Survey Sample Size Calculator
Find out how many responses you need for a statistically sound survey — and how many people to invite to get them.
For 95% confidence and a ±5% margin, from a large or unknown population.
| ±1% margin of error | 9,604 |
| ±2% margin of error | 2,401 |
| ±3% margin of error | 1,068 |
| ±5% margin of error | 385 |
| ±10% margin of error | 97 |
How many survey responses do you actually need?
Most teams either over-survey — burning reach and budget — or under-survey and end up with a result they can't defend. Statistical sample size depends on three things: how large your population is, how confident you want to be in the outcome, and how much error you can tolerate.
This survey sample size calculator turns those three inputs into a single number: the count of completed responses that makes your survey statistically sound. Change any input and the required sample size updates instantly.
Confidence level and margin of error, explained
Confidence level is how often a repeated survey would land within your margin — 95% is the near-universal default. Margin of error is the ± band around each result: a ±5% margin on a 60% finding means the true value is very likely between 55% and 65%.
Tightening the margin from ±5% to ±3% roughly doubles the responses you need, so choose the loosest margin your decision can live with. The calculator uses the standard sample-size formula with a worst-case 50% response proportion, then applies a finite-population correction whenever you supply a population size.
From a sample size to a real survey
Sample size is how many completed responses you need; the response-rate field works backward to how many people to invite, since real surveys never see 100% completion. Plan around the invite count, not just the target.
Once you have your number, the survey itself is the easy part. Build it in SimilarForm from a vertical template, or recreate a survey you already run from a URL or PDF, and start collecting responses.
Questions & answers
What confidence level and margin of error should I use?+
95% confidence with a ±5% margin of error is the standard default for most surveys. Tighten the margin to ±3% for higher-stakes decisions — it costs noticeably more responses. ±10% is fine for a quick directional read.
Why does it ask for population size?+
For large or unknown populations the population size barely changes the answer, so you can leave it blank. For small, finite groups — say 200 employees — the finite-population correction meaningfully lowers the number of responses you actually need.
What is the response-rate field for?+
Sample size is how many completed responses you need. Real surveys don't get 100% completion, so the calculator divides by your expected response rate to tell you how many people to invite in the first place.
Is my data sent anywhere?+
No. Every number is computed in your browser with standard statistical formulas — nothing is uploaded or stored.
You've got the number. Now collect the responses.
Build the survey in SimilarForm — recreate one from a URL or PDF, or start from a vertical template. Free to start.