Customer Churn Analysis using Python (EDA Project) 

In this project, I performed an Exploratory Data Analysis (EDA) on a telecom customer dataset to understand why customers are churning. Using Python, I analyzed demographic factors, service usage patterns, payment methods, and contract types to identify major churn influencers. 

 TOOLS USED:

• Python (Pandas, NumPy)
  
  

• Matplotlib & Seaborn (visualization)
  
  

• Google Colab
  
  

• EDA Techniques: cleaning, transformation, univariate & bivariate analysis
  
  

THINGS I DID FOR MEANINGFUL INSIGHTS:

• Loaded dataset into Google Colab.

• Cleaned missing values (Total Charges column fixed).

• Converted numeric categories (Senior Citizen → Yes/No).

• Checked duplicates & datatypes.

• Performed univariate analysis (Churn count, gender, tenure).

• Performed bivariate analysis with churn (contract type, payment method, services).

• Visualized relationships using bar plots, pie charts, stacked plots.

• Derived churn-driving patterns for business decision making.

 SOME SNAPSHOTS:

I found that month-to-month contract customers churn significantly more than those on one-year or two-year contracts.

This suggests that flexible customers are more sensitive to pricing, service quality, and promotions, while long-term contract users are more stable and satisfied. 

• Customers who pay via Electronic Check have the highest churn rate.

• This indicates friction or dissatisfaction associated with this payment method — possibly higher fees, poor convenience, or trust issues. In contrast, customers using Bank Transfer, Credit Card, or Mailed Check show significantly lower churn, indicating these methods are more stable and preferred.

.

CONCLUSION:

After going through the entire dataset and visualizing different patterns, I realised that churn isn’t random at all — it clearly increases in specific groups. Customers with short tenure, month-to-month contracts, and electronic check payments were leaving the most. On the other hand, people who stayed longer, had yearly contracts, or used secure payment methods were far more loyal.

So overall, this analysis helped me understand exactly who is at risk of leaving . These insights can genuinely help the business fix early-stage issues, design better plans, and improve the overall customer experience.