompts need categories. Add a Category for each prompt in your prompts file. See Keywords and prompt files.
How to use it
Open Prompt Clusters
In the sidebar, open LLM Tool, then click Prompt Clusters.
Set your filters
Pick a project and a date. You can filter by platform, category or sub-category.
Check best and worst
Read 🏆 Best Performing Category and 🔴 Worst Performing Category — Act Now.
Scan the grid
Look through All Categories (N) for High Gap and Critical Gap cards.
Understanding the screen
Best and worst category
🏆 Best Performing Category and 🔴 Worst Performing Category — Act Now appear when there are at least 4 categories.
All Categories grid
Each card in All Categories (N) shows:
- A status: You Lead, Low Gap, Medium Gap, High Gap or Critical Gap.
- N prompts tracked — how many prompts are in the category.
- Your Coverage — your visibility in this category.
- Leader: X (y%) — the leading brand and its coverage.
Understanding the data
The status shows how far you are from the category leader. You Lead means you are the leader. The gap labels go from Low Gap (close to the leader) up to Critical Gap (far behind).
A category is a group of related prompts, such as "pricing" or "reviews". A prompt is a question Prism asks the AI platforms.
Example
For example, the "integrations" card shows 12 prompts tracked, Your Coverage 10% and Leader: Competitor A (60%), with Critical Gap. The "pricing" card shows You Lead. You keep your pricing pages as they are and plan new integration guides first.
Tips
- Start with Critical Gap cards that have many prompts tracked.
- Open Competitors to learn more about the leader.
- Keep category names short and consistent in your prompts file.
Common issues
| Message | What to do |
|---|---|
| No prompt categories found for the selected filters. | Clear the category filter, or check that your prompts file has a Category column filled in. |
Frequently asked questions
Why don't I see the Best and Worst Performing Category cards?
They show only when there are at least 4 prompt categories for your filters.
Where do the categories come from?
From the Category column in your prompts file. You set it up in Project settings or Keyword Setting in the Client panel.
What does You Lead mean?
Your brand has the highest coverage in that category among the tracked brands.
Related articles
- AI VisibilityAI Visibility OverviewRead the AI Visibility Overview screen: KPI cards, visibility trends, the leaderboard, answer positions and top prompts.Read
- AI VisibilityCompetitorsSee how your brand compares with tracked competitors in AI answers: visibility, mentions, citations, best platform and rank.Read
- Rank TrackingKeywords and Prompt FilesUpload the keywords and AI prompts Prism tracks, using Excel files with the exact column headers Prism expects.Read
- AI VisibilityPlatformsCompare your AI visibility on ChatGPT, Perplexity, AI Mode, Gemini, Claude and AI Overview, and read each platform's status.Read
Was this article helpful?
Still stuck? Try troubleshooting or email support.