Research · original data
Survey benchmarks from real Claform responses
Completion rate by channel and by survey length, median time to complete and device split, computed from production surveys over the last 90 days. Anonymised aggregates only, published above strict privacy thresholds, and free to cite.
What is a good survey completion rate?
There is no single good completion rate: it depends on the channel, the number of questions and the device. This page measures exactly that from real Claform responses, with completion rate by channel and survey length, median and 75th-percentile time to complete, and the device split, and publishes each figure only once it clears strict privacy thresholds.
Collecting
First publication once a metric crosses the privacy thresholds
Data is being collected now. A figure appears on this page only when it aggregates enough distinct surveys, enough distinct customer accounts and enough responses that nothing about any customer, survey or respondent can be inferred from it. Until then the page documents the method, so you know exactly what you will be reading when the numbers arrive.
Methodology
How these figures are computed
The rules below are applied by the database itself, inside Claform's API. The page only prints what comes back.
01
Window and refresh
A trailing 90-day window on the time each response was started. The aggregates are recomputed once a day and this page refreshes within an hour of a new computation; the exact timestamp is shown at the top and in the citation.
02
Which responses count
Completed responses plus settled partials: responses marked abandoned, or still in progress and untouched for 24 hours. Test responses, responses quarantined by quality checks, surveys owned by Claform staff or company accounts, and surveys with fewer than 10 responses in the window are excluded.
03
Privacy thresholds
A figure is published only when it aggregates at least 20 distinct surveys from at least 10 distinct customer accounts and at least 500 responses. A segment below any one threshold is left out of the page entirely, not shown as zero and not flagged.
04
Rounding
Response counts are rounded to the nearest 100, survey and account counts to the nearest 10, rates to one decimal place and times to the second. Thresholds are checked on the exact counts before rounding.
05
Only aggregates leave the database
Every query returns counts, distinct counts or percentiles computed by the database. No identifier, title, answer text, e-mail address or per-survey row is read by this page or by the public endpoint behind it, and none is stored anywhere outside the production database.
06
Definitions
Completion rate is completed divided by completed plus settled partials. Survey length counts answerable questions only. Time to complete uses completed responses that took between 5 seconds and 2 hours. Device is the browser class the respondent app records at submission; responses without one are left out of the device split.
How to use these benchmarks
Reference points for planning a survey, not grades to pass. Pair them with the free tools and the channel guides.
- NPS calculatorTurn promoter, passive and detractor counts into a Net Promoter Score, then read it against the completion rates above when planning sample size.Open
- Sample size calculatorDecide how many completed responses you need, then divide by the completion rate for your channel and length to know how many people to invite.Open
- Channel guidesHow email, SMS, WhatsApp, phone, QR code and website surveys work on Claform, with the setup each channel needs.Open
- Claform solutionsSurveys, forms and quizzes by industry, team and use case, with the templates and distribution each one calls for.Open
Questions about this data
Where does this data come from?
From real responses submitted to surveys built on Claform, aggregated inside the production database by the API that serves this page. No sample, no panel and no third-party dataset is involved: the figures describe what actually happened on the platform during the window shown at the top of the page.
Can anyone identify a customer, a survey or a respondent from these numbers?
No. Each published figure covers many distinct surveys from many distinct customer accounts and a large number of responses, as set out in the methodology, and every count is rounded. A segment that does not meet all three thresholds is not published at all, so there is no small bucket to reason backwards from, and nothing but aggregates ever leaves the database.
Why is a channel, a length band or a device missing from a table?
Because it has not yet crossed the privacy thresholds in the current window. It is left out rather than shown as zero so that its absence carries no information about how much traffic it has. Segments are added automatically on the next computation once they qualify.
How often is the page updated?
The aggregates are recomputed once a day and the page refreshes within an hour of a new computation. The computation timestamp and the methodology version appear at the top of the page and in the citation block, so two readers quoting different numbers can tell which snapshot each saw.
Can I cite or reuse these figures?
Yes. The figures are published under the Creative Commons Attribution 4.0 licence. Use the suggested citation below, link to this page, and quote the computation date and methodology version alongside any number so readers can see when and how it was measured.
How to cite this page
The suggested form includes the computation date and the methodology version, so a quoted figure can always be traced to the snapshot it came from.
Claform (2026). Survey benchmarks from Claform data: methodology. https://claform.com/research/survey-benchmarks
Licence: Creative Commons Attribution 4.0 International (CC BY 4.0). CC BY 4.0
Measure your own completion rate with the same tools
Every Claform survey reports completion, drop-off by question and time to complete by channel, out of the box. Start free and compare your results with these benchmarks.