Evaluation System

The core of ClientCoach AI is its robust evaluation engine, designed to provide objective, structured, and actionable feedback.

✅ Scoring System

Condition Type
Score Impact

Critical Met

+20 points

Optional Met

+10 points

Red Flag Triggered

–40 points

Final Score Range

0–100 (normalized)

🧩 Evaluation Criteria

  1. Critical Aspects

    • Must be fulfilled to pass a phase

    • Examples: clarify goals, ask for use case, confirm need

  2. Optional Aspects

    • Good to include; boost score

    • Examples: empathize, mention integrations, provide examples

  3. Red Flags

    • Must avoid; damage the score

    • Examples: interrupting, overselling, ignoring objections

🔄 Phase-Based Evaluation

Each conversation is broken into structured phases. The Observer Agent tracks whether you meet the necessary conditions for success in each phase.

  • On Success → move to next intended phase

  • On Failure → the customer may show resistance or move toward closure

📈 Feedback Report

After each session, you get:

  • Phase-by-phase analysis

  • What was done right and wrong

  • Missed opportunities

  • Actionable recommendations for improvement

  • A score trendline over time

This transparent, educational feedback loop helps agents improve rapidly and consistently.

Last updated