👥 Customer Segmentation & Churn Prediction (Python Project)

Understanding customer behavior is vital for enhancing retention strategies. This project aims to:

• Segment customers based on purchasing behavior using RFM analysis.
• Identify high-risk customers likely to churn.
• Develop predictive models to anticipate customer churn.

---

📈 Project Overview

• This project leverages Python to perform customer segmentation using RFM (Recency, Frequency, Monetary) analysis and predicts customer churn using machine learning techniques. The analysis is based on e-commerce transaction data.

---

📦 Dataset

• Source: Kaggle – E-Commerce Data
• The dataset was originally created by the UC Irvine Machine Learning Repository.
• The dataset includes transactions from an online retailer between 01/12/2010 and 09/12/2011 and contains fields such as InvoiceNo, StockCode, Description, Quantity, InvoiceDate, UnitPrice, CustomerID, and Country.

---

💡 Key Steps

• 📊 RFM Analysis
  • Calculated Recency, Frequency, and Monetary Value for each customer
  • Standardized features for clustering
• 📈 Customer Segmentation (K-Means)
  • Used the Elbow Method to find optimal number of clusters
  • Segmented customers into 4 behavior-based groups
• 🔮 Churn Prediction
  • Labeled churned customers based on recency > 180 days
  • Trained a Random Forest classifier to predict churn
  • Evaluated model performance with precision, recall, F1-score

---

🔍 Key Python Queries

• STEP 1: Load and Explore Dataset

  import pandas as pd
  import matplotlib.pyplot as plt
  import seaborn as sns
  import datetime

• Read the Dataset

  df = pd.read_csv(r"C:\Users\sgand\OneDrive\Documents\Data Analysis\Python\Customer Segmentation and Churn 
  Prediction\ecommerce-data.csv")

• Basic Cleanup

  df.dropna(subset=['CustomerID'], inplace=True)
  df['InvoiceDate'] = pd.to_datetime(df['InvoiceDate'])
  df['TotalPrice'] = df['Quantity'] * df['UnitPrice']
  df = df[~df['InvoiceNo'].astype(str).str.startswith('C')]

• STEP 2: RFM Analysis

  snapshot_date = df['InvoiceDate'].max() + datetime.timedelta(days=1)
  rfm = df.groupby('CustomerID').agg({
      'InvoiceDate': lambda x: (snapshot_date - x.max()).days,
      'InvoiceNo': 'nunique',
      'TotalPrice': 'sum'
  })
  rfm.rename(columns={
      'InvoiceDate': 'Recency',
      'InvoiceNo': 'Frequency',
      'TotalPrice': 'MonetaryValue'
  }, inplace=True)

• STEP 3: Scaling and Clustering

  from sklearn.preprocessing import StandardScaler
  from sklearn.cluster import KMeans
  
  scaler = StandardScaler()
  rfm_scaled = scaler.fit_transform(rfm)

• Elbow Method

  sse = {}
  for k in range(1, 10):
      kmeans = KMeans(n_clusters=k, random_state=1)
      kmeans.fit(rfm_scaled)
      sse[k] = kmeans.inertia_
  
  plt.figure(figsize=(8,5))
  plt.plot(list(sse.keys()), list(sse.values()), marker='o')
  plt.xlabel("Number of clusters")
  plt.ylabel("SSE")
  plt.title("Elbow Method for Optimal K")
  plt.show()

• Apply KMeans with K=4

  kmeans = KMeans(n_clusters=4, random_state=1)
  rfm['Cluster'] = kmeans.fit_predict(rfm_scaled)

• Visualize Clusters

  sns.pairplot(rfm.reset_index(), hue='Cluster', palette='Set1', height=3)
  plt.savefig(r"C:\Users\sgand\OneDrive\Documents\Data Analysis\Python\Customer Segmentation and Churn 
  Prediction/clusters_plot.png")

• STEP 4: Churn Prediction

  from sklearn.model_selection import train_test_split
  from sklearn.ensemble import RandomForestClassifier
  from sklearn.metrics import classification_report
  import joblib

• Define Churn Label (Recency > 180 days)

  rfm['Churn'] = rfm['Recency'].apply(lambda x: 1 if x > 180 else 0)

• Train/Test Split

  model = RandomForestClassifier()
  model.fit(X_train, y_train)

• Evaluate

  y_pred = model.predict(X_test)
  print(classification_report(y_test, y_pred))

• Save Model

  joblib.dump(model, r"C:\Users\sgand\OneDrive\Documents\Data Analysis\Python\Customer Segmentation and Churn  
  Prediction/customer-segmentation-churn_model.pkl")

---

📂 Project Structure

project-python-customer-segmentation-churn-prediction/
├── Data/
│   └── ecommerce-data.csv                   # Original Kaggle dataset
│
├── Notebooks/
│   └── customer_segmentation.ipynb          # Complete Jupyter Notebook analysis
│
├── Source/
│   ├── data_preprocessing.py                # Cleaning, feature engineering, total price calculation
│   ├── rfm_analysis.py                      # Recency, Frequency, Monetary value calculation
│   ├── clustering.py                        # K-Means clustering logic and visualizations
│   └── churn_prediction.py                  # Model training, evaluation, and saving
│
├── Models/
│   └── churn_model.pkl                      # Trained Random Forest model
│
├── Visuals/
│   └── clusters_plot.png                    # Visualization of customer segments
│
├── Reports/
│   └── insights_summary.md                  # Business-style insights and summary report
│
├── requirements.txt                         # Python libraries needed to run this project
├── .gitignore                               # Ignore checkpoints, system files, and data
├── README.md                                # Full project overview and usage guide
└── LICENSE                                  # MIT License file

---

📊 Visualizations

---

📌 Key Insights

• High-value customers often have low recency (recent activity) and high frequency
• A cluster of customers showed high spend but long inactivity → ideal for retention targeting
• The churn prediction model achieved strong recall on identifying at-risk customers

---

🚀 How to Use

• Clone this repository:
  • git clone https://github.com/sgandhi797/project-python-customer-segmentation-churn-prediction.git
  • cd project-python-customer-segmentation-churn-prediction
• Install requirements:
  • Download Anaconda Navigator
  • Install Jupyter Notebook from the Navigator
• Open and run the notebook:
  • jupyter notebook/Project - Jupyter Notebook - Customer Segmentation and Churn Prediction.ipynb

---

📚 Tools & Technologies

• Python 3
• Pandas and NumPy for data handling
• Matplotlib and Seaborn for visualization
• Scikit-learn for clustering and classification
• Jupyter Notebook for interactive analysis

---

📄 License

• This project is licensed under the MIT License.