Final Report on Telecom Churn

Analysis

EDA REPORT

PAGE 1

Observation Summary

This report presents the analysis of a telecom company's customer data to identify patterns

of customer churn. Churn refers to customers discontinuing their services, which has a

significant impact on the company's revenue. After conducting detailed exploratory data

analysis (EDA), several key insights emerged:

1. High Churn Rate among Senior Citizens and Short-Tenure Customers:

Senior citizens and customers with less than a year of tenure tend to churn

more frequently. These groups may benefit from targeted retention

strategies, such as personalized offers or improved customer service.

2. Impact of Service Features on Churn: Lack of value-added services like

tech support, online security, and device protection strongly correlates with

churn. Customers who lack these services show a higher tendency to leave,

suggesting a need for bundling or promoting these services more effectively.

3. Price Sensitivity: Customers with higher monthly charges tend to churn at

a greater rate. This suggests that the pricing structure may be perceived as

too expensive or that these customers are not seeing enough value for the

cost. Offering loyalty rewards, discounts, or customizable packages may help

in retaining these customers.

PAGE 2

Full Report on Telecom Churn Analysis

1. Overview of the Dataset

− The dataset consists of customer information for a telecom company, aimed at

analysing churn. The key columns include:

 Customer Demographics: customerID, gender, SeniorCitizen, Partner, Dependents

 Services: PhoneService, MultipleLines, InternetService, OnlineSecurity, etc.

 Churn Information: Churn (target variable)

2. General Data Cleaning and Preprocessing

− The data cleaning involved:

 Handling missing values.

 Converting categorical data to numerical values for analysis.

3. Churn Distribution Analysis

A count plot was generated to understand the distribution of customers who churned versus

those who stayed with the company. The majority of customers did not churn, but a

significant portion did, which warrants further investigation.

Observation: The churn rate is moderately high and may indicate issues with customer

satisfaction or service features.

PAGE 3

5. Feature-Wise Churn Impact

 Senior Citizens: A higher proportion of senior citizens tend to churn.

 Dependents: Customers with dependents churn less frequently than those

without.

 Tech Support and Online Security: Customers without tech support and online

security services tend to churn at higher rates.

PAGE 4

6. Tenure and Monthly Charges

 Tenure: Customers with shorter tenure (less than a year) have a much higher

churn rate. Longer-tenured customers tend to stay loyal.

 Monthly Charges: Higher monthly charges are correlated with increased churn,

possibly indicating dissatisfaction with pricing.

7. Key Takeaways and Recommendations

 Service Quality Improvement: Customers without tech support and online

security churn more, suggesting that improving or bundling these services could

reduce churn.

 Pricing Strategy: High churn rates for customers with higher monthly charges

highlight a need to reassess pricing or offer loyalty discounts.

 Targeted Retention for New Customers: Since new customers tend to churn

more, introducing personalized retention programs early could help mitigate

churn.

 Retention Programs for new and senior customers,

 Offering discounted or tiered service packages, and

 Promoting value-added services like online security and tech support will

significantly reduce churn.

Implementing these strategies could not only improve customer satisfaction but also drive

long-term loyalty.