📊 E-commerce Revenue & Customer Analysis (RFM)

🔹 Overview

This project presents an end-to-end data analysis workflow focused on understanding revenue performance and customer behavior in an e-commerce dataset.

The analysis integrates Python (data processing), SQL (data querying), and Power BI (visualization) to deliver business insights through an interactive dashboard.

---

🎯 Objectives

• Analyze revenue trends over time
• Identify key geographic markets
• Understand customer purchasing behavior
• Segment customers using RFM (Recency, Frequency, Monetary)
• Build an executive-level dashboard for decision-making

---

🛠️ Tech Stack

• Python: Pandas, NumPy, Matplotlib
• SQL: SQLite
• Visualization: Power BI

📂 Project Structure

project/
│
├── data/
│   ├── raw/
│   └── processed/
│
├── notebooks/
│   ├── 01_data_cleaning.ipynb
│   ├── 02_eda.ipynb
│   ├── 03_sql_analysis.ipynb
│   ├── 04_exports.ipynb
│
├── models/
│   └── transactions.db
│
├── outputs/
│   ├── kpi_monthly.csv
│   ├── revenue_by_country.csv
│   ├── top_customers.csv
│   └── rfm_table.csv
│
└── dashboard/
    └── powerbi_dashboard.pbix

🔄 Data Processing

The dataset was cleaned and prepared to ensure analytical accuracy:

• Removed missing CustomerID values
• Excluded cancelled transactions (InvoiceNo starting with "C")
• Filtered invalid quantities and prices (≤ 0)
• Converted date fields to datetime format
• Created a Revenue metric (Quantity × UnitPrice)

---

📊 Analysis

The analysis was conducted in two stages:

Python (EDA)

• Revenue distribution and skewness
• Monthly revenue trends
• Customer purchasing patterns
• Orders per customer distribution

SQL (Validation & Aggregation)

• Revenue KPIs
• Revenue by country
• Monthly performance
• Customer-level aggregation

This ensured consistency and reproducibility across different analytical layers.

---

👥 Customer Segmentation (RFM)

Customers were segmented based on:

• Recency → Days since last purchase
• Frequency → Number of transactions
• Monetary → Total revenue generated

Key Findings

• Most customers exhibit low purchase frequency
• Revenue is highly concentrated in a small group of high-value customers
• A segment of customers shows inactivity, indicating potential churn risk

---

📈 Dashboard (Power BI)

The final solution is an interactive dashboard with two pages:

1️⃣ Executive Overview

• Revenue KPIs
• Monthly revenue trend
• Revenue by country
• Key business insights

2️⃣ Customer Analysis (RFM)

• Customer segmentation
• Revenue by segment
• Distribution of customer groups
• Identification of high-value and at-risk customers

---

🔍 Key Business Insights

• Revenue is highly concentrated in the UK (>80%), indicating geographic dependency
• Strong upward trend towards Q4 suggests seasonal demand patterns
• Customer base is dominated by low-frequency buyers
• A small segment of high-value customers drives the majority of revenue
• At-risk customers highlight opportunities for retention strategies

---

🚀 Conclusion

This project demonstrates the ability to:

• Build an end-to-end data pipeline (Python → SQL → Power BI)
• Translate raw data into business insights
• Design executive dashboards for decision-making
• Apply customer segmentation techniques (RFM)

---

📸 Dashboard Preview

Executive Revenue Overview

Customer Analysis

📦 Dataset

Due to file size limitations, the dataset is not included in this repository.

You can download it from:

• Data: https://1drv.ms/f/c/00af3ff0b1e3c79b/IgCbyjhVs6SITJNq2DHTUvGaASuPKPHyIHKb5FsFqHKC-mI?e=KANgko
• Models: https://1drv.ms/f/c/00af3ff0b1e3c79b/IgDO7x90NAMUQp4uZr6RxuIiAQ0TF9SX9APpQwJ_8BleaxM?e=mtQIug