Skip to main content
The Prompts tab is the raw evidence behind your GEO score. For each prompt you see what every AI platform actually said, and where you landed.

​
The four outcomes

Each prompt/platform pair resolves to one of four states. This is the vocabulary the whole report uses.
Named, not cited is the most fixable state. The model already knows who you are — it just didn’t have a page of yours worth quoting. That’s a content problem with a clear answer.

​
Filtering

Narrow the list by:
  • Platform — ChatGPT, Gemini, Claude, Perplexity, Copilot, Grok, DeepSeek, Meta AI, Mistral, Poe, AI Overview.
  • Outcome — mentioned, not mentioned, cites you, cites a competitor, cites both, neither.
  • Intent type — Informational, Commercial, Comparison, Buyer intent, Branded, Locational, Adjacent, Direct, General.

​
Intent types

Every prompt is classified by what the person asking actually wants. Coverage across intent types matters as much as raw volume: winning every informational prompt while losing every comparison prompt is a losing position.

​
Reading an answer

Open any prompt to see the full response from each platform, the sources it cited, which competitors it named, and the sentiment of your mention. Draft reply turns a weak answer into a content brief.

​
Managing tracked prompts

Prompt Explorer, reachable from the toolbar, holds the tracked prompt set:
  • Tracked prompts are generated and refilled by Botric to your plan’s cap on every run.
  • Add-on prompts are yours — fixed slots per site, available on paid plans. Locked slots show as Locked.
A run bills the full tracked cap plus your add-on prompts, regardless of how many prompts are currently visible. See How credits work.
Saved prompts don’t appear in results until the next run. When that happens, the overview shows a stale prompts banner.

​
Next steps

Prompt Research

Trial a prompt before you track it.

Content gaps

Turn losing prompts into a content plan.