📘

Call scoring helps teams improve call quality without adding manual work. After each conversation, calls are automatically evaluated using a scoring template called a scorecard. This guide explains how Scorecards works, how scores are calculated, and how supervisors and agents can use the results for faster, more consistent coaching.

Overview

Call scoring automatically evaluates each eligible call and provides consistent feedback. The system uses a company-defined scorecards to score key parts of a conversation (for example: Agent Call Closing, Call Opening & Introduction, etc.). This helps supervisors focus on calls that need a human review, while agents get quick coaching moments after every call.

Where to find the scorecards in Ziwo?

How scoring works

Scorecard-driven

Each score is generated from a template that defines the categories to evaluate (for example, Agent Call Closing or Call Opening & Introduction). Within each category, one or more metrics are defined. Each metric is evaluated as True or False: if the call meets the metric definition, it is validated; otherwise, it is rejected.

Score calculation

For a given call:

• The system looks at all the metrics that are enabled in the scorecard.
• Each time a metric is detected and validated during the call, it is counted as scored.

The final score represents how many of the expected metrics were successfully met during the call.

If the final score meets or exceeds the minimum score defined in the scorecard, the call passes the evaluation. Otherwise, it does not.

Example evaluation

Assume a scorecard with the following setup:

• Enabled metrics in the scorecard: 10
• Metrics validated during the call: 7
• Pass threshold: 70%

Score calculation:

• 7 validated metrics ÷ 10 enabled metrics = 70%

Result:

• Final score: 70%
• Threshold: 70%
• Outcome: ✅ Pass

If only 6 metrics had been validated, the score would be 60%, and the call would fail the evaluation.

Set up the scorecard

TBD

Best practices for building scorecard

Make metrics observable: Avoid vague terms like "Strong communicator". Instead, write a clear definition of the metric and add one or more concrete examples.

• Metric definition: "Did the agent start the call with a proper greeting like "Hello, thank you for calling Ziwo support, how can I help you today?"
• Examples:
  • "Hello, this is John from Ziwo support, how may I assist you?",
  • "Hi there, thanks for calling our customer service line."

Avoid context-dependent metrics: Only include the questions that can be evaluated from the call transcript.

• Metric definition: "Makes resolution steps explicit to the customer."
• Examples:
  • "Before ending the call, the agent summarizes the issue, states the actions taken, and confirms the outcome."

Limit the scope: Prefer a small number of well-defined categories over many broad ones. Each category should include focused, clearly defined metrics with concrete examples to ensure consistent evaluation. Keep the scorecard simple and easy to manage.

Use multiple examples where possible: Adding more than one example per metric helps make expectations clearer and avoids ambiguity.

FAQ