Free CRO Tool
A/B Test Significance Calculator
Find out if your split test results are statistically significant before making decisions. Uses a two-proportion z-test with two-tailed probability.
Enter your test results above to calculate significance.
-%Confidence
-Uplift (B vs A)
-P-value
-Z-score
Results Breakdown
| Metric | Variant A | Variant B | Status |
|---|---|---|---|
| Visitors | - | - | Info |
| Conversions | - | - | Info |
| Conversion rate | - | - | Info |
| Relative uplift (B vs A) | - | - | N/A |
| Statistical confidence (threshold: 95%) | - | - | N/A |
| P-value (two-tailed) | - | - | N/A |
| Z-score | - | - | Info |
Test Quality Checklist
| Check | Value | Status | |
|---|---|---|---|
| · | Confidence above 95% | No data | N/A |
| · | P-value below significance level | No data | N/A |
| · | Sample size >= 100 per variant | No data | N/A |
| ⚠ | At least 7 days of test runtime | Check manually | Warning |
| · | Conversions >= 25 per variant | No data | N/A |
| · | Equal traffic split between variants | No data | N/A |
A/B test quick reference
95%+ confidence - Safe to ship the winner
85-95% confidence - Collect more data first
Below 85% - Result likely due to chance
P-value - Must be below 0.05 at 95% confidence
Runtime - Always run for at least 7 full days
Sample size - Calculate before starting, not after
