Customer Churn Analysis using Logistic Regression

Project Overview

This project focuses on analyzing customer churn for a California-based Telco company. The aim is to understand the factors influencing churn and devise strategies to reduce the churn rate by at least 10% by year-end.

Introduction

Customer churn, a vital metric for business sustainability, especially in the telecom sector, signifies the loss of clients or customers. In response to a 15% increase in churn rate after Q3, this study utilizes a data-driven approach to mitigate customer attrition.

Dataset Overview

• Source: Kaggle-Telco-Customer-Churn and IBM Community
• Composition: 7043 observations across 33 attributes, encompassing demographic, service, and financial data
• Significant Attributes: 'Churn Label' as the dependent variable and various customer-related features as independent variables
• Preprocessing: Conversion of data types and handling of missing values, particularly in the 'Churn Reason' attribute

Exploratory Data Analysis (EDA)

• Geographical Influence: Heatmap analysis revealed higher churn rates in major cities like Los Angeles, San Francisco, and San Jose.
• Numerical Data Insights: Strong correlation of 'Churn Value' with 'Churn Score' and 'Tenure Months'
• Categorical Data Analysis: Crossplots highlighted the impact of contract type on churn, with month-to-month contracts showing higher churn rates.

Regression Analysis

Logistic regression was employed to model the probability of churn, considering over 30 features, including demographics, service usage, and geographical data.

Model Diagnosis and Selection

• Challenges: Addressed multicollinearity and influential data points to refine the model
• Final Model Selection: Based on Recall score, the reduced model (0.85) outperformed the full model (0.828), leading to its selection for further analysis

Key Findings

• Factors such as senior citizen status, phone service, contract length, and payment method significantly affect churn probabilities.
• Each additional month of tenure and certain payment methods notably decrease the odds of churning.

Recommendations

1. Senior Citizen Engagement: Implement targeted strategies to retain senior citizens.
2. Phone Service Enhancement: Improve and promote phone service features.
3. Contract Length Optimization: Encourage longer contract terms.
4. Payment Method Diversification: Offer incentives for preferred payment methods.
5. Churn Score Monitoring: Regularly evaluate churn scores to identify at-risk customers.

Conclusion

This analysis provides actionable insights into reducing customer churn, emphasizing the need for tailored strategies to enhance customer retention and satisfaction.

References

• Kaggle Dataset
• IBM Community