This project analyzes the results of product experiments using statistical methods in Python and visualizes the outcomes in Tableau.
The goal of this project is to evaluate whether changes in product features lead to statistically significant improvements in conversion metrics.
Four key metrics were analyzed:
- add_payment_info / session
- add_shipping_info / session
- begin_checkout / session
- new_accounts / session
- Python
- Pandas
- NumPy
- Statsmodels (Two-proportion Z-test)
- Tableau
- Data preparation and aggregation
- Conversion rate calculation
- Two-proportion Z-test
- Statistical significance evaluation
- Visualization of results in Tableau
| Metric | Conversion Rate Control | Conversion Rate Variant | Metric Change | P-value |
|---|---|---|---|---|
| add_payment_info | 4.38% | 4.93% | +12.5% | 0.00009 |
The results are visualized in Tableau (Link to review) using an interactive dashboard.
- Some metrics showed statistically significant improvements.
- Statistical testing confirmed whether observed differences were meaningful.
Viacheslav Kovalchuk
Aspiring Data Analyst