Helpful tools
CRO Tool

A/B Test Sample Size Calculator

Estimate how many visitors each variant needs before your experiment can detect a meaningful conversion lift.

Uses a two-sided two-proportion z-test approximation with equal traffic allocation. Treat the result as a planning estimate, then keep tests running through full business cycles.
Experiment inputs
Planning estimate
%
Current conversion rate for the control.
%
Relative lift you need the test to detect.
Two-sided statistical confidence.
Chance of detecting the lift if it is real.
Total eligible visitors across all variants.
Control plus treatment variants.
Per variant
-
visitors needed
Total sample
-
across all variants
Estimated duration
-
at current traffic

Enter valid inputs

The calculator updates automatically as you change the plan.

When to use it

Use this before launching a landing page, pricing, onboarding, checkout, or signup-flow A/B test.

What the lift means

A 10% lift on a 5% baseline means the variant must reach 5.5%, an absolute gain of 0.5 percentage points.

Planning guardrails

  • Run tests through complete weekday/weekend cycles.
  • Avoid stopping early when results first look good.
  • Use fewer variants when traffic is limited.