Telco Customer Churn Analysis

This project focuses on analyzing customer churn for a telecommunications company using exploratory data analysis (EDA). It provides key insights and actionable recommendations to help improve customer retention strategies.

Objective

Identify key factors influencing customer churn and recommend strategies to reduce churn.

Dataset

• Source: IBM Sample Data Sets (via Kaggle)
• This dataset is provided by IBM and publicly available on Kaggle.
• After downloading, rename the file to: TelcoCustomerChurn.csv
• Place it in the same directory as the notebook: Telco_Customer_Churn.ipynb
• For a detailed description of the dataset (columns, data types, etc.), please refer to the notebook.

Methodology

• Understanding the dataset
• Data Cleaning and Preprocessing
• Univariate and bivariate analysis
• Feature engineering
• Key insights and recommendations

Tools Used

• Python
• Pandas
• Matplotlib
• Seaborn
• JupyterLab

Key Insights

• Customers with month-to-month contracts and short tenure are more likely to churn
• Internet service (especially Fiber optic) users show higher churn risk, particularly those lacking online security, tech support, online backup or device protection.
• Customers paying via electronic check have higher churn probability
• Service-related features have stronger impact on churn than demographic features

View the Jupyter Notebook

• The notebook can be previewed directly on GitHub (along with outputs and graphs). Just navigate to the repository and open the Telco_Customer_Churn.ipynb file by clicking on it. Or you can click here.