A/B Test Sample Size Calculator
Plan an A/B test before you launch it. Enter your baseline conversion rate, the smallest lift you want to be able to detect, plus your statistical power and significance level. The calculator uses the standard two-proportion formula to tell you how many visitors each variant needs and how many in total.
How to use the A/B test sample size calculator
- Enter your current baseline conversion rate as a percentage.
- Enter the minimum detectable effect: the absolute lift in points you want to catch.
- Set the statistical power and significance level, then read the sample size per variant.
Examples
Sizing a 10% to 12% test
Baseline 10%, detect a 2 point lift, 80% power, 5% significance, two-tailed.
About 3,841 visitors per variant, so roughly 7,682 in total.
A bigger lift needs less traffic
Baseline 10%, detect a 5 point lift, 80% power, 5% significance, two-tailed.
Far fewer visitors per variant, because a larger gap is easier to detect.
Frequently asked questions
What is the minimum detectable effect?
The minimum detectable effect (MDE) is the smallest change in conversion rate you want your test to be able to spot. This tool uses an absolute MDE in points: a 2 point MDE on a 10 percent baseline means you are sizing the test to detect a move from 10 percent to 12 percent. Smaller effects need much larger samples.
What is statistical power and why does it matter?
Statistical power is the probability that your test correctly detects a real difference of the size you specified, rather than missing it. The convention is 80 percent power, meaning a 20 percent chance of a false negative. Raising power to 90 percent makes the test more sensitive but requires a larger sample, which is exactly the trade-off this calculator quantifies.
What is the significance level?
The significance level, or alpha, is the false positive rate you are willing to accept: the chance of calling a result a winner when the variants are actually equal. The common default is 5 percent, which pairs with 95 percent confidence. A stricter 1 percent significance level reduces false positives but pushes the required sample size up.
Should I use a one-tailed or two-tailed test?
A two-tailed test checks whether the variation differs from the control in either direction and is the safer default. A one-tailed test only asks whether the variation is better, which lowers the required sample size but should be reserved for cases where a worse result would be treated the same as no change. This tool supports both.
How is the sample size calculated?
It uses the standard normal-approximation formula for comparing two proportions. From your baseline and target rates it derives a pooled variance, combines the z-values for your significance level and power, and divides by the squared absolute lift. The result is rounded up to whole visitors per variant.
Does my data leave my browser?
No. Every calculation runs entirely in your browser. The rates and settings you enter are never uploaded or stored on a server.
Related tools
A/B Test Significance Calculator
Check if your A/B test result is statistically significant. Enter visitors and conversions for each variant to get the p-value, z-score and verdict.
Conversion Rate Calculator
Calculate your conversion rate from visitors and conversions. Add spend and revenue to see cost per conversion and revenue per conversion instantly.
CAC Calculator
Calculate customer acquisition cost from your marketing and sales spend and new customers. Add lifetime value to see your LTV:CAC ratio instantly.
Break-even Calculator
Find your break-even point fast. Enter fixed costs, price and variable cost per unit to see the units and revenue needed to cover your costs.
Break-even Units Calculator
Find how many units you must sell to break even, hit a target profit, and see your margin of safety from fixed costs, price and variable cost.
Churn Rate Calculator
Calculate customer churn and retention from customers lost. Add MRR to get gross and net revenue churn, including net negative churn, in your browser.