Telecom Customer Churn Analysis | SQL + Power BI
This project analyzes customer churn behavior for a telecom company using a multi-layer SQL analytics pipeline and an interactive Power BI dashboard. The objective is to identify key churn drivers, revenue impact, and high-risk customer segments.
- SQL: PostgreSQL
- Python: Pandas, SQLAlchemy
- BI Tool: Power BI
- Data Modeling: Star-style analytical table (Gold Layer)
- Raw Layer: Source datasets loaded from Pandas into PostgreSQL
- Silver Layer: Cleaned and standardized customer, service, and churn tables
- Gold Layer: Unified churn analytics table for BI & modeling
- Overall churn rate & revenue impact
- Churn by tenure and contract
- Customer distribution & CLTV comparison
- Top churn categories and reasons
- Churn by age group, gender, internet type
- Marital status & data plan segmentation
- Behavioral and demographic churn patterns
- ~27% churn rate with $3.68M revenue loss
- 53% churn occurs within first 6 months
- Month-to-month contracts dominate churn
- Competitor offerings and service quality are top churn reasons
- Satisfaction score is a strong churn predictor
- Strengthen onboarding in first 6 months
- Promote long-term contracts
- Improve competitive pricing and device offerings
- Monitor low satisfaction scores proactively
- Maven Analytics (Telecom Churn Dataset)
telecom-churn-analysis/
│
├── data/
│ └── gold_churn_analytics.csv
│
├── analysis/
│ ├── EDA.ipynb
│
├── images/
│ └── dashboard_screenshots.png
│
├── README.md
└── requirements.txt
Project Author: Koustubhya Singh
Learning Focus: data analytics, dashboard visualization
Project Duration: [4/02/26] - [8/02/26]
Connect:
- LinkedIn: [My Profile]
- GitHub: [My GitHub]
- Portfolio: [My Website]
⭐ If this project helped you learn data analysis, please star the repo!