📊 Customer Churn Analysis

🔎 Project Overview

This project analyzes customer churn behavior using a telecom customer dataset to identify key factors influencing customer attrition. The objective is to uncover patterns in customer behavior and provide data-driven retention strategies to reduce churn.

The dataset represents a snapshot of customer data at the end of a reporting quarter.

---

🎯 Business Problem

Customer churn directly impacts revenue and long-term business growth.
The goal of this analysis is to:

• Identify high-risk customer segments
• Understand the drivers behind churn
• Provide actionable business recommendations
• Support data-driven retention strategies

---

📁 Dataset Summary

• ~7,000 customer records
• 34 features including:
  • Tenure Months
  • Contract Type
  • Monthly Charges
  • Total Charges
  • Tech Support
  • Internet Service
  • Payment Method
  • Churn Label

---

🛠 Tools & Technologies Used

• Python
• Pandas
• NumPy
• Matplotlib
• Jupyter Notebook
• Exploratory Data Analysis (EDA)
• Customer Segmentation & Multi-Factor Analysis

---

📂 Repository Structure

customer-churn-analysis/
├── data/
│   └── Telco_customer_churn.xlsx
├── notebooks/
│   └── churnAnalysis.ipynb
├── .gitignore
└── README.md

---

▶️ How to Run

1. Clone the repository:

git clone https://github.com/anupamkajith/customer-churn-analysis.git
cd customer-churn-analysis

2. Install the required Python libraries:

pip install pandas numpy matplotlib openpyxl jupyter

3. Start Jupyter Notebook:

jupyter notebook

4. Open the notebook inside the notebooks/ folder.

5. Run the notebook cells sequentially to reproduce the customer churn analysis and visualizations.

Note: The dataset is stored inside the data/ folder. The notebook should use the relative dataset path:

../data/Telco_customer_churn.xlsx

---

📊 Key Insights

1️⃣ Overall Churn Rate

• Overall churn rate: 26.5%
• Approximately 1 in 4 customers leave the service.

---

2️⃣ Tenure Impact on Churn

• 0–6 months: 53% churn
• 48+ months: 9% churn

Insight: Customers in their early lifecycle stage are significantly more likely to churn.

---

3️⃣ Contract Type Impact

• Month-to-month: 42.7% churn
• One-year: 11.2% churn
• Two-year: 2.8% churn

Insight: Longer contract commitments dramatically reduce churn risk.

---

4️⃣ Tech Support Influence

• No Tech Support: 41.6% churn
• With Tech Support: 15.1% churn

Insight: Access to technical support significantly improves customer retention.

---

5️⃣ Pricing Sensitivity

• High monthly charges segment: ~33% churn
• Low monthly charges segment: ~9.8% churn

Insight: Customers paying higher monthly charges show increased churn likelihood, indicating possible price sensitivity.

---

🔍 Multi-Factor Risk Segmentation

The highest churn risk segment identified:

• Tenure: 0–6 months
• Contract: Month-to-month
• No Tech Support
• High Monthly Charges

This segment shows churn rates exceeding 50%, making it a priority for retention strategies.

---

💡 Business Recommendations

1. Strengthen onboarding and engagement for new customers.
2. Incentivize migration from month-to-month to long-term contracts.
3. Offer bundled technical support for high-risk segments.
4. Target high-paying customers with proactive retention campaigns.
5. Develop churn risk monitoring dashboards for early intervention.

---

📈 Future Enhancements

• Build predictive churn model (Logistic Regression / Random Forest)
• Perform feature importance analysis
• Develop churn risk scoring framework
• Create interactive dashboard (Power BI / Tableau)

---

🚀 Project Outcome

This analysis identified key churn drivers across tenure, contract type, technical support, and pricing, and translated them into actionable retention recommendations.

👤 Author

Anupam K Ajith
AI/ML & GenAI | M.Tech AI & Data Science @ IIIT Kottayam | Python | RAG | MLOps