Survey Length Estimator
Enter how many questions of each kind your survey has — yes/no, single choice, rating grids, rankings, open-ended — plus the words in your consent and instructions, and get an estimated completion time that adds reading time to answering time. Every assumption is shown and adjustable, and the result gives a figure for a typical respondent, a slower one, and a length to announce. Free, no sign-up.
Your questions
How many of each kind does the survey have? Leave blank for none.
Reading and extras
Count the text a respondent reads once, such as the information sheet and consent.
238 is the mean for silent reading of English non-fiction in Brysbaert's (2019) meta-analysis. Survey respondents often skim, so you may set it higher.
Shows how long someone who takes longer than typical would need.
Adjust the timing for each question type
Seconds to decide and respond, not counting reading; and the words a respondent reads per question (the question and its options). These are planning defaults, not measurements: time a few people on a pilot and replace them with what you see.
Estimated completion time
Enter your question counts to see the estimate.
How to Use the Survey Length Estimator
- 1. Count your questions by type. Count a grid as one question per row and a ranking as one per list. Leave a type blank if you have none.
- 2. Add the text respondents read once. Enter the words in your information sheet, consent, instructions and thank-you page.
- 3. Adjust the assumptions if you have data. The defaults are planning figures. If a pilot showed that your single-choice items take 8 seconds, change them.
- 4. Read three numbers. The typical time, a slower respondent's time, and the whole number of minutes to announce, rounded up.
How the Estimate Is Built
Each question costs reading time plus answering time. Reading time is the words a respondent reads (the question and its options) divided by their reading speed. Answering time is a per-type allowance in seconds for deciding and responding. Add the one-off reading of consent and instructions, and you have the typical completion time. A slower respondent's time is that total multiplied by a factor you choose, 1.5 by default.
| Question type | Answering (seconds) | Reading (words) |
|---|---|---|
| Yes / No | 3 | 12 |
| Single choice | 5 | 30 |
| Multiple select | 9 | 35 |
| Single rating or Likert item | 5 | 18 |
| Grid row | 4 | 8 |
| Number or slider | 6 | 12 |
| Ranking (per list) | 25 | 45 |
| Short text | 15 | 15 |
| Open-ended | 60 | 20 |
These figures are the tool's planning defaults, not measurements from a study. Response times depend on how complex an item is and on who is answering: Yan and Tourangeau (2008) found that response times in web surveys differ with question complexity and with respondents' age and experience. A grid row is cheaper than a stand-alone rating because the instructions and scale are read once. A pilot with five or ten people from your target group, timed on their own device, is the best way to replace these defaults.
Reading Speed
The default is 238 words per minute, the mean for silent reading of English non-fiction in Brysbaert's (2019) meta-analysis of 190 studies with 18,573 participants. Fiction came out at 260, and reading aloud at 183. Real survey respondents often skim instructions and read answer options selectively, so a higher speed is reasonable for a motivated panel, and a lower one for respondents reading in a second language or on a small screen.
How Long Should a Survey Be?
There is no universal answer, but there is evidence that length costs responses. In an experiment by Galesic and Bosnjak (2009), people invited to a web survey were told it would take 10, 20 or 30 minutes; the longer the stated length, the fewer started and completed it, and answers to questions later in the questionnaire were faster, shorter and less differentiated. Not every study finds an effect of stated length, but the direction is consistent enough to plan around. In a study of an opt-in online panel in Mexico (Revilla & Ochoa, 2017), respondents' median ideal length was 10 minutes and their median maximum was 20. One panel, one country, one time: use it as context, not a threshold.
A Worked Example
Press Try an example: 43 questions — 3 yes/no, 10 single-choice, 2 multiple-select, 20 grid rows, 3 numeric, 1 ranking, 2 short-text and 2 open-ended — with a 300-word consent and instruction text. Reading the consent takes 76 seconds. The typical estimate is 10:06, a slower respondent needs 15:10, and the length to announce is 11 minutes. The two open-ended questions are 21% of the total, the largest single share, even though they are only 5% of the questions; they are the first place to cut if you need to shorten the survey.
References
- Brysbaert, M. (2019). How many words do we read per minute? A review and meta-analysis of reading rate. Journal of Memory and Language, 109, Article 104047. https://doi.org/10.1016/j.jml.2019.104047
- Galesic, M., & Bosnjak, M. (2009). Effects of questionnaire length on participation and indicators of response quality in a web survey. Public Opinion Quarterly, 73(2), 349–360. https://doi.org/10.1093/poq/nfp031
- Revilla, M., & Ochoa, C. (2017). Ideal and maximum length for a web survey. International Journal of Market Research, 59(5), 557–565. https://doi.org/10.2501/IJMR-2017-039
- Yan, T., & Tourangeau, R. (2008). Fast times and easy questions: The effects of age, experience and question complexity on web survey response times. Applied Cognitive Psychology, 22(1), 51–68. https://doi.org/10.1002/acp.1331
Announce a figure you can keep
Round up and tell respondents the length of a typical completion, not your own fastest run-through. A survey that takes longer than announced raises drop-out and rushed answers late on. If your estimate is for a mobile audience, pilot on phones.
Frequently Asked Questions
- How long does a survey question take to answer?
- It depends on the type: a yes/no may take a few seconds, a single-choice question around five, and an open-ended question a minute or more, plus the time to read it. The tool's defaults reflect those orders of magnitude and are adjustable, but timing a pilot gives better figures for your own audience.
- How many questions can I ask in a 10-minute survey?
- With the tool's defaults, about 40 mostly closed questions fit in 10 minutes, fewer if there are open-ended questions or long stems. Enter your mix and the estimate will tell you; swapping a few open-ended questions for closed ones saves the most time.
- How fast do people read?
- Brysbaert's (2019) meta-analysis found a mean of 238 words per minute for silent reading of English non-fiction. Individuals vary a great deal, and survey respondents often skim, so treat the figure as a starting point and change it for your audience.
- How long should I say the survey will take?
- State the length of a typical completion, rounded up to a whole minute, based on a pilot if you can. Underestimating invites complaints and drop-outs; some researchers also give a range.
- Why is a grid question quicker per row?
- The instructions and the response scale are read once for the whole grid, so each row costs only the item text and the response. Very long grids encourage straight-lining, though, so quick is not always better.
- Does the estimate include page loading and mobile effects?
- No. It covers reading and answering only. Page transitions, slow connections and typing on a phone all add time, particularly for open-ended questions, so pilot on the devices your respondents will use.
- Does this tool store my survey questions?
- It never sees them. You enter counts and word totals, the calculation runs in your browser, and nothing is uploaded.