This project focuses on customer segmentation for an online retail dataset using the RFM (Recency, Frequency, Monetary) model and K-Means clustering.
The goal is to identify distinct customer groups based on purchasing behavior and derive actionable business insights.
The methodology is inspired by the research paper:
Chen et al., “Data mining for the online retail industry: A case study of RFM model-based customer segmentation”, Journal of Database Marketing, 2012.
- Online retail transactional data
- Each transaction includes invoice details, purchase quantity, price, invoice date, and customer identifier.
- Data is preprocessed to compute:
- Recency: Time since last purchase
- Frequency: Number of transactions
- Monetary: Total spending per customer
-
Data Cleaning & Preprocessing
- Removed invalid and missing records
- Filtered relevant transactions
- Aggregated transactional data at customer level
-
Feature Engineering
- Computed RFM features for each customer
- Applied feature scaling to handle differing magnitudes
-
Clustering
- Applied K-Means clustering
- Evaluated multiple values of k
- Interpreted clusters based on RFM characteristics
-
Visualization & Interpretation
- Visualized cluster distributions
- Identified high-value, loyal, medium-value, and low-value customer segments
- Successfully segmented customers into distinct behavioral groups
- Identified a small segment contributing disproportionately high revenue
- Highlighted customer groups requiring retention or re-engagement strategies
- Demonstrated practical application of clustering for customer-centric marketing
- Python
- Pandas, NumPy
- Scikit-learn
- Matplotlib / Seaborn
- Jupyter Notebook
Chen, D., Sain, S. L., & Guo, K. (2012).
Data mining for the online retail industry: A case study of RFM model-based customer segmentation.
Journal of Database Marketing & Customer Strategy Management.
Harshit Paramhans