Support Quality Review Form and Calibration Session
Score conversations consistently across reviewers.
The Open buttons send the prompt text as currently shown below to that site only when you click. Fill in the placeholders first.
Create a quality review form. What good support means here: [QUALITY_DEFINITION]. Channels: [CHANNELS]. Reviewers: [REVIEWERS]. Number of conversations reviewed per agent per month: [REVIEW_VOLUME]. Design categories: accuracy, resolution, tone, clarity, process and compliance, each scored 0 to 2 with descriptions. Add automatic fail rules for serious errors such as sharing another customer's data. Include a comments field requiring a quote. Add a calibration session plan: all reviewers score the same five conversations, compare, and adjust definitions until scores differ by no more than one point. Return the form, the plan, a sampling method that avoids only reviewing complaints, and a coaching note template that begins with what went well. Add a worked scoring example with an invented conversation: the agent answered the wrong question but was polite, so accuracy scores 0 and tone scores 2, giving a total below the pass mark, with a comment quoting the line where the misunderstanding occurred. Provide the pass mark logic: total points divided by maximum points as a percentage, with the arithmetic for a score of 9 out of 12 giving 75 percent. Describe how many conversations a reviewer can score per hour and how to plan time. Include a rule that scores are used for coaching and trends, not for ranking people publicly. Add guidance on reviewing conversations in other languages or about sensitive topics, privacy rules for reviewers, and how agents can challenge a score. Provide a quarterly summary layout with the top three strengths and the top three themes for improvement, and a plan for changing the form when the product or policy changes. Include a note on recording the calibration outcomes so definitions improve over time, and on sampling conversations fairly across shifts, channels and ticket types. Add a short rule for what happens when a reviewer finds a serious compliance breach: stop scoring, tell the lead the same day, and keep the record restricted to those who need it.
Fill in before use
Optional: type your details and the prompt above updates. Nothing is saved or sent.
Replace each bracketed placeholder with your own details, then paste into Any AI or another assistant.
See an example result (fictional data)
Invented review of one conversation, scores 0 to 2 across six categories. Accuracy 0 (answered the wrong question; quote: "Your plan renews on the 5th" when the customer asked about cancelling). Resolution 1. Tone 2. Clarity 2. Process 2. Compliance 2. Total 9 of 12 = 75 percent. Pass mark 80 percent, so below target. Comment begins with strengths: warm tone and clear steps. Coaching point: restate the question before answering. Calibration: three reviewers scored 9, 10 and 8, within the allowed spread of 2 points.
Illustrative only, made-up data. Your results will differ and must be reviewed.