🏦 Bank Customer Analytics & Risk Intelligence

📊 Project Overview

An end-to-end banking analytics project analyzing 45,211 customers to predict term deposit subscription, identify customer segments, and assess credit risk.

🛠️ Tools Used

Tool	Purpose
Excel / Power Query	Data cleaning & scenario analysis
SQL Server (SSMS)	Database design & analytical queries
Python 3.10	EDA, ML prediction, RFM segmentation
Power BI	5-page interactive dashboard

📁 Dataset

• Source: UCI Machine Learning Repository
• Dataset: Bank Marketing Dataset
• Size: 45,211 customer records
• Features: 17 columns

🔑 Key Findings

Subscription Analysis

• Overall subscription rate: 11.7%
• Students have the highest rate: 28.7%
• March is the best month: 52.0%
• Calls over 10 mins convert at: 48.4%

ML Model Performance

• Algorithm: Random Forest Classifier
• Accuracy: 84.3%
• AUC Score: 0.916 (Excellent)
• Top predictor: Call Duration (50.4%)

RFM Segmentation

• Champions (3.4%): 31.6% subscription rate
• Average balance: $3,366
• Lost customers (40%): Only $3 avg balance
• Biggest opportunity: Need Attention segment

Risk Analysis

• Default rate: 1.8% (815 customers)
• High Risk customers: all have a negative balance
• Previous campaign success: 64.7% conversion

📊 Dashboard Pages

1. Executive Summary — KPIs, job analysis, risk
2. Customer Segmentation — RFM segments
3. Churn & Risk Analysis — ML predictions
4. Campaign Performance — Monthly trends
5. ML Predictions — Feature importance

💡 Business Recommendations

1. Focus on March — 52% subscription rate
2. Train agents — longer calls = more subscriptions
3. Re-target previous successes — 64.7% rate
4. Champion retention — highest value segment
5. Re-engage Need Attention — high balance, low engagement

🚀 How to Run

Prerequisites

pip install -r requirements.txt

Steps

1. Clone the repository
2. Download the UCI Bank Marketing dataset
3. Run Power Query cleaning in Excel
4. Import data to SQL Server
5. Run Python scripts in order (01-04)
6. Open Power BI dashboard

Python Scripts Order

python 01_eda.py
python 02_eda_advanced.py
python 03_churn_prediction.py
python 04_rfm_segmentation.py

📸 Dashboard Screenshots

Executive Summary

Customer Segmentation

Churn & Risk Analysis

Campaign Performance

ML Predictions