What is this conversion rate & sample size calculator?
A free, browser-based calculator that does two jobs: it computes a simple conversion rate from visitors and conversions, and it tells you how many visitors per variant an A/B test needs to reliably detect a given improvement. Plan your test before you run it so you don't stop too early or chase an effect too small to measure. Everything runs locally in your browser, so your data is never uploaded.
Methodology
The required sample size uses the standard two-proportion formula for a two-sided test. From the baseline rate (p1) and the minimum detectable effect, we derive the target rate p2 = p1 × (1 + MDE), then combine the z-scores for your confidence level and power. Confidence guards against false positives; power against missing a real effect. The result is per variant — a full test needs roughly double.
How to use it
- Enter visitors and conversions to get your current conversion rate.
- Enter the baseline rate and the relative effect you want to detect.
- Pick power and confidence, then read the required visitors per variant.
Common use cases
- Planning how long an A/B test must run before you can trust it.
- Checking whether your traffic can realistically detect a small lift.
- Setting expectations with stakeholders before launching a test.