Customer Segmentation Analysis

Project Overview

This project involved analyzing customer behavior patterns from a UK-based online retail dataset to identify distinct customer segments using RFM (Recency, Frequency, Monetary) analysis. The goal was to help the business understand their customer base better and develop targeted marketing strategies.

The analysis revealed 6 distinct customer segments, each with unique characteristics and business implications. The findings were visualized using Tableau dashboards to provide actionable insights for business decision-making.

Python Pandas NumPy Scikit-learn Tableau RFM Analysis K-means Clustering Data Visualization

Problem Statement

The UK online retail company needed to understand their customer base better to:

Identify high-value customers for retention strategies
Develop targeted marketing campaigns for different customer groups
Optimize resource allocation based on customer profitability
Predict customer churn and implement preventive measures
Improve overall customer lifetime value

Methodology

1. Data Preprocessing

Cleaned and prepared the dataset by handling missing values, removing duplicates, and standardizing data formats. The dataset contained transaction records with customer IDs, purchase dates, quantities, and monetary values.

# Data preprocessing example
import pandas as pd
import numpy as np

# Load and clean data
df = pd.read_excel('Online Retail.xlsx')
df = df.dropna(subset=['CustomerID'])
df = df[df['Quantity'] > 0]
df['TotalAmount'] = df['Quantity'] * df['UnitPrice']
            

2. RFM Analysis

Implemented RFM (Recency, Frequency, Monetary) analysis to evaluate customer value:

Recency: Days since last purchase
Frequency: Number of transactions
Monetary: Total amount spent

# RFM Analysis
from datetime import datetime, timedelta

# Calculate RFM metrics
snapshot_date = df['InvoiceDate'].max() + timedelta(days=1)
rfm = df.groupby('CustomerID').agg({
    'InvoiceDate': lambda x: (snapshot_date - x.max()).days,
    'InvoiceNo': 'count',
    'TotalAmount': 'sum'
}).rename(columns={
    'InvoiceDate': 'Recency',
    'InvoiceNo': 'Frequency',
    'TotalAmount': 'Monetary'
})
            

3. Customer Segmentation

Applied K-means clustering to group customers into distinct segments based on their RFM scores. The optimal number of clusters was determined using the elbow method and silhouette analysis.

# K-means clustering
from sklearn.cluster import KMeans
from sklearn.preprocessing import StandardScaler

# Standardize features
scaler = StandardScaler()
rfm_scaled = scaler.fit_transform(rfm)

# Apply K-means clustering
kmeans = KMeans(n_clusters=6, random_state=42)
rfm['Cluster'] = kmeans.fit_predict(rfm_scaled)
            

Key Results

6

Customer Segments Identified

23%

Revenue from Champions

18%

Customers at Risk

£2.1M

Total Revenue Analyzed

Customer Segments Identified:

Champions: High-value, frequent customers who recently made purchases
Loyal Customers: Regular customers with good recency and frequency
At Risk: Previously valuable customers showing declining engagement
Can't Lose: High-value customers who haven't purchased recently
About to Sleep: Customers with declining frequency and recency
Lost: Customers with low recency, frequency, and monetary value

Data Visualizations

Created comprehensive Tableau dashboards to visualize customer segments and their characteristics:

Revenue by Segment Bar Chart

Interactive Tableau visualization showing revenue distribution across customer segments

Customer Distribution Pie Chart

Visual representation of customer segment proportions

Monthly Revenue Trends

Time series analysis showing revenue patterns by segment over time

RFM Scatter Plot

Three-dimensional visualization of Recency, Frequency, and Monetary values

Business Impact & Recommendations

Strategic Recommendations:

Champions: Implement VIP programs and exclusive offers to maintain loyalty
Loyal Customers: Upsell and cross-sell opportunities to increase value
At Risk: Immediate re-engagement campaigns with personalized offers
Can't Lose: Win-back campaigns with significant incentives
About to Sleep: Reactivation campaigns with product recommendations
Lost: Low-cost retention efforts or consider them as acquisition targets

Expected Outcomes:

15-20% increase in customer retention rates
25% improvement in marketing campaign effectiveness
10-15% increase in average customer lifetime value
Better resource allocation and ROI on marketing spend

Technical Implementation

Data Pipeline:

Data extraction from Excel files
Data cleaning and preprocessing using Pandas
RFM metric calculation and scoring
K-means clustering implementation
Segment analysis and interpretation
Tableau dashboard creation and deployment

Key Libraries Used:

# Key imports
import pandas as pd
import numpy as np
import matplotlib.pyplot as plt
import seaborn as sns
from sklearn.cluster import KMeans
from sklearn.preprocessing import StandardScaler
from sklearn.metrics import silhouette_score
import warnings
warnings.filterwarnings('ignore')
            

Lessons Learned

Data quality is crucial for meaningful segmentation results
Business context is essential for interpreting clustering results
Visualization plays a key role in communicating insights to stakeholders
Regular model updates are necessary as customer behavior evolves
Combining statistical analysis with business knowledge yields better results