Customer Segmentation using RFM Analysis

This project implements Recency, Frequency, and Monetary (RFM) analysis to segment customers into distinct groups. The aim is to gain actionable insights to enhance marketing strategies and improve customer retention.

Project Overview

The primary goal of this project is to analyze customer purchase data using RFM metrics:

• Recency: How recently a customer made a purchase.
• Frequency: How often the customer makes purchases.
• Monetary: How much money the customer spends.

These metrics are used to classify customers into clusters that represent different value segments, guiding targeted marketing actions.

Contents

• Objectives
• Data Processing
• Segmentation
• Cluster Insights & Recommendations
• Marketing Recommendations
• Dataset Information
• Visualization
• Analysis Summary

Objectives

• Segment customers based on RFM metrics
• Provide targeted recommendations for marketing and customer retention
• Identify key characteristics of each customer segment

Data Processing

The project involves calculating RFM scores based on customer transaction history:

• Recency: Days since last purchase
• Frequency: Number of purchases
• Monetary: Total spend per customer

Segmentation

Using clustering algorithms, customers are segmented into:

1. High-Value Customers (Recent, frequent, and high monetary spending)
2. Mid-Value Customers (Moderately recent, regular frequency, moderate spend)
3. Low-Value Customers (Infrequent, low spend, less recent)
4. Dormant Customers (Rarely purchase, minimal recent interaction)

Cluster Insights & Recommendations

High-Value Customers

• Implement exclusive loyalty programs
• Offer personalized recommendations
• Conduct VIP events
• Special anniversary offers

Mid-Value Customers

• Introduce loyalty programs
• Run limited-time promotions
• Execute cross-sell campaigns
• Collect feedback surveys

Low-Value Customers

• Run targeted reactivation campaigns
• Offer bundle deals
• Provide educational content
• Limited-time discounts

Dormant Customers

• Execute win-back campaigns
• Provide reactivation incentives
• Collect feedback to understand dormancy
• Personalized re-engagement offers

Marketing Recommendations

• Tailor communication based on segments
• Engage actively on social media
• Implement feedback loops for strategy refinement
• Continuously monitor and adapt strategies

Dataset Information

• Total Records: 4372 rows, 8 columns
• Time Period: From December 1, 2010, to December 9, 2011 (373 days)
• Columns:
  • InvoiceNo
  • StockCode
  • Description
  • Quantity
  • InvoiceDate
  • UnitPrice
  • CustomerID
  • Country

Visualization

Visual summaries clearly depict segment distributions:

• High-value customers (1415)
• Mid-value customers (1466)
• Low-value customers (786)
• Dormant customers (705)

Analysis Summary

• Most orders occur on Thursdays around 12 PM.
• Maximum sales are observed in Autumn, likely due to holiday seasons.
• Top countries by order volume and product trends have been identified.
• Insufficient data available for payment, profitability, and customer satisfaction analyses.