Survey Length Calculator
A 20-question survey takes one respondent five minutes and another fifteen. The difference is never the question count alone: it is the mix of question types, who is answering, and how many screens stand between them and the submit button. This calculator estimates survey completion time from those inputs, forecasts the completion rate for the resulting length, and writes the invite line you can paste into your survey email.
Questions not counted above are treated as multiple choice (8 seconds each). All per-type times and multipliers are planning estimates, not measured medians.
Recommended maximum for your audience:
Estimated completion rate:
Optimization:
Participant invite line (paste into your email or intro)
Show your work
Quick Answer
A 20-question survey with 10 multiple choice, 5 yes/no, 3 rating-scale and 2 open-ended questions takes about 5 minutes (246 seconds) for a student audience answering online at moderate complexity. The calculator forecasts an 85% completion rate for that length, and students tolerate 15-20 minutes, so the verdict is “Good length for your target audience”. Change any input above and the estimate, forecast and invite line update together.
Key takeaways
- Survey length is not question count alone: the tool weights each question type by its own time (open-ended costs 45 seconds, yes/no costs 3) before applying audience, platform and complexity multipliers.
- The worked example below comes out at 5 minutes and 85% completion; the same survey spread over 3 pages with an intro screen rises to 6 minutes and 75%.
- Completion-rate forecast falls in steps as length grows: 85% at 5 minutes, 75% at 10, 65% at 15, 50% at 20, 35% beyond that, with a 10-point lift if you offer an incentive.
- Every estimate is a planning figure, not a measured median: four of the five per-type seconds sit inside published benchmark ranges (Drag’n Survey, Henricks Consulting, Dynata, AskYourTeam and Pew Research Center), and the yes/no figure is a disclosed estimate, not a benchmark.
- Quick Answer
- How to use the survey length calculator
- Where the time-per-question numbers come from
- Worked example: the 20-question student survey
- Reading the completion-rate forecast
- Audience, platform and complexity multipliers
- What competing calculators miss
- Trimming a survey that comes out too long
- Frequently asked questions
- Add This Calculator to Your Website
How to use the survey length calculator
Enter your survey’s shape and read four answers: how long it takes, the longest it should be for your audience, the share of starters likely to finish, and what to fix. The results stay hidden until you click the button.
- Enter the total question count, then split it across the five question types. Anything you do not itemize is counted at the multiple-choice rate of 8 seconds. The student-satisfaction example ships pre-filled.
- Set the structure: how many pages the survey spans, plus whether it has an intro/consent screen (+20s) or a thank-you screen (+10s).
- Pick the audience, platform, complexity and incentive. A phone interview with older adults is a very different clock from a desktop survey of students, and the multipliers capture that.
- Click “Estimate completion time”. The result shows the rounded-up minutes, the audience’s recommended maximum, the completion-rate forecast, one optimization tip, a ready-to-paste invite line, and a full “show your work” breakdown.
Blank or zero totals are rejected with a message instead of a guess, and negative counts are clamped to zero, so a stray minus sign cannot quietly shrink your estimate.
Where the time-per-question numbers come from
Four of the five per-type seconds sit inside published benchmark ranges. The fifth, yes/no, has no per-type published benchmark anywhere, so it stays a disclosed estimate:
| Question type | Tool estimate | Benchmark or basis |
|---|---|---|
| Yes / No | 3s | Disclosed estimate: no publisher gives a per-type figure for binary questions. 3s sits well below Pew Research Center’s roughly 10 seconds for an ordinary stand-alone question, and in line with practitioner guidance that a simple binary question takes under three seconds. AskYourTeam’s 8.5 seconds for all other questions was deliberately not applied here, because it would erase the binary-versus-multiple-choice difference every source agrees exists. |
| Rating scales | 6s | Drag’n Survey publishes 5 to 7 seconds for rating scales; 6s sits inside that published range. |
| Multiple choice | 8s | Drag’n Survey publishes 7 to 10 seconds for multiple-choice questions and AskYourTeam uses 8.5 seconds for non-text questions; 8s sits inside that published range. |
| Matrix / grid | 20s | Dynata’s rule is 3-4 questions per minute with each grid row counted as an individual question, which is 15 to 20 seconds per row; 20s applies that per-row rate once per grid, so it is the conservative single-line estimate for a small grid. For large grids, count each row as its own question instead. |
| Open-ended | 45s | Henricks Consulting’s published timing data puts qualitative answers at 45 seconds to 1 minute; 45s sits at the bottom of that published range. AskYourTeam budgets 30 seconds per free-text question, and Pew notes that challenging open-ended questions can take respondents minutes, so 45s still understates long-form answers. |
The honest caveat: these are planning constants, not medians measured from your audience, and yes/no at 3s is a disclosed estimate rather than a benchmark figure. Your own pilot timings always beat a generic table, which is why the tool keeps every constant visible in the “show your work” panel instead of hiding them.
