This project analyzes user behavior in an e-commerce platform to identify conversion bottlenecks, customer segments, and revenue opportunities. The analysis follows an end-to-end workflow using Snowflake for data transformation and Power BI for visualization, simulating a real-world analytics pipeline.
- Identify where users drop off in the funnel
- Understand customer behavior and segmentation
- Evaluate category and pricing performance
- Provide actionable recommendations to improve conversion and revenue
- Event-level transactional data
- ~885,000 records
- Time range: Sep 2020 – Feb 2021
- Each row represents a user interaction (view, cart, purchase)
Raw Event Data ↓ Snowflake (SQL Transformation) ↓ Analysis-Ready Dataset (View) ↓ Power BI Dashboard
- Converted timestamps for time-based analysis
- Removed duplicate records using window functions
- Handled missing values:
- ~27% missing category
- ~24% missing brand
- Standardized categorical fields
- Created derived features:
- price buckets
- customer segments
- session-level metrics
- 90% views → 6.11% carts → 4.22% purchases
- Major drop-off at the view to cart stage
- Cart to purchase conversion is strong (~69%)
Insight:
The main issue is user engagement, not checkout.
- ~95% of users are non-buyers
- Revenue is driven by:
- One-time buyers
- High-value customers
- Repeat customers
Insight:
Revenue is concentrated among a small group of users.
- Strong performance: computers, electronics
- Weak performance: apparel, appliances
- Low-priced items convert best
- Luxury items have the lowest conversion
Insight:
Conversion is influenced by price and product positioning.
Cart Abandonment:
- ~50% of users abandon after adding to cart
Session Behavior:
- Short sessions: 1.94% conversion
- Long sessions: 47.40% conversion
Insight:
Conversion increases with engagement depth.
- Improve product pages (images, descriptions, reviews)
- Enhance product discovery and recommendations
- Optimize pricing (discounts, bundling, financing)
- Reduce decision friction (comparison tools, clarity)
- Recover abandoned carts via retargeting
- Focus on converting low-intent users
- Strengthen retention strategies
- Built SQL transformation pipelines using CTEs in Snowflake
- Created analysis-ready datasets using views
- Designed structured data model for reporting
- Developed interactive dashboards in Power BI
- Data Warehouse: Snowflake
- Query Language: SQL
- Visualization: Power BI
- Data Modeling: Star schema
Conversion is not a checkout problem — it is a decision-making problem at the top of the funnel.
- Review SQL scripts in
/sql - Open Power BI dashboard file (
.pbix) - Explore dashboards and insights