Customer churn costs telecom companies millions annually. This project simulates a TELUS-style analytics workflow to identify which customers are most likely to churn, why they churn, and what revenue is at risk.
Raw Data → Staging Layer → Core Layer → Analytics Mart → Power BI
A 3-layer SQL data architecture was designed to:
- Ingest and validate raw customer data (Staging)
- Clean, standardize, and join datasets (Core)
- Build churn-ready analytical tables (Analytics Mart)
- Customer churn rate and KPI metrics
- Churn drivers: contract type, internet service, tenure
- Customer risk segmentation: Low / Medium / High
- Revenue exposure quantification by segment
- Month-to-month contracts show significantly higher churn rates than annual contracts
- Customers with tenure under 12 months represent the highest churn risk segment
- Specific service combinations correlate with improved retention
- Revenue exposure can be estimated and prioritized by risk tier for proactive intervention
| Tool | Purpose |
|---|---|
| SQL Server | Data Architecture & Transformation |
| Power BI | Executive Dashboard |
| DAX | KPI Measures |
| Star Schema | Dimensional Data Modeling |
| Power Query | Data Cleaning |
- Full SQL scripts (staging → core → analytics mart)
- Power BI dashboard (.pbix)
- Technical workflow documentation
- Project insights PDF
SQL Data Modeling · Churn Analysis · Customer Segmentation Revenue Impact Analysis · Business Intelligence · DAX