Worked example: the 20-question student survey
A customer-satisfaction survey has 20 questions, all on one page with no intro or outro screen: 10 multiple choice, 5 yes/no, 3 rating scales and 2 open-ended. The audience is students, complexity is moderate, the platform is online, and there is no incentive. Here is exactly how the tool reaches 5 minutes:
Now change only the structure: spread the same survey over 3 pages and add an intro screen. Fixed overhead becomes (3 − 1) × 20 + 20 = 60 seconds. Total = 245.63 + 60 = 305.63 seconds, which rounds up to 6 minutes and drops the forecast to 75%. Page turns and screens cost real minutes, which is why single-page short surveys consistently outperform multi-page ones of equal question count. You can reproduce both runs by editing the page and intro fields above.
Reading the completion-rate forecast
The forecast is a stepped heuristic, not a prediction of your survey: 5 minutes or less forecasts 85%, up to 10 minutes 75%, up to 15 minutes 65%, up to 20 minutes 50%, and anything longer 35%. Offering an incentive adds 10 points, capped at 95%.
| Estimated length | Forecast | What research says at that length |
|---|---|---|
| Up to 5 min | 85% | Safely inside every benchmark; Zoho’s analysis puts the abandonment spike at 7-8 minutes. |
| 6-10 min | 75% | Still under the 10-minute engagement ceiling SurveyMonkey recommends; abandonment starts climbing past 7-8 minutes. |
| 11-15 min | 65% | Pew caps its own panel surveys at 15 minutes based on prior research; Revilla and Ochoa found respondents’ stated maximum at a median of 20 minutes. |
| 16-20 min | 50% | Respondents start satisficing: SurveyMonkey measured per-question time nearly halving on surveys above 30 questions, degrading data quality. |
| Over 20 min | 35% | Kantar found surveys over 25 minutes lose more than three times as many respondents as ones under 5 minutes. Split it or pay for the length with incentives. |
Pair the forecast with the audience’s recommended maximum, shown beside every result: mobile users get no more than 5-8 minutes, older adults 8-10, professionals 10-12, the general public 10-15, students 15-20, and subject experts up to 20-25. If the estimate sits above that band, the optimization tip says so plainly.
Audience, platform and complexity multipliers
After the question seconds are summed, three multipliers and two fixed adjustments shape the total:
| Setting | Options and multipliers |
|---|---|
| Question complexity | Simple 0.9 × · Moderate 1.1 × · Complex 1.3 × · Very complex 1.6 × |
| Target audience | Students 1.0 × · Experts 0.8 × · Professionals 0.9 × · General public 1.2 × · Mobile users 1.3 × · Older adults 1.5 × |
| Platform | Online desktop 1.0 × · In-person 1.1 × · Mobile app 1.2 × · Paper 1.3 × · Phone interview 1.4 × |
| Fixed adjustments | Reading and transitions +10% on the question subtotal; each page after the first +20s; intro screen +20s; outro screen +10s. The final minute figure always rounds up. |
Watch the extremes: an elderly audience on a phone interview multiplies question time by 1.5 × 1.4 = 2.1 before the reading adjustment, which is why the same questionnaire can be a 5-minute desktop survey and an 11-minute phone one. If you run a completion rate calculator on your fielded results later, compare its actual figure against this forecast to calibrate future estimates.
What competing calculators miss
Survey-length tools fall into a few camps, and each leaves something on the table:
- Single-average tools (SavvyCalculator): one “average seconds per question” field plus intro/outro and a complexity picker. Honest and simple, but a survey of 20 yes/no questions and a survey of 20 open-ended questions get the same answer unless you hand-tune the average yourself. The per-type split here does that tuning for you.
- Thin formula pages (CalculatorDoc): a two-input formula page with a made-up-looking equation and no breakdown of where the seconds go. No audience or platform adjustment, no completion forecast, no worked numbers you can check.
- Platform help docs (AskYourTeam, Workday Peakon): genuinely useful constants (30s free-text / 8.5s others; 15s per question flat), but they live inside one vendor’s help centre and only cover that vendor’s surveys. They are better used as calibration data, which is what the timings table above does.
