Overview

This project analyzes transactional data from an e-commerce marketplace to evaluate revenue growth, customer purchasing behavior, and retention patterns.
Using PostgreSQL and SQL, I designed an analytics-ready data model and conducted end-to-end analysis to identify what drives revenue and where growth opportunities exist.

Tools: PostgreSQL • SQL • Data Modeling • Cohort Analysis • RFM Segmentation

Problem

Marketplace growth alone does not indicate business health. Stakeholders needed clarity on:

• Is revenue growing sustainably?

• Are customers returning after their first purchase?

• What products and customers generate most revenue?

• Where should the business focus to improve performance?

Approach

The project followed a structured analytical workflow:

• Built a star schema analytics model from raw transactional data

• Validated data quality and relationships before analysis

• Analyzed revenue trends and customer purchasing patterns

• Performed cohort analysis to measure retention

• Segmented customers using RFM methodology

Key insights

• Growth is primarily driven by new customer acquisition, not retention.

• ~97% of customers make only one purchase.

• Average basket size is low (~1.14 items), indicating mission-based shopping.

• A small subset of products generates a large share of revenue.

• High-value customers are frequently becoming inactive (“at risk”).

Recommendations

• Focus on retention rather than acquisition alone.

• Introduce engagement campaigns within 60–90 days after purchase.

• Target high-value at-risk customers with reactivation strategies.

• Encourage multi-item purchases through bundling and promotions.

What this project demonstrates

• End-to-end analytical thinking

• SQL-driven business analysis

• Data modeling for analytics

• Translating data into actionable decisions