Longer surveys collect more data per respondent but fewer respondents finish them - the trade-off is real and worth thinking about deliberately rather than by accident.
A rough guide
- 1-3 questions: near-zero abandonment risk. Good for quick pulse checks and post-interaction CSAT.
- 4-10 questions: the sweet spot for most feedback surveys - enough room for a few open-ended questions alongside quick ones.
- 10-20 questions: expect meaningfully lower completion. Reserve this length for cases where respondents are highly motivated (paid research participants, required internal surveys).
- 20+ questions: significant drop-off risk for most general audiences. Worth splitting into a shorter core survey plus an optional follow-up for engaged respondents.
Question type matters too
A 10-question survey that's all single-choice takes far less time than a 10-question survey with several open-text fields - respondents feel survey length as much as literal question count.
Tell people how long it'll take
Put an honest time estimate in your survey description. People tolerate a known 3-minute commitment far better than an open-ended one, even if the actual time is similar.