Route real estate buyer inquiries with GPT-4o, HubSpot, Twilio and Slack

Quick overview

This workflow runs daily to pull 30-day account usage from your product API, compute a health score, and use OpenAI to classify churn risk and recommend a retention play, then routes actions to Slack, Gmail, HubSpot, and Google Sheets.

How it works

1. Runs every morning at 7am on a schedule.
2. Requests the last 30 days of account usage metrics from your product API.
3. Calculates a 0–100 health score per account from usage and support signals.
4. Sends the account context to OpenAI (GPT-4o) to classify churn risk (high/medium/low) and recommend one retention play.
5. Routes high-risk accounts to a Slack alert for the #customer-success channel, medium-risk accounts to a nurture email via Gmail, and low-risk accounts to a CRM update in HubSpot.
6. Appends the account’s signals, score, and chosen action to a Google Sheets “Churn Signals” tab for tracking and reporting.

Setup

1. Provide a reachable product usage endpoint and update the HTTP request URL (and any required auth headers) to match your product API.
2. Add an OpenAI credential and ensure the selected GPT-4o model is available to your account.
3. Connect Slack credentials and confirm the target channel (for example, #customer-success) exists and is selectable.
4. Connect a Gmail account authorized to send emails and verify the account owner email field returned by your API is correct.
5. Connect HubSpot credentials and configure the contact lookup/update fields you want to write (for example, a custom health score property).
6. Connect Google Sheets credentials and replace REPLACE_WITH_SHEET_ID with your spreadsheet ID (and ensure a “Churn Signals” sheet/tab exists).