Every business has customers who buy regularly, customers who are about to leave, and customers who have already gone quiet. This project uses a technique called RFM Analysis to automatically sort 1,000 customers into behaviour-based groups — so a business knows exactly who to focus on, who to win back, and who is at risk of churning.
RFM stands for: Recency (how recently did they buy?), Frequency (how often do they buy?), Monetary (how much do they spend?)
- Python — for all data processing and analysis
- Pandas — to clean, merge, and calculate RFM scores
- Matplotlib & Seaborn — to visualise the customer segments
- Google Colab / Jupyter Notebook — development environment
| File | What It Contains |
|---|---|
Customer_Master_Data.csv |
1,000 customers — name, age, city, gender, join date |
Customer_Transactions.csv |
23,050 transactions with dates and amounts |
| Customer Segment | What It Means (Plain English) | Count | % of Base |
|---|---|---|---|
| 🏆 Champions | Buy often, spend a lot, bought recently — your best customers | 127 | 12.7% |
| 💛 Loyal Customers | Buy regularly and spend well — strong relationship | 114 | 11.4% |
| 🌱 Potential Loyalists | Recently active but not yet frequent — nurture them | 252 | 25.2% |
| Used to buy often but haven't recently — about to leave | 207 | 20.7% | |
| 💤 Hibernating | Low activity across all three metrics — largely inactive | 273 | 27.3% |
| ➖ Others | Mixed signals — need further monitoring | 27 | 2.7% |
Total Revenue Analysed: ₹2.3 Crore across 23,050 transactions At-Risk Revenue (money that could be lost): ₹53.5 Lakh
1. The top 24% of customers (Champions + Loyal) are holding the business together. Only about 1 in 4 customers is truly engaged. This is a risky concentration — if these customers leave, revenue drops sharply.
2. ₹53.5 Lakh is sitting in the "At Risk" group right now. These 207 customers spent well in the past but have gone quiet. They have not left yet — but they will if no action is taken. A targeted win-back campaign could recover a significant portion of this revenue.
3. Hibernating customers (273 people) are the biggest re-engagement opportunity. This is the largest single group. A simple discount or loyalty offer could reactivate a meaningful percentage of them.
- Protect your Champions first — Create a VIP or loyalty reward program for the 127 Champions before they drift into the At Risk category.
- Launch a win-back campaign for At Risk customers — Personalised offers or reminders sent to the 207 At Risk customers could recover up to ₹16L+ if even 30% respond.
- Re-engage Hibernating customers with a low-cost campaign — A simple email with a discount code costs almost nothing but could reactivate hundreds of inactive buyers.
- Invest in Potential Loyalists — The 252 Potential Loyalists are the future Champions. Targeted engagement now will move them up the segment ladder.
📁 RFM-Customer-Segmentation-Python/
│
├── 📁 Data/
│ ├── Customer_Master_Data.csv
│ └── Customer_Transactions.csv
│
├── 📁 Pictures/
│ ├── segment_distribution.png
│ ├── revenue_per_segment.png
│ └── rfm_scatter.png
│
├── 📁 Notebook/
│ └── python_mini_project.ipynb
│
└── README.md