Skip to content

Project 1: Customer Segmentation Analysis using K-Means clustering

Objective:
This project analyzes customer purchase behavior using RFM analysis and K-Means clustering, uncovering key customer groups and behavioral patterns.
The insights are designed to support marketing and CRM teams in personalized targeting, retention, and campaign optimization.

Data Overview

Data Period: April 1st – June 30th, 2025
Datasets Used: - sales: order-level data (sale_id, sale_date, customer_id, order_amount, country, channel)
- sales_item: item-level data (product_id, price, quantity, customer_id, sale_id)
- customer: demographic data (customer_id, age_range, signup_date, country)

Analytical Approach

  1. RFM Segmentation

  2. Recency, Frequency, Monetary scoring to categorize customers:

    • 🏆 Champions — high frequency, high recency, high value
    • 💎 Loyal customers — frequent and recent buyers
    • ⚠️ At-risk customers — long inactive but historically high spenders
    • 💔 Lost customers — inactive, low engagement

  3. Behavioral Deep Dive

  4. Purchase interval and time-of-day trends
  5. Channel reliance (e.g., website vs. app)
  6. Product category preference and average order value
  7. Demographic persona differences

  8. Lifecycle analysis: registration-to-first-purchase time

  9. Clustering with K-Means

  10. Behavioral feature normalization
  11. Elbow method for optimal k
  12. Cluster interpretation by product preference, spend, and recency

Key Insights

  • High-value customers (Champions) drive ~60% of total revenue; mostly aged 25–34, prefer direct web channel.
  • At-risk customers show a declining trend in frequency; targeted reactivation campaigns can be designed around seasonal promotions.
  • Loyal customers respond strongly to product category discounts, suggesting potential for personalized bundling.
  • Lost customers tend to have short early engagement — onboarding optimization could extend lifecycle.

GitHub Repository

🔗 View the full project on GitHub →

Includes: - Jupyter Notebook (customer_segmentation.ipynb) - Data structure overview - Python scripts and visualization code - Tableau export instructions

Tools & Stack

Category Tools Used
Data Cleaning pandas, numpy
Analysis scikit-learn (KMeans), matplotlib, seaborn
Segmentation RFM scoring, clustering
Visualization Tableau, matplotlib
Documentation MkDocs (Material theme)

Next Steps

  • Integrate real-time RFM updates into CRM dashboard
  • Automate customer churn alerts using scheduling tools
  • A/B test personalized marketing strategies per segment