Telecom Customer Churn Analysis

This project explores customer churn behavior using a telecom dataset, with the goal of identifying the key drivers behind customer attrition and translating them into actionable business insights.

The analysis begins with data preparation and cleaning to ensure reliability. Key transformations include consolidating the Phone Service and Multiple Lines columns into a single, more meaningful categorical feature (Single Line, Multiple Lines, No Service), reducing redundancy and improving interpretability. Additionally, missing values in Total Charges were addressed by deriving them from Monthly Charges and tenure-related logic, ensuring consistency across financial metrics rather than relying on arbitrary imputation.

From there, the analysis focuses on how different variables influence churn rate. Financial factors such as monthly and total charges were examined to understand price sensitivity and whether higher-paying customers are more or less likely to leave. Contract type was a major focal point, revealing how commitment levels (monthly vs. long-term contracts) correlate with retention. Payment methods were also analyzed to uncover behavioral patterns—particularly whether automated payments reduce churn compared to manual methods.

Customer tenure was another critical dimension, helping segment users into lifecycle stages and highlighting when churn risk is highest. In parallel, the impact of subscribed services (e.g., internet service, add-ons, and bundled offerings) was explored to determine whether increased product engagement leads to stronger retention or potential dissatisfaction.

The project ultimately provides a multi-dimensional view of churn, combining behavioral, financial, and service-based factors. The insights can be used to inform retention strategies, such as targeting high-risk segments, optimizing pricing structures, and promoting contract upgrades or service bundles.

Dataset: https://www.kaggle.com/da...