📉 Subscription Churn Analysis (OTT / SaaS)

An end-to-end churn analytics project that identifies churn drivers, predicts churn probability, and segments high-risk customers using the Telco Customer Churn dataset (adapted to OTT/SaaS PRD requirements).

✨ Features

• 📊 Churn KPI Analysis: Calculates overall churn rate and group-level churn trends.
• 🧹 Data Cleaning Pipeline: Handles missing values and type corrections.
• 🧠 Feature Engineering: Builds engagement score, tenure groups, and churn-ready features.
• 📈 Rich Visualizations: Tenure, usage, pricing, support behavior, and correlation heatmap.
• 🤖 ML Modeling: Logistic Regression and Random Forest model training + evaluation.
• 🎯 Risk Segmentation: Generates churn probabilities and classifies users into Low/Medium/High risk.
• 💼 Business Insights: Produces actionable retention recommendations.

🧰 Tech Stack

• Python
• Pandas, NumPy
• Matplotlib, Seaborn
• Scikit-learn
• Jupyter Notebook

📁 Project Structure

Subscription_Churn_Analysis/
├── Telco_Customer_Churn.csv
├── README.md
├── analysis_summary.md
├── subscription_churn.csv
├── subscription_churn.ipynb
├── subscription_churn.py
├── outputs/
│ ├── churn_rate.png
│ ├── tenure_churn.png
│ ├── usage_churn.png
│ ├── charges_churn.png
│ ├── support_churn.png
│ ├── correlation_heatmap.png
│ ├── model_accuracy.png
│ └── feature_importance.png
└── requirements.txt

🚀 Installation

1. Clone the repository

git clone https://github.com/https://github.com/bikram73/Subscription_Churn_Analysis.git
cd Subscription_Churn_Analysis

2. Create and activate virtual environment (recommended)

python -m venv .venv

Windows PowerShell:

.venv\Scripts\Activate.ps1

3. Install dependencies

pip install -r requirements.txt

▶️ Usage

Run Python Pipeline

python subscription_churn.py

This generates:

• subscription_churn.csv (normalized + scored dataset)
• analysis_summary.md (insights + model performance + recommendations)
• outputs/*.png (all required charts)

Run Notebook Workflow

Open subscription_churn.ipynb and run all cells in order.

• Notebook visualizations are shown inline.
• Notebook cells are designed to analyze without saving charts to outputs/.

📌 Key Objectives Covered

• Calculate churn rate
• Identify major churn factors
• Segment high-risk users
• Build and evaluate prediction models
• Recommend retention strategies

🧾 Dataset Note

Source dataset (Kaggle): https://www.kaggle.com/datasets/blastchar/telco-customer-churn

The Telco dataset does not directly contain OTT telemetry fields like exact app usage frequency, last login days, and customer support call counts.

📜 License

This project is licensed under the MIT License.

• See LICENSE