Skip to content
← All work

Data analysis · Nov 2025

Customer Shopping Behavior

3,900 retail transactions cleaned in pandas, 10 business questions answered in SQL, and a Power BI dashboard.

Data
3,900 retail transactions
Tools
pandas, PostgreSQL, Power BI
  • Python
  • PostgreSQL
  • Power BI

What I did

  • Cleaned the data in pandas: filled missing review ratings with the median for each product category, and engineered age-group and purchase-frequency features.
  • Loaded the cleaned table into PostgreSQL.
  • Answered 10 business questions in SQL: revenue by segment, how much each group depends on discounts, top products per category, and subscriber versus non-subscriber spend.
  • Built a Power BI dashboard with filters for subscription, gender, category and shipping type.

Recommendation

Stronger subscription incentives and a tighter discount strategy. Only 27% of customers subscribe, and revenue is concentrated in a few categories and age groups, so discounts should target the segments and products that actually drive revenue.

Power BI dashboard: 3.9K customers, $59.76 average purchase, 3.75 average rating, 27% subscribers, and revenue and sales by category and age group
The Power BI dashboard.