Complete AI Training

Prompt

Design Customer Service QA Scorecard

Use this when you need a practical quality rubric that evaluates empathy, accuracy, resolution, and compliance.

How to use it

  1. Copy the prompt and paste it into ChatGPT, Claude, Gemini or any other AI.
  2. Replace every {{placeholder}} with your own details, or let the AI ask you for them.
  3. Use the follow-ups below to go deeper.
Prompt

Role You are a customer service quality lead who builds practical QA scorecards that coaches can apply consistently and agents can understand.

Context you provide

  • {{channel_type}} - phone, email, chat, or social.
  • {{interaction_type}} - the main reason customers contact you.
  • {{team_size}} - number of agents being scored.
  • {{existing_metrics}} - current KPIs like CSAT, FCR, AHT.
  • {{compliance_requirements}} - required disclosures, privacy rules, or scripts.
  • {{company_tone}} - brand voice, e.g., warm, concise, formal.
  • {{score_scale}} - points or levels per criterion.
  • {{weighting_priorities}} - which dimensions matter most.
  • {{review_frequency}} - how often QA reviews happen.
  • {{coaching_process}} - how feedback reaches agents.

Instructions

  1. Ask for any missing inputs, then confirm the scorecard's purpose and scope.
  2. Define 4 to 6 criteria covering empathy, accuracy, resolution, and compliance. For each, write a one-sentence description and 3 to 5 observable behaviors.
  3. Assign point values or levels using {{score_scale}} and weights based on {{weighting_priorities}}.
  4. Add scoring guidance for partial credit and automatic failures, such as a missed compliance disclosure.
  5. Include a short calibration note for reviewers and a feedback template linked to {{coaching_process}}.
  6. Format as a table plus a one-page summary.

Output format A markdown table with columns: Criterion, Description, Observable Behaviors, Points, Weight. Below the table, add a 5-bullet calibration guide and a 3-sentence feedback template. Keep total length under 700 words. Use plain language, no jargon.

Guardrails

  • Do not invent legal or regulatory requirements. Ask the user to confirm with their compliance team.
  • Flag any assumption about scoring weights or automatic failures.
  • Remind the user that local privacy laws and company policies must be checked before rollout.

Example Channel: phone; interaction: billing dispute; team size: 12; existing metrics: CSAT, FCR; compliance: verify identity before account changes; tone: friendly and clear; scale: 1-5; priorities: resolution and compliance; review: weekly; coaching: 1:1 within 48 hours.