Completion Rate and Average Time: What Your Survey Management Page Shows You¶

Your Survey Management page tracks more than how many people answered - it also tracks how many finished versus started, and how long a completed response actually took. Both numbers are useful, and both are easy to misread if you don't know what's feeding into them.

What Feeds Completion Time¶

Average completion time is calculated from when a respondent actually started the survey to when they actually finished it, using timestamps recorded independently of the respondent's own device clock - which is what keeps the number accurate even when a respondent's browser and the server disagree about the exact time. A respondent who starts, closes their browser, and comes back later to finish (see How Respondents Resume an In-Progress Survey) still gets an accurate total duration, including the gap while they were away - not an artificially short number reflecting only the time after they returned.

Why a Long Average Doesn't Always Mean a Long Survey¶

Because the timer runs across the full span from start to finish, a handful of respondents who started a survey, got distracted for an hour, and came back later can pull your average completion time up noticeably - even though the survey itself takes most people only a few minutes to actually complete. Looking at the spread of individual completion times, not just the average, gives a more accurate sense of typical experience.

Completion Rate: Finished vs. Started¶

Completion rate reflects how many respondents who started your survey actually finished it, distinct from a screened-out respondent (see Screening and Disqualification: What a Respondent Sees), who was disqualified rather than simply abandoning the survey partway through. A low completion rate is worth investigating - it can point to a survey that's too long, a confusing question, or a technical issue, rather than simply respondent disinterest.

Using These Numbers Together¶

A short average completion time paired with a low completion rate is a different signal than a long average time paired with a high completion rate - the first suggests people are giving up quickly (possibly right at the start), while the second suggests people are willing to invest real time and mostly seeing it through. Reading the two together tells you more than either number alone.

FAQ¶

Does a screened-out respondent count against my completion rate?
Check your Survey Management page for how screened-out responses are categorized separately from abandoned ones - this distinction affects how you should read the completion rate.

Can I see completion time for individual responses, not just the average?
Check your Survey Management page for response-level detail - the average is a summary, but individual response data is typically also available.

Does resuming a survey after a long break skew the average unfairly?
It reflects real elapsed time, including the gap - which is accurate but can make the average look longer than the actual active time respondents spent answering. Consider this when interpreting the number, especially for a survey where mid-survey breaks are common.

Is there a benchmark for what counts as a "good" completion rate?
This varies enormously by survey length, audience, and distribution method - there's no universal benchmark. Comparing your own survey's completion rate against a similar prior survey of yours is usually more useful than an external benchmark.

For a related read on interpreting scale-based results, see Interpreting NPS and Scale Question Results.