This project analyzes customer churn patterns using SQL and Power BI.
The goal is to identify factors contributing to customer churn and provide actionable insights for reducing churn.
Tools Used
- SQL
- Power BI
- Python (for preprocessing)
Key Metrics
- Total Customers
- Churned Customers
- Churn Rate
- Average Monthly Spend
A churn prediction model was built using Python to identify customers at risk of churn.
Steps:
- Data cleaning using pandas
- Feature engineering
- Train-test split
- Logistic regression model training
Libraries Used:
- pandas
- scikit-learn
- numpy
Output: The model predicts the probability of customer churn based on:
• Tenure • Monthly spend • Support tickets • Contract type
Dashboard Insights
- Customers with monthly contracts have the highest churn rate.
- Customers with shorter tenure show higher churn probability.
- Higher support tickets correlate with increased churn.
Visualizations
- Churn by Contract Type
- Churn by Tenure
- Support Tickets vs Churn
- Monthly Spend vs Churn
Project Structure
- SQL queries used for churn calculations
- Power BI dashboard for visualization
- Dataset sample used for analysis
Outcome
The dashboard helps stakeholders quickly identify high-risk customer segments and design retention strategies.