Sample Size Estimator
Estimate sample size per variant from baseline rate, minimum detectable effect, and confidence.
Avoid underpowered tests before they start
Small experiments often end with ambiguous results because they never had enough traffic to detect a realistic change. A sample size estimate helps you decide whether a test is worth running or whether the effect would need to be much larger.
A smaller baseline rate or a smaller effect you want to detect both demand more traffic per variant; running an undersized test is how teams ship changes on noise that later reverse.
Enter a baseline conversion rate, the relative lift you want to detect, and a confidence level. The estimator returns a rough visitors-per-variant target for planning.
Private planning math
The estimate is calculated locally. Analytics assumptions are not uploaded.
Frequently Asked Questions
What does sample size mean?
It is the approximate number of visitors or responses needed per variant to detect the target effect.
Is this exact?
No. It is a simplified planning estimate, not a replacement for a full power analysis.
What is minimum detectable lift?
It is the smallest relative improvement you care enough to detect.
Is this private?
Yes. It runs locally in your browser.
Need sharper pricing, projections, or growth planning? Let us help.