The "Scoring Card" allows you to automate the evaluation of your calls using AI by defining precise quality and performance criteria. Below are the steps to configure it.

To get a live guide, click on the icon

Creating the Evaluation Scorecard

1. Accessing the scorecard creation

To get started, go to Empower Settings, then click on AI Customization.
In the left-hand menu, select Call Score and click on Create a scorecard.

2. Choosing your creation method

Three options are available to generate your scorecard:

1. Create a form with the AI agent

This is the guided method. Simply answer the chat questions (main objective, call type, key criteria such as empathy, existence of a script, etc.). The AI will automatically generate a tailored scorecard.

2. Create from a template

Use an existing scorecard where all fields are already filled in. You just need to customize it to fit your needs.

3. Create with your own instructions

This is the most flexible option. You name your form, choose the use case, and write the instructions. It is recommended to include the role the AI should take, the scoring methodology, and the key steps of the call.

In the instructions, it is recommended to include the following information:

The scorecard title: what will be evaluated
The role: the posture the AI should adopt to evaluate the call
The methodology: how the call will be scored
(for example: active listening, score from 1 to 5 for each step)

To build your scorecard, you should consider the main objective of call evaluations, the types of calls you want to assess, the most important criteria, behaviors to avoid, etc.

💡 We recommend following our evaluation scorecard template to create the most effective scoring card possible.

💡 If you would like help from your CSM to build your scoring card, please fill out this questionnaire and send it to us.

3. Customizing evaluation criteria

Once the basic structure is created, you can refine the technical criteria. You can enable, disable, or weight the importance of each element:

Objection handling: Measures the agent’s ability to respond to issues.
Product knowledge: Evaluates understanding of the service.
Call sentiment: Analyzes the contact’s words to determine sentiment.
Moment tracking: Uses predefined keywords to analyze specific moments.
Negative words: Checks whether the agent uses negative words.
Duration: Analyzes call length to estimate ideal engagement time.
Speech rate: Calculates speaking speed to identify optimal pacing.
Talk-to-listen ratio: Ensures balanced communication.
Call closing: Measures the effectiveness of the conclusion.
Interactions: Measures interruptions, monologues, filler words, etc.
Script compliance: Calculates compliance with a reference script.
Silences: Detects awkward silences.

4. Assigning and activating the scorecard

Before finishing:

Add the users or groups concerned by this scorecard. You can enable automatic assignment for new users.
Don’t forget to Save and Activate your scoring card so it applies to future calls.

🎉 That’s it! You now know how to configure your evaluation scorecard.