CRO Consulting
About Varify
Contact
Blog
Webinars Live
Success Stories
Card Set
Varify.io
Functions Pricing For agencies Try for free
Get a demo

A/B Test Runtime Calculator

How long until your A/B test reaches significance? Enter your monthly traffic, baseline conversion rate, the effect you expect and the number of variants — and get the required sample size and test duration instantly.

Enter your traffic and test parameters above, then click Calculate to see the required runtime and the days-per-uplift chart.

Plan your test before you start it

Know upfront how much traffic and time a reliable result will take.

Required sample size

See how many visitors and conversions you need per variant to detect the effect you care about.

Time to significance

Get the test duration in days for your expected effect — and how it shifts as the uplift gets smaller or larger.

Plan with confidence

Choose 90%, 95% or 99% confidence and add variants to see how each choice changes the timeline.

Understanding A/B test runtime

How is the test runtime calculated?

The calculator first works out the sample size needed to detect your expected effect at the chosen confidence level and 80% statistical power, using a two-proportion sample-size formula. It multiplies that by the number of variants to get the total visitors needed, then divides by your monthly traffic and converts to days: roughly total visitors ÷ monthly visitors × 30.

Why does a smaller expected effect take longer to test?

Small differences are harder to distinguish from random noise, so they need a larger sample to reach significance. Halving the effect you want to detect roughly quadruples the required sample size — which is why realistic effect estimates matter a lot for planning.

How many visitors do I need?

The calculator shows both the sample size per variant and the total across all variants. The total is what you divide by your traffic to get the runtime, so more variants or lower traffic both push the duration up.

What confidence level and power does this use?

You pick the confidence level (90%, 95% or 99%, two-sided). Statistical power is fixed at 80% — the common default, meaning an 80% chance of detecting a real effect of the size you specified. Higher confidence means a larger sample and a longer test.

Does it account for multiple variants?

Yes — the total sample size scales with the number of variants (including the control). Note that it does not apply a multiple-comparison correction, so with several variants treat the runtime as a lower bound and plan a little extra to stay statistically safe.

One price. Unlimited A/B testing.

Varify.io is your GDPR-compliant, cookie-less alternative to traffic-priced testing tools. Flat rate from €149/month — no matter how much traffic or how many experiments you run.

Start your 30-day free trial

No credit card required · Made & hosted in Germany