Build AI-Informed Workflow Automation in HubSpot

Requirements

Marketing Hub Professional+ or Sales Hub Professional+
Breeze AI enabled
Workflow access
Ability to create custom contact properties

What This Enables

Automation based on likelihood, intent, or risk
Predictive behavior without relying on static rules alone
Sales and marketing actions triggered from AI-informed signals

Examples:

“This lead is unlikely to engage”
“This contact shows buying intent”
“This deal is at risk of stalling”

Step 1: Select a Breeze AI Insight (Source Signal)

Choose an AI insight that is visible in the UI, even if it’s not workflow-triggerable:

Common sources:

Engagement likelihood
Predictive lead insights
Sales activity trends
Email sentiment summaries
Deal risk indicators

These signals cannot directly trigger workflows, they must be persisted first.

Step 2: Create a Writable “AI Signal” Property

Create a custom contact property:

Object: Contact
Name: AI Engagement Tier
Type: Dropdown
Values:
- High
- Medium
- Low

(Optional: repeat this pattern for Intent, Risk, or Priority)

Step 3: Translate AI Insight → Property (The Hack)

Option A: Workflow-Based Translation (No Code)

Create a contact-based workflow using proxy conditions that reflect the AI insight.

Example: Low Engagement Detection

Enrollment triggers:

HubSpot Score < 30
Last activity date > 7 days ago
Marketing emails opened = 0 (last 14 days)

Action:

Set AI Engagement Tier = Low

Repeat with additional workflows for:

Medium engagement
High engagement

This creates a persistent AI-informed signal workflows can rely on.

Option B: Sales-Assisted AI Capture (Optional)

For Sales Hub teams:

SDR reviews Breeze insights
Updates AI Engagement Tier via record sidebar
Workflow handles the response automatically

This keeps a human-in-the-loop for high-value records.

Step 4: Build the Automation Workflow

Example: Low Engagement Rescue Workflow

Workflow type: Contact-based

Enrollment triggers:

AI Engagement Tier = Low
Lifecycle stage = Lead

Actions:

Enroll in re-engagement nurture
Create follow-up task for SDR
Send internal Slack/email notification
Reduce lead score (optional)

This workflow is now:

Stable
Auditable
Fully supported

Step 5: Add AI-Driven Branching Logic

Use the same pattern for intent or priority.

If/then branch:

If AI Intent Tier = High
- Assign senior rep
- Notify sales manager
Else
- Continue automated nurture

All branching is now property-based and reliable.

Step 6: Monitor and Tune

Track:

Workflow enrollment volume
Conversion rate (before vs after)
Time to first response
Sales acceptance rate

Adjust:

Thresholds monthly
Proxy logic quarterly
Property values as needed

Why This Works

Breeze provides directional intelligence
Properties provide automation stability
Workflows remain deterministic
AI improves decisions without breaking governance

Common Pitfalls (Avoid These)

Triggering workflows directly from Breeze UI insights
Treating AI scores as the absolute truth
No human escalation for high-value leads
Over-triggering without volume controls