Customer Segmentation & Personalized Offer Optimization - Banking CRM Analytics

Segments retail banking customers using behavioral data and recommends next-best financial offers to improve campaign conversion, cross-sell, and retention.

---

Business Problem

Banks run marketing campaigns to promote savings accounts, credit cards, and deposits. Generic campaigns result in low conversion and high costs.

Goal:

• Identify meaningful customer segments
• Personalize offers
• Improve marketing ROI

---

Solution Overview

This project simulates a real banking CRM analytics workflow:

1. Exploratory Data Analysis
2. Feature Engineering (behavioral metrics)
3. Customer Segmentation (KMeans clustering)
4. Persona Profiling
5. Offer Recommendation Engine (Next Best Offer)
6. Campaign Effectiveness Evaluation (A/B Testing Simulation)
7. Power BI Dashboard
8. Interactive Streamlit App

---

Tech Stack

• Python (Pandas, Scikit-learn)
• Clustering (KMeans)
• Power BI
• Streamlit
• GitHub

---

Project Structure

data/ → raw & processed datasets

notebooks/ → analytics pipeline

src/ → reusable recommendation logic

dashboards/ → Power BI reports

app/ → Streamlit demo

---

Dashboard

Interactive Power BI dashboard shows:

• Segment distribution
• Persona characteristics
• Offer targeting
• Conversion insights

Streamlit App

Interactive tool to input customer behavior and get:

• Segment classification
• Recommended offer

Run locally:

streamlit run app/streamlit_app.py

Business Impact (Simulated)

• Identified 4 actionable customer personas
• Personalized offers by segment
• Improved expected conversion by ~30–40%
• Demonstrated additional uplift using A/B testing simulation

---

How to Run

pip install -r requirements.txt
streamlit run app/streamlit_app.py