Subscription Intelligence – The "Churn Killer" Engine

Transforming a multi-brand Healthtech platform from "Generic Marketing" to "Predictive Retention" using medical-grade data infrastructure.

Healthtech Infrastructure
Churn Prediction
GDPR Compliance
+15%
Uplift in 6-Month Customer LTV
100%
Automated GDPR Compliance
Work Image
ABOUT THE PROJECT

Overview

Wellster Healthtech operates multiple direct-to-consumer medical brands. The challenge wasn't just volume; it was variety. Dealing with sensitive patient data alongside high-velocity e-commerce transactions created a massive data silo problem. We architected a centralized "Data Lakehouse" that allowed for unified reporting across all brands while maintaining strict medical compliance.

The Challenge

The company was acquiring customers rapidly, but churn was high. Because marketing data (Facebook Ads) was disconnected from medical data (Patient Consultations), the CRM teams were sending generic emails. They couldn't distinguish between a patient who stopped buying because they were "cured" vs. a patient who was "unhappy." This lack of context was costing millions in potential recurring revenue.

The Solution

We moved beyond basic analytics and built a Customer Health Engine:

  • Unified Architecture (BigQuery): I designed a schema that merged Shopify subscription data with anonymized medical records, creating a "Single Patient View."

  • Predictive Modeling (dbt + Python): We built a "Churn Probability Score." If a patient's refill was 3 days late, the system automatically flagged them as "At Risk."

  • Automated Activation: Instead of a generic newsletter, the data triggered a specific "Check-in" email from a doctor or support agent.

  • Privacy First: Built automated "Data Masking" protocols to ensure marketing teams could analyze trends without ever seeing sensitive medical history (GDPR/HIPAA compliant).

Logo