Customer Churn Analysis

Customer churn is a major concern for businesses that rely on subscription models. This analysis aims to identify customers likely to churn and develop a predictive model that provides actionable insights to reduce churn rates.

Data Collection

The dataset consists of customer demographics (age, gender, location), subscription details (start/end dates, service plans), usage patterns, customer service interactions, and churn status (churned or not churned).

Data Preprocessing

The following preprocessing steps were applied:

• Handling Missing Data: Missing demographic information, such as age and gender, was imputed using median values.
• Feature Engineering: Features like customer tenure, number of service interactions, and payment method were added.

Exploratory Data Analysis

The EDA revealed several key trends:

• Churn by Contract Type: Customers on month-to-month contracts exhibited the highest churn rates.
• Churn by Customer Service Interactions: Customers with frequent customer service requests were more likely to churn.
• Churn by Payment Method: Customers paying by credit card churned at a higher rate compared to those on automatic payments.

Model Development

The following models were developed:

• Logistic Regression: Used as the baseline model for classifying customers as churned or not churned.
• Random Forest & XGBoost: Applied to capture non-linear relationships and improve model accuracy.
• Hyperparameter Tuning: GridSearchCV was used to optimize the number of trees and the maximum depth for Random Forest and XGBoost models.

Model Evaluation

• Accuracy: 87%
• AUC-ROC: 0.92, indicating strong performance in distinguishing between churn and non-churn customers.
• Precision-Recall: The model achieved a precision of 0.85 and a recall of 0.81, showing a good balance between false positives and false negatives.

Conclusion & Recommendations

Key drivers of churn include contract type and customer service interactions. Businesses should offer incentives for customers to switch to annual contracts and improve customer service efficiency. A targeted retention campaign for customers interacting frequently with support is recommended.