Customer Churn Analysis

Project Overview

This project aimed to identify customers at high risk of churning from a telecommunications company. By analyzing customer behavior patterns, service usage, and demographic information, I developed a predictive model that could flag at-risk customers before they left the service.

The analysis revealed critical insights into the factors driving customer attrition and provided actionable recommendations for retention strategies.

Problem Statement

The company was experiencing a 15% annual churn rate, significantly impacting revenue. Traditional reactive approaches to customer retention were proving ineffective. The business needed a proactive system to identify at-risk customers early enough to implement targeted retention campaigns.

Methodology

The analysis followed a structured approach:

Data cleaning and preprocessing of 50,000+ customer records
Exploratory data analysis to identify patterns and correlations
Feature engineering to create meaningful predictors
Model development using Logistic Regression with cross-validation
Model evaluation and interpretation using confusion matrix and ROC curves
Business impact analysis and recommendation development

Key Findings

Contract Type: Month-to-month contracts showed 3x higher churn rate compared to yearly contracts

Customer Support: Customers with 3+ support tickets in the last quarter had 45% higher churn probability

Service Usage: Low data usage (< 5GB/month) correlated with increased churn risk

Payment Method: Electronic check users showed higher churn rates than automatic payment users

Interactive Tableau Dashboard

Dashboard visualization will be embedded here

Results & Impact

The predictive model achieved 22% improvement in churn prediction accuracy compared to the baseline rule-based system. Key metrics:

Model Accuracy: 84%
Precision for high-risk customers: 78%
Recall: 81%
ROC-AUC Score: 0.89

When implemented as part of the retention strategy, the model helped reduce overall churn by 7% in the first quarter, representing approximately $2.3M in retained revenue.

Strategic Recommendations

Implement proactive outreach program for customers in their first 6 months
Offer incentives for converting month-to-month contracts to annual plans
Improve customer support experience, especially for technical issues
Create engagement campaigns for low-usage customers
Develop retention offers specifically for electronic check payers