How many questions should a survey have?

Because questions are not equal. A one-tap rating takes a few seconds; an open-ended "tell us why" can take a minute or more. Ten multiple-choice questions and ten essay questions are both "10 questions", but one is a two-minute survey and the other is a fifteen-minute one. Respondents experience time and effort, not a count.

Who is answering matters as much. SurveyMonkey found people tolerated longer surveys when they were work- or school-related and less when they were customer-related. An employee who wants their team's workload fixed will give you ten minutes; a shopper who bought socks will give you one.

Two figures appear in many survey-length articles without a dataset behind them. The first says almost half of consumers are willing to answer about 7–10 questions. Jotform attributes it to HubSpot, but HubSpot's article presents 7–10 questions as a practitioner's rule of thumb, not the result of a consumer survey. We could not find an original study, so treat it as opinion.

The second says employee surveys should have 20 to 35 questions. It is vendor guidance; we found no study behind the range. Employees do tolerate longer surveys, per SurveyMonkey, but the exact band is a judgment call.

Also check the date on any figure you use. The SurveyMonkey drop-off data covers 2009–2010, before most surveys were answered on phones. Qualtrics' separate, lower mobile threshold suggests phone respondents tolerate less, so treat the older numbers as generous.

Telling the truth helps; telling a long truth costs you starts. In Galesic and Bosnjak's experiment, respondents told a survey would take 10, 20 or 30 minutes started and finished less often as the stated length rose. The fix is to shorten the survey, not to understate it: an estimate that turns out wrong costs you completions midway and trust next time.

A progress indicator does a similar job during the survey. SurveyMonkey recommends one, and it keeps the promise visible. If your survey is honestly long, say so, explain why it matters, and consider an incentive.

Yes. Beyond losing people, long surveys change how the remaining people answer. SurveyMonkey found respondents on surveys longer than 30 questions spent nearly half as much time per question as on shorter ones. Galesic and Bosnjak found answers to questions placed later were faster, shorter and more uniform — including less variation across grid items, a sign of straight-lining.

Put the questions that matter most early, keep grids short, and do not save your most important open-ended question for the end. Our guide on how to write survey questions covers the order effects in more detail.

Zunoform saves partial responses by default, so you can see the exact question where people stopped rather than guessing. Conditional logic is on every plan, including Free, which is how you keep a long questionnaire short for each person. The one-question-at-a-time layout and a progress bar you can switch on or off help people see how far they have come.

What we do not have yet is our own benchmark of completion rates by survey length; we will publish one only with a stated sample size. The free plan's 500 responses a month suits most feedback surveys, but for very large distributions Google Forms and Tally have no stated response cap.