Overview
AI Scorecards help streamline your call evaluations by automatically providing initial answers, allowing you to focus on responses that need closer attention. By identifying areas where agents can improve, AI Scorecards save you time and make your coaching process more effective.
Creating AI Scorecard Templates
To get started, navigate to Scorecard Templates and select Create New Scorecard. Under Scoring Type, choose Score with REGAL AI.
Configuring Autoscore Criteria for AI Scorecard Templates
AI Scorecard Templates can score calls automatically once they’ve concluded. To set up automatic scoring, add Autoscore Criteria such as Task Type, Agent Team, and Campaign to specify which calls the AI Scorecard should evaluate.
Note: Each call will only be scored by one AI Scorecard Template. If multiple templates apply, the one with the most specific match will be used. For instance, a template matching both Task Type and Campaign will take precedence over one that matches only Task Type. If there’s still a tie, the most recently updated template will be applied.
Creating Questions for Your AI Scorecard Templates
Once you've set the Scoring Type and Autoscore Criteria, it’s time to add questions. Here are some tips for crafting questions that REGAL AI can best respond to:
-
- Use Clear Questions: Begin with phrases like “Did the agent…” or “How well did the agent…” to ensure accurate AI responses.
- Provide Context: Since REGAL AI doesn’t access your scoring rubrics, add relevant criteria directly in each question.
- Anticipate "N/A" Responses: If you expect the AI to return an "N/A" for some calls, specify the conditions for when that answer would apply.
Example Questions
Here are a few example questions that would perform well in an AI Scorecard. Note how they're all phrased as questions:
- Did the agent introduce themselves?
- Did the agent say the name of the company they're calling on behalf of?
- Did the agent indicate that the call will be recorded or that they were calling on a recorded line?
- How well did the agent identify the customer's concerns or objections?
- Did the agent clearly state the purpose of their call?
Using AI Scorecard Templates Manually
Once you’ve saved your AI Scorecard, it’s ready for use. When scoring a call, simply select the scorecard from the dropdown and click Score to see REGAL AI’s responses. After reviewing and confirming accuracy, click Save to finalize the scorecard.
Reviewing and Validating AI Scorecards
You can update AI Scorecards if, after review, you disagree with any of the answers or the AI’s reasoning. Once you save edits to an AI Scorecard, it’s marked as validated, indicating it has been reviewed and confirmed as accurate. You can also filter Insights Reports by validation status, allowing you to show agents only the Scorecards you’ve manually validated if desired.
If you review a Scorecard and agree with all responses—perfect! Simply click Confirm (next to Save) to validate the Scorecard without making any changes.