Average completion time

On average, respondents spend 4 minutes 20 seconds completing a survey — from the first question to submission. Average completion time is the average duration of a session from the start to the end of a survey. The metric helps gauge how "long" a survey feels to the respondent, spot overly fast sessions (possible "speeding" or click-through) or overly long ones, and plan the burden placed on respondents. Together with the completion rate and the dropout rate, completion time is one of the basic survey metrics.

Average time is calculated only over completed responses: incomplete sessions can be cut off for various reasons and would distort the figure. In SurveyNinja, completion time is visible in reports and can be used to decide whether to shorten the survey or to check response quality.

It is useful to state the metric in the respondent invitation ("the survey will take about 5 minutes") and in the report methodology — that way the client and colleagues understand the real burden on participants.

What average completion time means in simple terms

Average completion time is the average duration from the moment a survey begins (opening the first page or answering the first question) to the moment the answers are submitted (survey completion), calculated across all completed sessions. It is measured in minutes and seconds (or just seconds). It is used to assess the "length" of a survey from the respondent's point of view, to detect abnormally short sessions (risk of inattentive completion) or abnormally long ones, and to compare survey versions.

Put simply: average completion time is "how long, on average, people spend on a survey from start to finish." If the average is 5 minutes, the survey is perceived as roughly a five-minute one.

Why measure average completion time

Assessing survey length. It shows the real duration for the respondent. If the instructions say "3 minutes" but the average time is 12 minutes, the survey's length was underestimated, which can increase fatigue and dropouts.

Quality control. An average time that is too short (for example, 30 seconds for 20 questions) may indicate click-through without reading. A time that is too long suggests that respondents are getting stuck or distracted.

Comparing versions. After shortening a survey or adding logic jumps, the average completion time should decrease; comparing before and after helps assess the effect.

Planning. Knowing the average time helps plan surveys in field conditions (for example, at an event) and make realistic promises to respondents ("the survey will take about N minutes").

How it is calculated and what affects it

Average time is usually calculated as the arithmetic mean over all completed sessions: the sum of session durations divided by the number of completed ones. Only those who reached form submission are counted. The result is affected by survey length, the number of questions, logic jumps (some respondents see fewer questions), the complexity of the wording, and whether the survey is taken on mobile or desktop.

How to interpret it

Time that is too short. If the average time is noticeably below what you expected (for example, one question per minute with complex scales), some respondents may have answered inattentively. It makes sense to look at the distribution: how many "outliers" with very short times are there.

Time that is too long. It may mean that respondents are reading and thinking (good) or that the survey is overloaded or inconvenient (bad). A breakdown by question or screen helps you understand where the time grows.

Spread. A large spread in time (some finish in 1 minute, others in 15) may indicate different path lengths due to logic jumps or different behavior (attentive vs. click-through respondents). It is useful to look at the median and percentiles, not just the average.

Common mistakes

Including incomplete sessions. If you calculate time over all sessions, including abandoned ones, the average gets distorted: someone left after 10 seconds, someone after 8 minutes. The standard is to count only completed ones.

Relying on the average alone. A single average value does not show the spread. When needed, supplement it with the median or percentiles.

Ignoring it during the pilot. In a pilot survey, average completion time is one of the signals: if the survey turns out too long, it can be shortened before the main data collection.

Not accounting for the device. On mobile, surveys often take longer because of typing and scrolling. If a significant share of respondents answer from a phone, breaking down the average time by device type can be more useful than a single overall number.

How it looks in SurveyNinja

In SurveyNinja, completion time is recorded for each response; in reports you can see the average time over completed sessions and use it to assess survey length and data quality. If needed, you can export the data and calculate the median or the time distribution in third-party tools. Tracking average time helps you make realistic promises to respondents ("the survey will take about N minutes") and reduce dropouts through an honest estimate of length.

If in one survey some respondents follow a short path (thanks to logic jumps) while others follow the full one, the average time will be an average of the two groups. In such cases it is useful to look at a breakdown by segment or by the number of questions answered, so as not to interpret a single number as "the time for everyone."

Example of use

A survey of 15 questions (mostly scales and single choice). Across 200 completed responses, the average completion time is 4 min 10 s. The invitation stated "about 5 minutes" — the promise was kept. When checking the distribution, it turned out that 15 respondents completed the survey in under 1.5 minutes; their answers can, if necessary, be excluded from the substantive analysis as possible click-through. After shortening the survey to 10 questions, the average time dropped to 2 min 50 s — the metric confirmed the effect of the reduction.

Average completion time makes sense to view together with the completion rate and the dropout rate: if a survey takes 15 minutes on average and the dropout rate is high, the reason may be overload. Shortening a survey usually reduces both the average time and the dropout.

Practical recommendations

State it in the report. In the methodology you can write: "The average survey completion time was N minutes (over completed responses)."

Compare it with the stated length. If the invitation says "5 minutes" but the average time is 12, adjust the promise or shorten the survey.

Use it in the pilot. From the pilot, estimate the average time and, if necessary, shorten or simplify the survey before the main collection.

Supplement it with the median when the spread is large. If the spread in time is large (for example, from 1 to 25 minutes), the median reflects the "typical" session better than the average. When exporting from SurveyNinja, you can calculate the median in Excel or another tool.

Relationship with other metrics

Average completion time complements the completion rate and the dropout rate: if a survey takes a very long time on average, a high dropout rate may be a consequence of fatigue or overload. State the average time in the methodology together with the number of questions and the stated length — that way the reader can assess the burden on respondents and the reproducibility of the conditions. In SurveyNinja, the average time over completed responses is shown in reports without any additional calculations.

For long surveys with branching, it is useful to record not only the overall average time but also the average by the number of questions answered or by segment (for example, B2B vs. B2C), so that the interpretation is more accurate.

Bottom line: average completion time is a simple, intuitive metric of survey length; state it in the methodology and in the respondent invitation.

Average completion time is the average duration from the start to the end of a survey across all completed responses. It helps assess the real length of a survey, spot overly fast or slow sessions, and plan the burden on respondents. In SurveyNinja, the metric is available in response reports.