Sample Size Calculator

How this works

This estimates the number of completed responses needed to measure a single proportion (like a top-two-box score or repurchase-intent rate) at your chosen confidence and margin of error. The formula is n = z²·p·(1−p) / e², with a finite-population correction applied when you enter a population size.

Rules of thumb for product testing

For a directional read-and-react on a new formula, 50 to 150 skin-matched testers is usually enough to make a launch or reformulate decision. Tighten the margin of error (and raise the count) when a claim or a big spend rides on the number. Remember: who you test matters as much as how many. A hundred verified, skin-matched testers beat a thousand unscreened panelists.