E-Commerce Customer Intelligence Dashboard

An advanced Business Intelligence (BI) and Machine Learning solution designed to transform raw e-commerce data into strategic growth levers. This dashboard focuses on RFM Segmentation, Churn Prediction, and Automated Retention Strategies to maximize profitability and optimize marketing ROI.

---

Strategic Business Overview

In the e-commerce sector, the cost of customer acquisition (CAC) is rising. Retaining an existing customer is 5x more cost-effective than acquiring a new one.

This project provides a Profit-Driven Framework to:

• Maximize Revenue: Identifying high-value segments for targeted VIP treatment.
• Minimize Revenue Leakage: Predicting customer churn before it happens.
• Operational Excellence: Automating decision-making through an "Action Engine" that suggests immediate business interventions.

---

Tech Stack & Architecture

• Language: Python 3.x
• Dashboard: Streamlit (Interactive UI)
• Data Science: Pandas, NumPy
• Machine Learning: Scikit-Learn (Logistic Regression, KMeans Clustering)
• Visualization: Plotly, Seaborn (Dynamic and Publication-quality charts)

---

Core Business Intelligence & Analysis

RFM Data Intelligence

We analyze the customer base through three fundamental dimensions to understand the "Health" of our customer relationships:

• Recency: Time since last purchase (Engagement indicator).
• Frequency: Total transactions (Loyalty indicator).
• Monetary: Total revenue (Financial value).

Customer Segmentation (Targeting Strategy)

Using KMeans Clustering, we categorize customers into 4 personas to enable Precision Marketing:

1. Loyal Customers: High-frequency assets; focus on cross-selling.
2. Big Spenders: High monetary contribution; focus on premium upsells.
3. At Risk: Declining engagement; requires immediate reactivation offers.
4. New / Occasional: Recently acquired; requires onboarding nurture sequences.

Churn Prediction (Risk Mitigation)

We implemented a Logistic Regression model to calculate a Churn Risk Score (%).

• Key Discovery: Our analysis shows that Frequency (-2.62) has a significantly higher impact on retention than Monetary (-1.67).
• Business Insight: Building shopping habits is more valuable for long-term stability than isolated high-ticket sales.

Time-Based Behavioral Insights

Understanding when and how customers interact with the platform helps in timing our interventions.

• Recency Distribution: Helps management define the "Defection Point" to trigger recovery campaigns.
• Seasonality: Tracks monthly revenue trends to align marketing budgets with high-activity periods.

---

Business Impact & Value Proposition

• Enhanced ROI: By targeting only at-risk or high-value customers, we optimize marketing spend and avoid "blanket discounts."
• Revenue Recovery: The Action Engine identifies VIP customers at risk, triggering "Urgent VIP Calls" to save high-stakes accounts.
• Data-Driven Culture: Shifts the organization from "guessing" to "knowing" based on statistical significance.

---

Project Structure (Modules)

Designed for scalability and clean code maintenance, the project follows a modular architecture:

project/
│
├─ app.py                    # Main Business Dashboard (Streamlit UI)
├─ requirements.txt          # Environment dependencies
├─ data/                     # Transactional Datasets (CSV)
├─ reports/
│   └─ figures/              # Visual assets and performance charts
└─ src/                      # Production-grade modular logic
    ├─ load_and_clean_data.py
    ├─ compute_rfm.py
    ├─ scale_rfm.py
    ├─ apply_kmeans.py
    ├─ assign_segments.py
    └─ segment_statistics.py

How to Deploy

Clone the repository:

git clone <your-repo-url>

Install dependencies:

pip install -r requirements.txt

Launch the Dashboard:

streamlit run app.py