Skip to main content

Conditions

Learn best practices and how to use Conditions in workflows.

Swyft AI avatar
Written by Swyft AI
Updated this week

Overview

The Condition action lets you control whether later steps in your workflow should run based on rules you define. Rules can reference CRM fields, trigger data, AI-generated insights produced by earlier actions, and more. This gives you precise control over workflow execution, ensuring actions only occur when the right criteria are met.

Conditions are often used to:

  • Gate automation until AI has generated a required insight.

  • Ensure a trigger event includes all necessary context.

  • Apply logic to complex qualification frameworks (e.g., MEDDIC, BANT, etc.).


Components

Rule Logic (AND/OR)

  • AND – All rules in the group must be true for the condition to pass.

  • OR – Any single rule in the group being true will allow the condition to pass.

  • Nested Groups – Use Add Rule Group to combine AND/OR blocks together for advanced logic (e.g., (A AND B) OR C).


Add Rule

Click Add Rule to define a new condition. Each rule has two parts:

  1. Variable Selection – Choose the data point you want to evaluate.

    • CRM fields (e.g., Opportunity Stage, Renewal Date).

    • Trigger data (e.g., “Call Completed,” “Meeting Outcome”).

    • AI-generated insights (e.g., MEDDIC fields like Metrics, Pain, Decision Process).

    • Any variable produced by earlier workflow actions.

    Example: After running “Generate MEDDIC Insights,” you’ll see available fields like Metrics, Identify Pain, and Decision Criteria.

  2. Operator – Define how to evaluate the value. Common operators include:

    • Exists / Does Not Exist – Checks if a value has been generated or is empty.

    • Equals / Does Not Equal – Compares against a specific value.

    • Contains / Does Not Contain – Checks if a field includes a keyword or phrase.

    • Greater Than / Less Than – Compares numeric values (e.g., ARR > 10,000).


Rule Groups

  • Add Rule Group lets you nest multiple rules under different logic.

  • Useful for modeling real-world qualification logic.

  • Example:

    • Group 1: MEDDIC Metrics Exists AND Identify Pain Exists

    • Group 2: Stage = “Closed Won”

    • Overall Logic: (Group 1 OR Group 2)


Best Practices

  1. Reference earlier actions.

    • Rules can directly check AI-generated variables from previous workflow steps.

    • Example: Only update CRM if Decision Process was successfully generated in the “Generate MEDDIC Insights” action.

  2. Fail early.

    • Place condition checks before expensive actions (like CRM updates or Slack messages) to prevent bad data, noise, and unnecessary credit usage.

  3. Use Exists/Does Not Exist.

    • Helpful when gating actions on whether Swyft generated an insight (e.g., only update CRM if a “Decision Criteria” was captured).

  4. Keep logic transparent.

    • Name each Condition action clearly (e.g., “Check MEDDIC Complete” instead of just “Condition”) so workflow logs are easy to interpret.


Example

Use Case: Only update the CRM with MEDDIC fields if Swyft AI has generated all required insights.

  1. Add a Condition action.

  2. Create a rule group:

    • Metrics Exists

    • Identify Pain Exists

    • Decision Process Exists

    • Logic = AND

  3. If the condition passes → proceed to Update CRM Record action.

  4. If the condition fails → workflow ends, ensuring no partial data is written.


Summary

The Condition action is your workflow’s decision gate. By combining CRM fields, trigger data, and AI-generated insights, you can model both simple and highly complex logic to control exactly when later actions run.

Did this answer your question?