Free tool
Survey sample size calculator
Choose a margin of error and a confidence level to see how many answers you need. Or enter the answers you have to see how precise your results are.
- Free
- No sign-up
- Runs in your browser
Everyone you could ask, such as monthly readers.
Leave at 50% if you don't know: it gives the safest number.
The share of people who see the survey and answer.
Answers you need
for ±5% at 95% confidence.
- ±10% margin of error 97 answers
- ±5% margin of error 385 answers
- ±3% margin of error 1,068 answers
Collect the answers on your own pages
Asklet asks a short question on the page your visitors are reading, while the experience is fresh. Answering takes one tap, and results are grouped by page, with no cookies or sign-in.
Start freeFree plan, no card needed.
How to use the sample size calculator
- 01
Choose what to work out
The answers you need for a margin of error, or the margin of error of the answers you already have.
- 02
Set the confidence level and margin
95% confidence and ±5% are the usual starting points. Add your population if it is small, such as the readers of one page.
- 03
Add your response rate
Optional: enter the share of people who answer when they see the survey, and you get how many people need to see it.
What the numbers mean
The calculator uses three settings:
- Margin of error: how far the true answer could be from your result. With ±5%, a result of 60% means the true value is likely between 55% and 65%.
- Confidence level: how sure you can be that the true value falls inside that margin. At 95%, about 19 surveys in 20 would land inside it.
- Expected answer share: the result you expect, such as 70% saying yes. 50% gives the largest, safest number, so keep it when you are unsure.
How the sample size is calculated
For a large population, the answers needed are n = z² × p × (1 - p) ÷ e², where z is 1.96 at 95% confidence, p the expected share and e the margin of error. At 95% confidence and ±5%, that gives 385 answers.
When you enter a population, the calculator applies the finite population correction, n ÷ (1 + (n - 1) ÷ N). It matters when the population is small: for 1,000 people, 278 answers give the same ±5%.
Precision is not the only risk
The formula assumes the people who answer are like the people who do not. If only your happiest or angriest readers answer, more answers won't fix that. Ask at a natural moment, keep the survey short and ask everyone the same way.
Common questions
How many responses do I need for a survey?
For a large audience, about 385 answers give a margin of error of ±5% at 95% confidence, and about 100 give ±10%. For a small, known group, fewer are needed: enter its size in the calculator.
Is 100 survey responses enough?
For many decisions, yes. 100 answers give a margin of error of about ±10 points at 95% confidence, which is enough to spot clear differences but not small ones.
What is a good margin of error for a survey?
±5% at 95% confidence is the common standard. For quick feedback on a page, ±10% is often fine. To compare two pages or track small changes, aim for ±3% or less.
Does population size matter?
Only when it is small. Above a few tens of thousands of people, the answers you need barely change. For a group of a few hundred, the finite population correction lowers the number a lot.
How do I turn the sample size into visitors?
Divide the answers you need by your response rate. With 385 answers and 2% of visitors answering, about 19,250 visitors need to see the survey.
Do you store what I enter?
No. The calculator runs in your browser and nothing is sent to us.