Scorecard Actions are powerful, automated rules that you can set up to run immediately after your AI completes an evaluation.
Think of them as an "if this, then that" system for your quality assurance process. By setting up actions, you can eliminate manual follow-up tasks, ensure consistency, and instantly notify the right people about critical interactions, saving you valuable time.
Actions are triggered based on the results of a scorecard. You can create actions based on the overall score or on the answer to a specific question, giving you precise control over your automated workflows.
How to Create a Scorecard Action
You can add an action at two different levels within any scorecard:
Scorecard-level: The action is triggered based on the overall result of the entire scorecard (e.g., total points, final percentage, or pass/fail status).
Question-level: The action is triggered based on the specific answer your AI gives for that individual question.
To get started:
Navigate to Scorecards from the main menu and open the scorecard you wish to edit, or create a new one.
To create a Scorecard-level action, click the "Add Action" button in the main scorecard header.
To create a Question-level action, find the specific question and click its "Add Action" button.
After clicking, a dialog box will appear. First, define the trigger condition. This tells the system when to perform the action.
Next, choose the action you want to automate. Actions are grouped into several categories to help you find what you need.
Available Actions
Here are the actions you can automate, broken down by category:
Evaluations
Queue a new AI evaluation: Automatically trigger a different AI scorecard. Useful for a multi-stage review process.
New human evaluation: Assign the interaction to a human for manual review with a new scorecard. Perfect for in a hybrid QA workflow as a filter that assign high priority interactions to a human team.
Require human review: Assign the evaluation for a manager or admin to review and approve.
Organization
Add interaction to a List: Automatically group similar calls together. For example, add all perfect scores to a "Training Examples" list.
Tag the interaction: Apply a specific tag to the interaction for easier filtering and reporting later.
Third Party
Send a Slack message: Instantly notify a specific channel in Slack with a customizable message.
Post to a webhook: Send the evaluation, transcript, and audio data to another application you use, like a CRM, Zapier, or a project management tool.
Clean up
Exclude evaluation's Score: Keep the evaluation data but exclude its score from all reporting and agent dashboards.
Delete call recording: Permanently delete the audio file.
Delete evaluation: Permanently delete the entire evaluation submission.
Delete transcript: Permanently delete the call's transcript.
Use Cases & Creative Ideas
Not sure where to start? Here are a few practical examples of how you can use Scorecard Actions to build a more efficient QA process.
Use Case: Flag Critical Failures for Immediate Attention
Trigger: On a question like "Did the agent provide the required compliance disclosure?", set the trigger to fire if the answer is "No".
Action 1: Send a Slack message to the #managers channel with a link to the call.
Action 2: Add interaction to a List called "Urgent Compliance Review".
Use Case: Escalate Low Scores for Review
Trigger: At the Scorecard-level, set the trigger to fire if the final Score % is less than 75%.
Action: Require human review and assign it to an evaluator to manually review the call.
Use Case: Identify Top Performers and Best Practices
Trigger: At the Scorecard-level, set the trigger to fire if the final Score % is 100%.
Action 1: Tag the interaction with "Top_Performance".
Action 2: Add interaction to a List called "Gold Star Calls" to use in future team training.
Use Case: Automate Follow-up with Other Systems
Trigger: On a question "Did the customer express interest in Product X?", set the trigger to fire if the answer is "Yes".
Action: Post to a webhook that connects to your CRM, automatically creating a follow-up task for your sales team.
Use Case: Auto Export Evaluation Data to BI tools: After each submission set a trigger to send the completed evaluation data to your BI tool, eg. Domo, Looker, PowerBI, etc.
Action: Post to a webhook that connects to directly to your BI tool, an internal service, or via a platform like Zapier.