Mention Rate Tool

← Return to Entity Measurement Framework

Mention Rate Tool Beta v2.0

An LLM output is not a rank; it is an observation from a probability distribution. Use this tool to calculate your baseline Mention Rate across structured prompt portfolios as outlined in the Entity Measurement Framework.

Operational Rules for This Tool

  • Presence First: This utility calculates Mention Rate ($Mentions \div Total Runs$), measuring pure brand presence in the response pool. Recommendation stance and citation share must be evaluated in companion analysis.
  • Breadth Over Repetition: Ensure your tested prompt portfolio represents multiple intents (Discover, Research, Compare, Validate) across your 5D taxonomy before burning compute on high repetition numbers.
  • Separation of Platforms: Run tests independently across engines (ChatGPT, Perplexity, Gemini, Copilot). Never pool multi-engine responses into a single blended score.

// Mention Rate Sampling Engine Loaded Below

Interpreting Your Output

A stable Mention Rate confirms whether your entity exists in the model’s extraction pool for a given prompt portfolio. If unbranded presence is low, review your Topic Entity Framework to address identity disambiguation and third-party consensus gaps.