🧠 Customer Intelligence & Churn Risk Dashboard

An end-to-end Machine Learning + Power BI project that transforms customer shopping behavior data into actionable retention insights.

This project uses Python for data analysis and feature engineering, builds a churn-risk prediction workflow, and prepares a Power BI-ready dataset for interactive business dashboards.

---

📌 Project Overview

This project analyzes customer shopping behavior and builds a customer intelligence solution that helps identify:

• Customers with high retention risk
• Customer segments with the highest business value
• Revenue potentially at risk
• Recommended retention actions for each customer group

The final output is a Power BI dashboard supported by a machine learning workflow built in Jupyter Notebook.

---

🎯 Project Objectives

• Clean and analyze customer shopping data
• Engineer business-focused customer features
• Build a churn-risk prediction model
• Generate customer-level predictions and recommendations
• Create a Power BI dashboard for decision-making
• Present a business-ready portfolio project for BI / Data / ML roles

---

📂 Dataset

Source file used: customer_shopping_behavior.csv

The dataset contains customer-level shopping behavior such as:

• Age
• Gender
• Item purchased
• Category
• Purchase amount
• Review rating
• Subscription status
• Shipping type
• Discount usage
• Promo code usage
• Previous purchases
• Payment method
• Frequency of purchases

---

🛠️ Tools & Technologies

Area	Tools Used
Data Analysis	Python, Pandas, NumPy
Visualization	Matplotlib, Seaborn, Power BI
Machine Learning	scikit-learn
Development	Jupyter Notebook
Dashboarding	Power BI Desktop
Version Control	Git, GitHub

---

🔄 Project Workflow

1. Data Understanding

• Loaded the CSV dataset in Jupyter Notebook
• Inspected shape, columns, null values, and data types
• Cleaned and standardized column names

2. Exploratory Data Analysis

• Analyzed customer behavior patterns
• Compared spending, subscription status, review ratings, and purchase frequency
• Identified trends useful for retention-risk analysis

3. Feature Engineering

Created new business-focused features such as:

• frequency_score
• engagement_score
• clv_proxy
• discount_dependency
• risk_score

4. Risk Modeling

• Built a churn-risk / retention-risk prediction workflow
• Generated customer-level probabilities and risk labels
• Classified customers into segments such as:
  • Champions
  • Regular
  • Potential Loyalist
  • At Risk
  • High Value - At Risk

5. Business Output Generation

Created a final Power BI-ready dataset with:

• risk_level
• churn_probability
• prediction_confidence
• customer_segment
• recommended_action
• revenue_at_risk
• retention_opportunity

6. Dashboard Development

Built a Power BI dashboard with pages such as:

• Executive Overview
• Risk Breakdown
• Customer Segments
• Action Center

---

📊 Power BI Dashboard Highlights

The dashboard helps answer business questions like:

• How many customers are at high risk?
• Which customer segments have the highest churn probability?
• Which product categories show higher retention risk?
• How much revenue is at risk?
• What action should the business take for each customer group?

Dashboard Pages

1. Executive Overview

• Total customers
• High-risk customers
• Revenue at risk
• Retention opportunity
• Risk-level distribution

2. Risk Breakdown

• Risk by category
• Risk by subscription status
• Risk distribution matrix
• Payment and shipping behavior analysis

3. Customer Segments

• Segment distribution
• CLV comparison by segment
• Engagement comparison by segment
• Scatter analysis of customer value vs churn risk

4. Action Center

• High-risk customer list
• Recommended actions
• Revenue at risk by action type
• Retention opportunity tracking

---

📁 Project Structure

customer-intelligence-dashboard/
│
├── data/
│   ├── customer_shopping_behavior.csv
│   └── customer_predictions_powerbi.xlsx
│
├── notebooks/
│   └── analysis.ipynb
│
├── screenshots/
│   ├── executive_overview.png
│   ├── risk_breakdown.png
│   ├── customer_segments.png
│   └── action_center.png
│
├── README.md
└── requirements.txt

---

🚀 How to Run This Project

1. Clone the repository

git clone https://github.com/YOUR_USERNAME/customer-intelligence-dashboard.git
cd customer-intelligence-dashboard

2. Install dependencies

pip install -r requirements.txt

3. Run the notebook

Open Jupyter Notebook and run:

jupyter notebook

Then open analysis.ipynb.

4. Open Power BI

• Open Power BI Desktop
• Import customer_predictions_powerbi.xlsx
• Build the dashboard using the prepared fields and measures

---

📈 Key Business Value

This project demonstrates how raw customer behavior data can be converted into business insights that support:

• customer retention strategy
• risk-based targeting
• revenue protection
• customer segmentation
• decision-making through BI dashboards

---

💼 Why This Project Matters

This is a portfolio-ready project for roles such as:

• Data Analyst
• Business Intelligence Intern
• Power BI Developer Intern
• Machine Learning Intern
• Data Science Intern

It shows practical skills in:

• data cleaning
• exploratory analysis
• feature engineering
• predictive analytics
• dashboard storytelling
• business insight generation

---

⚠️ Note

The churn field in this project is a risk-oriented modeled output, built from customer behavior indicators in the dataset. It should be interpreted as a customer retention risk score / churn-risk proxy, not as a confirmed real-world churn event label.

---

👨‍💻 Author

Rajdeep Chatterjee
Computer Science Student | Data Analytics | ML | Power BI | Full-Stack Development

---

⭐ Repository Usage

If you found this project useful, consider starring the repository.