Predict Customer Churn with Claude AI Using Mixpanel, Stripe & HubSpot Data
Calculate real-time customer health scores with Claude AI by analyzing multi-signal data from Mixpanel usage, Stripe payments, Intercom sentiment, and HubSpot engagement.
The workflow at a glance
Use this template as a starting point, connect the apps it needs, then adapt the steps to your own workflow.
Everything this workflow handles
Transform your customer success operations with AI-powered health scoring that predicts churn before it happens. This puq.ai template aggregates signals from Mixpanel, Stripe, Intercom, and HubSpot, then uses Claude AI to calculate comprehensive health scores — automatically alerting your team to at-risk accounts with actionable intervention recommendations.
Designed for customer success teams managing SaaS retention and expansion, this workflow ensures no account silently churns while identifying your best candidates for upsell conversations.
How This Automation Works
1. Trigger on Customer Data Updates
The workflow activates when customer records are created or updated in your Airtable CRM, capturing the latest multi-signal data.
2. Multi-Signal Health Analysis with Claude AI
Claude AI calculates health scores using weighted signals:
- Product Usage (Mixpanel) — 35% weight: Login frequency, feature adoption, session depth
- Payment Health (Stripe) — 25% weight: Payment success, plan tier, billing issues
- Support Sentiment (Intercom) — 20% weight: Ticket volume, sentiment, resolution satisfaction
- Engagement (HubSpot) — 20% weight: Email opens, meeting attendance, content engagement
3. Health Band Classification
Customers are classified into actionable segments:
- Critical (0-30) — Immediate intervention required
- At Risk (31-50) — Proactive outreach needed
- Warning (51-70) — Monitor closely
- Healthy (71-85) — Standard care
- Champion (86-100) — Expansion opportunity
4. AI-Generated Insights
For each customer, Claude AI provides:
- Primary risk reason identification
- Specific intervention recommendations
- Owner segment assignment (CSM, Account Manager, Support)
- Churn probability percentage
- Expansion potential score
5. Airtable Health Snapshots
Complete health assessments are logged to Airtable including scores, bands, MRR, usage metrics, and AI analysis for trend tracking.
6. Health-Based Team Routing
Alerts are routed based on health band:
- Critical/At Risk → Immediate Slack alert to #cs-critical-alerts + urgent Gmail to assigned CSM
- Warning → Slack notification to #cs-monitoring
- Healthy/Champion → Slack post to #cs-wins with expansion opportunities
Key Benefits
- AI-powered multi-signal health scoring
- Predictive churn identification before cancellation
- Automated CSM alerts for at-risk accounts
- Expansion opportunity detection for healthy customers
- Complete health history for trend analysis
- Weighted scoring across usage, payment, support, engagement
- Actionable intervention recommendations
- Owner segment routing for appropriate follow-up
Use Cases
- SaaS companies tracking customer health metrics
- Customer success teams managing large account portfolios
- Revenue teams focused on net retention
- Organizations with multi-source customer data
- Teams needing predictive churn analytics
- Companies prioritizing expansion revenue
Integrations Used
- Airtable — Customer database and health snapshot logging
- Claude AI — Multi-signal analysis and health scoring
- Mixpanel — Product usage data (via Airtable sync)
- Stripe — Payment health metrics (via Airtable sync)
- Intercom — Support sentiment data (via Airtable sync)
- HubSpot — Engagement tracking (via Airtable sync)
- Slack — Team alerts by health band
- Gmail — CSM notifications for critical accounts
Why Use This puq.ai Template?
Reactive customer success waits for cancellation requests. This template enables proactive retention by surfacing at-risk accounts with enough time for meaningful intervention — while simultaneously identifying your best expansion opportunities.
With puq.ai, you deploy enterprise-grade customer health automation in minutes — no code, no surprise churn, no missed expansion.
Get Started in Minutes
- Import the template into puq.ai
- Connect your Airtable, Claude AI, Slack, and Gmail accounts
- Configure your Airtable base with multi-signal customer data
- Set up Slack channels for each alert tier
- Customize health band thresholds for your business
- Activate and start predicting churn automatically
Score health. Predict churn. Retain revenue.
Use this template to transform reactive customer success into proactive retention — powered by Claude AI multi-signal analysis and orchestrated by puq.ai.
Questions before you start
Do I need to write code to use this?
No. This is a pre-built workflow — you connect your accounts and adjust settings in puq.ai's visual editor. No custom code required.
How long does setup actually take?
Most people have this workflow running in about 12 min setup. You're connecting accounts and mapping a few fields for Airtable, Claude, Mixpanel, Stripe, Intercom, HubSpot, Slack, and Gmail — not building the automation from scratch.
What if I don't use one of these apps?
You can swap any step for another supported app inside the workflow builder after you open the template — the trigger-to-action structure stays the same.
What happens after I click "Use this template"?
You'll be taken to your puq.ai workspace with this workflow pre-loaded — connect Airtable, Claude, Mixpanel, Stripe, Intercom, HubSpot, Slack, and Gmail, adjust any settings, and turn it on.
Can I customize the workflow after I import it?
Yes — this template is a starting point, not a fixed script. Add, remove, or rewire steps freely once it's in your workspace.
Is my data safe when I connect an app?
You control which accounts you connect, and you can disconnect any integration at any time from your puq.ai dashboard.