What to enter
Enter two numbers: how many people you invited to take the survey, and how many completed it. The calculator divides the second by the first and multiplies by 100.
If your survey platform reports partial completions separately from full completions, decide whether you're counting partials as responses before you type a number in. Switching the definition between survey rounds makes period-over-period comparisons meaningless.
Calculate your response rate
A response rate is a data-quality check, not a score. Lower participation raises the risk that the people who answered are not representative of everyone who was asked — read results from small samples with more caution.
The formula
Response rate = (completed responses / people invited) x 100.
Platforms differ on what counts as "invited." If 40 of 500 email invitations bounced, you can divide by 500 (everyone you attempted to reach) or 460 (everyone who actually received the invitation). Either is defensible. Pick one basis and keep using it so your rate is comparable across survey cycles.
What the percentage does and doesn't tell you
A response rate by itself says nothing about who responded. A 70% rate from an eight-person team and a 22% rate from a 4,000-person division aren't measuring the same thing; population size and structure matter as much as the percentage.
We're not publishing a single benchmark figure for a "good" response rate here. Published benchmarks vary by industry, survey length, whether responses are anonymous, and whether the survey is mandatory, and applying an outside number to your own survey without knowing how it was calculated can mislead more than it clarifies.
Nonresponse bias is a separate issue from a low rate. If the people who skip your survey differ systematically from the people who complete it, by tenure, department, shift, or location, even a rate that looks respectable can misrepresent the group you're trying to measure.
What this calculator doesn't do
It performs one division. It doesn't weight responses or adjust for stratified sampling, and it doesn't produce a margin of error. A confidence interval needs your sample variance and population size, neither of which this tool asks for.
What to do after you get a number
If your rate looks low, check who didn't respond before you assume something is wrong with the results. Pull a quick headcount by department or shift and compare it to who actually opened the invitation, if your platform tracks that.
If you're also running the closing recommend-this-employer question from the employee engagement survey template, use the eNPS calculator on those responses once you've confirmed your response rate is high enough to trust the sample.
Log the invited count, the completed count, and which denominator you used somewhere you'll find it next cycle. That's the only way to compare rates survey to survey instead of guessing whether this quarter's number is actually better or worse than last quarter's.