Navigation

Case StudiesExperimentationTools

Appearance

A/B Test Significance

Enter your test results below to determine if your uplift is statistically significant and get prescriptive advice on whether to ship or keep testing.

Control Group

Test Variations

Verdict
Not Significant

The observed difference could be due to chance. You need more data or a stronger effect.

The lift, with its range

At 95% confidence. The dot is what you measured; the bar is where the true value probably lies.

Lift over control

Variant B+12.5% (-1.5% to +26.5%)

Conversion rate

Control4.00% (3.6% to 4.4%)
Variant B4.50% (4.1% to 4.9%)

A lift whose bar crosses “No change” could be no change at all. Judge by the lift, not by whether the rates' bars overlap: they can overlap even when the lift is significant.

Control
4.00% CR
Variant B
4.50% CR
+12.50%
Confidence92.0%