Significance Calculator
Is your test result actually significant?
A complete significance calculator for the three metrics that matter — conversion rate, average order value, and revenue per visitor.
Control (A)
— enter visitors & conversions
$
$
— enter orders & revenue
$
$
— enter visitors & revenue
Variant (B)
— enter visitors & conversions
$
$
— enter orders & revenue
$
$
— enter visitors & revenue
Confidence threshold
Hypothesis
Enter your test data above
Pick a metric, fill in both variants, and your significance verdict appears here.
Confidence
—
—
Relative uplift
—
—
Z-score
—
—
P-value
—
—
Confidence level—
95%
Sample distribution
A
—
B
—
How to read these results
Conversion Rate
Pooled-proportion z-test. Use this when your KPI is a binary
outcome — purchased / didn't, signed up / didn't.
AOV (Average Order Value)
Welch's t-test on order values. Sample size = number of
orders. Tests whether the average basket size moved.
RPU (Revenue Per User)
Welch's t-test on revenue per visitor (zeros included).
Combines CR and AOV — the metric that ultimately moves the topline.
Statistical significance
The probability the result is real and not random chance. 95% is industry
standard — only a 5% chance the difference is a fluke.
Relative uplift
How much better (or worse) variant performs vs. control. This is the number
you report to stakeholders.
Two-tailed vs one-tailed
Two-tailed tests for any difference. One-tailed tests only for improvement.
Default to two-tailed.
Need someone to build and QA the test?
BrillMark has run 15,000+ A/B tests on Optimizely, VWO, Convert, AB Tasty, and more.