Creating a new Scorecard Form

4 min

We can create customized scorecard forms tailored to specific quality parameters and call evaluation needs. With this feature, you can:

✅ Define custom sections and criteria for each evaluation form

🎯 Align scorecards with campaign, team, or process-level KPIs

📊 Set scoring types — numerical, weighted, or checkbox-based

🔄 Reuse and modify templates for consistent quality measurement

👨‍💼 Assign forms to evaluators for agent call reviews in real time

This powerful form builder helps streamline your quality assurance process while maintaining flexibility and accuracy across teams.

To create a scorecard form -

The view will display all the previously created scoreacrd forms.

The details shows us which channel/interaction is the form created for. Status , and last changes done. You can edit , delete any scorecard form.

To create a new scorecard form for any channel - Click on the "Create New Scorecard form" to create new. Fill in Form Details

a. Form Name

Enter a unique and descriptive name for your scorecard. This helps evaluators identify the purpose of the form. b. Interaction Type

Select the type of interaction this scorecard applies to.

Common options include:

Audio Call

Video Call

Chat

c. LOB (Line of Business)

Choose the business unit or domain where this scorecard will be used.

d. Threshold

Enter the minimum passing score or threshold percentage for the evaluation.

e. Form Status

The form status appears automatically as: Draft

f. Add Parameters

Click the + Add Parameter button to include scoring parameters or evaluation criteria.

Each parameter may include:

Parameter Header - The name that the parameter holds for. Eg Greetings etc.

Definition - This field is used to describe what this parameter measures

Weight (0–1) - Weightage of this parameter on the overall form.

Reasoning Prompt - This field allows you to write a detailed instruction or question that guides automated scoring or evaluator reasoning.

Build Prompt Button - This option helps users automatically generate a structured prompt using pre-built templates or assistive suggestions.

The form will pop up . Fill in

Goal - One-line instruction that captures the essence of the parameter you want evaluated.

Output - The expected format or type of the response. Use this to control how the reasoning result is returned (for example, a simple Yes/No, a numeric score,)

Constraints - Optional free-text field for rules that limit or shape the generated prompt. Constraints let you force specific behaviours, exclude certain words, or require a particular style.

- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -

Example A — Binary check (Yes/No) with explanation

Goal: Did the agent greet the customer and use the customer’s name?

Output: Yes/No (Explanation allowed)

No explanation: OFF

Constraints: If greeting is partial, return "No". Limit explanation to 2 sentences.

Generated behaviour: returns Yes plus a 1–2 sentence reason, e.g. Yes — Agent greeted the customer and used the name "Ravi" within the first 10 seconds.

- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -

Critical Error Toggle - Enable this option if failing this parameter should automatically fail the entire evaluation.

Coach Tag Toggle - Enable this option when the parameter should be highlighted for coaching purposes.

Rules Section - Rules allow you to define logic-driven scoring criteria for the parameter.

Load Defaults – On top of scorecard form , you have an option of automatically populating the scorecard form with predefined standard parameters. These settings can also be preconfigured and reused across multiple forms for consistency and efficiency.

In case you need to configure a new metrics after the defaults, click on Add Parameter. and fill in the details.

Fill in all the details and click on "Save"

Error Message

Weights do not sum to 1

Your parameter weights currently total: 1.20. Adjust weights so they add up to 1.

This means the combined weights of all parameters in your scorecard do not equal 1.0, which is the required standard for normalization. Review the form again to ensure the weights match to 1.