Predict customer churn risk & create interventions with GPT-4, Zendesk & HubSpot
Needs changesNeeds changes. This template uses 1 thing(s) that n8n 3.0 removes: Cron. 1 breaking item(s) found. Static check, last run 2026-10-19.
What to replace before upgrading
| Removed in 3.0 | Replace with |
|---|---|
| Cron Daily Risk Analysis | Schedule Trigger |
Worth a look
- AI Transform (AI Risk Prediction, Generate Critical Intervention, Generate High Risk Intervention, Generate Medium Risk Check-in): Automatically migrated to a Code node on upgrade, keeping the generated JavaScript. You can no longer add new AI Transform nodes.
Nodes used (10)
- Cron
n8n-nodes-base.cron - HTTP Request
n8n-nodes-base.httpRequest - Zendesk
n8n-nodes-base.zendesk - Stripe
n8n-nodes-base.stripe - Merge
n8n-nodes-base.merge - AI Transform
n8n-nodes-base.aiTransform - If
n8n-nodes-base.if - Slack
n8n-nodes-base.slack - Send Email
n8n-nodes-base.emailSend - HubSpot
n8n-nodes-base.hubspot
What it does
How it works This AI Customer Success Risk Prediction workflow revolutionizes customer retention by predicting churn risk 30-90 days before it happens. Here's the high-level flow: Daily Data Collection → AI Multi-Signal Analysis → Risk Scoring & Prediction → Smart Risk Routing → AI-Generated Personalized Interventions → CRM Updates & Team Alerts The system automatically gathers data from your product analytics, support system, billing platform, and email tools, then uses GPT-4 to analyze patterns and predict which customers are at risk. It creates personalized intervention strategies and routes them based on urgency level. Set up steps Time to set up: Approximately 45 minutes Prerequisites: Active accounts with your analytics platform, support system, billing provider, CRM, and AI provider Step 1: Import & Configure Workflow (5 minutes) Import the workflow JSON into your n8n instance Review the 3 comprehensive sticky notes for context Understand the AI analysis logic and intervention strategies Step 2: Set Environment Variables (10 minutes) Configure these critical variables: ANALYTICSAPIURL and ANALYTICSAPIKEY HIGHRISKSLACKCHANNEL (for critical alerts) CSTEAMEMAIL (intervention sender) CRMBASEURL and CALENDARBOOKINGURL Step 3: Configure API Credentials (20 minutes) Set up secure credential connections for: OpenAI/Anthropic API (AI analysis engine) Analytics platform (Mixpanel/Amplitude/GA) Support system (Zendesk/Intercom) Billing platform (Stripe/Chargebee) HubSpot CRM (risk data storage) Slack API (team notifications) SMTP/SendGrid (email delivery) Step 4: Customize AI Prompts & Risk Thresholds (8 minutes) Review and adjust the AI analysis prompts for your business Modify risk score thresholds (Critical 90+, High 70-89, Medium 40-69) Customize intervention email templates and tone Set your specific risk factors (usage patterns, support indicators) Step 5: Test & Activate (2 minutes) Run a test execution with sample customer data Verify AI analysis generates appropriate risk scores Check that interventions are routed correctly Activate the daily cron schedule