🛒 Retail Customer Analytics Dashboard

Customer Segmentation & Purchase Pattern Analysis

📌 Project Overview

This project analyzes retail transaction data to:

• Segment customers based on their purchasing behavior using RFM Analysis and K-Means Clustering
• Visualize customer clusters using PCA (Principal Component Analysis)
• Discover product purchase patterns using Market Basket Analysis (Apriori Algorithm)
• Provide an interactive analytics dashboard using Streamlit

The goal is to support data-driven marketing and business strategy decisions.

---

🎯 Business Objectives

• Identify different customer segments (e.g., loyal, inactive, high spenders)
• Support marketing strategies such as:
  • Loyalty programs
  • Personalized promotions
  • Product bundling
• Understand which products are frequently purchased together

---

📂 Dataset

Online Retail Dataset
Transaction data with the following columns:

• InvoiceNo
• StockCode
• Description
• Quantity
• InvoiceDate
• UnitPrice
• CustomerID
• Country

Source: https://www.kaggle.com/datasets/ulrikthygepedersen/online-retail-dataset

---

🧹 Data Cleaning

The following preprocessing steps were applied:

• Removed duplicate records
• Removed rows with missing CustomerID and Description
• Converted InvoiceDate to datetime format
• Removed cancelled transactions (InvoiceNo starting with "C")
• Removed invalid values (Quantity ≤ 0, UnitPrice ≤ 0)
• Removed outliers using the IQR method
• Standardized text fields (Description, Country)

Feature engineering:

• Created TotalPrice = Quantity × UnitPrice

---

📊 RFM Analysis

RFM variables:

• Recency → Number of days since the last transaction
• Frequency → Number of unique invoices
• Monetary → Total spending

These variables represent customer purchasing behavior and are used as input for clustering.

---

🤖 Customer Segmentation

• Algorithm: K-Means Clustering
• Input features: Scaled RFM variables
• Visualization: PCA (2D scatter plot)
• Output: Cluster label for each customer

Cluster profiling is performed using the average RFM values for each cluster.

---

🛍 Market Basket Analysis

• Algorithm: Apriori
• Metrics used:
  • Support
  • Confidence
  • Lift
• Output:
  • Association rules such as:
    {Product A} → {Product B}

These rules can be used to support product recommendation and bundling strategies.

---

🖥 Dashboard Features

• Interactive selection of number of clusters (k)
• PCA scatter plot for cluster visualization
• Bar chart for cluster profiling
• Association rules table

---

🛠 Tech Stack

• Python
• Pandas, NumPy
• Scikit-learn
• Mlxtend
• Matplotlib
• Streamlit

---

🚀 How to Run

1. Clone the repository:

git clone https://github.com/yourusername/retail-customer-analytics.git
cd retail-customer-analytics

2. Install dependencies:

pip install -r requirements.txt

3. Project structure:

retail-customer-analytics/
│
├── app.py
├── requirements.txt
├── README.md
└── data/
    └── online_retail.csv

4. Run the Streamlit app:

streamlit run app.py