AI Workflows Automation #4973: NPS Collection with Intercom + HubSpot + Make

Problem

Model output from Intercom is only useful when nps collection results land reliably in HubSpot with trace logs.

The flow below documents that production path.

Workflow

Intercom trigger → transform/map fields → HubSpot action → optional alert via Make.

Tools Used

• Intercom
• HubSpot
• Make

Setup Steps

1. Connect Intercom and HubSpot with scoped API permissions.
2. Configure the nps collection entry condition (Medium difficulty in this library entry).
3. Set field transforms and default values between tools.
4. Add a dead-letter or retry path for failed runs.
5. Validate with sample data before go-live.

Expected Outcome

• nps collection runs without manual copy-paste between Intercom, HubSpot, Make.
• Status updates stay aligned across the connected tools.
• Failures surface in one place instead of silent drift.

Benefits & ROI

• Ranked as High ROI in our template dataset for AI Workflows.
• Typical implementation complexity: Medium.
• Frees ops time from repetitive nps collection tasks in this stack.

Variations

• Add a manual approval step before writes to HubSpot.
• Insert a deduplication check on Intercom record IDs.

Troubleshooting

• Pause the workflow before rotating API keys, then resume after credentials update.
• Log model inputs/outputs for traceability; never send secrets to LLM nodes.
• Add human review for low-confidence classifications.