- Research write-ups (Pew, SurveyMonkey, Zoho, Dynata): the best evidence on the planet about how long surveys take and when respondents quit, but none of them is a calculator. You read the article, then do the arithmetic by hand.
- The gap this page fills: per-type timing, audience/platform/complexity multipliers, page-and-screen overhead, a completion-rate forecast with an incentive adjustment, audience-specific maximums, an optimization tip, and a paste-ready invite line, with every constant shown in the “show your work” panel.
Trimming a survey that comes out too long
When the estimate lands above your audience’s maximum, the fixes below are ordered by how much time they buy:
- Cut open-ended questions first. At 45 seconds each, two open-ended questions cost as much as eleven multiple-choice ones. Keep at most 2-3, or move them to a follow-up interview. If open-ended questions pass 30% of the total, the tool flags it explicitly.
- Collapse pages. Every page after the first costs 20 seconds of transitions; intro and outro screens add 30 more. A one-page survey with no screens is the fastest structure for the same questions.
- Convert grids to single questions. Matrix questions are slow on mobile and expensive everywhere. If a grid has more than a few rows, Dynata’s guidance is to count each row as its own question, which usually argues for splitting it up or dropping rows.
- Front-load and prune. Put the must-have questions first so partial responses still carry signal, then delete every question that does not drive a decision. Qualaroo’s field guidance puts the in-app ceiling at 2-5 questions and the email ceiling at 7-10 before abandonment climbs sharply.
- Add a progress bar and save-and-resume once you cross 10 minutes, which is exactly what the tool recommends at that length, and consider an incentive: the forecast adds 10 points for one.
- Price the fielding. Longer surveys cost more per complete; run the numbers through a cost per survey calculator before you commit to a 20-minute instrument when a 10-minute one would do.
Frequently asked questions
Each question type is multiplied by its built-in time: multiple choice 8 seconds, yes/no 3, rating scales 6, open-ended 45, matrix/grid 20. Questions you do not itemize count at 8 seconds. The subtotal is multiplied by your complexity, audience and platform factors, increased 10% for reading and transitions, then page and screen overhead is added (20 seconds per extra page, 20 for an intro screen, 10 for an outro). The final figure rounds up to whole minutes.
Typing an original answer is cognitively and physically slower than tapping a choice. The tool budgets 45 seconds per open-ended question, and the research behind that is stark: Pew Research Center notes that challenging open-ended questions can take respondents minutes to answer while an ordinary stand-alone question takes about 10 seconds, and AskYourTeam budgets 30 seconds per free-text question versus 8.5 seconds for everything else. Henricks Consulting’s published timing data puts qualitative answers at 45 seconds to 1 minute, which is the range the tool’s 45 seconds sits in. Two open-ended questions cost roughly as much as eleven multiple-choice ones, which is why the tool warns you when they pass 30% of the survey.
Under 10 minutes for high engagement is the most-cited rule, with the sweet spot at 7-8 minutes before abandonment starts climbing. Mobile surveys should stay under 7 minutes and under 10 questions. The tool pairs this with audience-specific maximums: mobile users 5-8 minutes, older adults 8-10, professionals 10-12, the general public 10-15, students 15-20, and subject experts up to 20-25. Revilla and Ochoa’s web-survey research found respondents’ ideal length at a median of 10 minutes and their stated maximum at 20.
It is a stepped heuristic, not a prediction of your survey: 85% at 5 minutes or less, 75% up to 10, 65% up to 15, 50% up to 20, and 35% beyond, with a 10-point lift for an incentive capped at 95%. Treat it as a planning band. Your real completion rate depends on topic relevance, panel quality, device mix and how the invitation is written, so compare the forecast against your fielded results and calibrate future estimates from your own data.
For small grids, count the grid once at 20 seconds in the matrix field. For large grids, count each row as its own question instead, following Dynata’s rule that each grid row is an individual question estimated at 3-4 questions per minute. A 10-row grid counted once would estimate 20 seconds; counted per row it estimates roughly 2.5-3 minutes, which is far closer to reality.
Because the same questionnaire runs on very different clocks. An older adult answering a phone interview multiplies question time by 1.5 for audience and 1.4 for platform, more than doubling it before the reading adjustment. A student on desktop is the baseline at 1.0 for both. Ignoring who answers and how is the most common reason length estimates miss.
The tool writes an exact figure, rounded up: “Please allow about 5 minutes to complete this survey.” Rounding up is deliberate, since under-promising time is the fastest way to lose trust and completions. If your estimate sits near a forecast boundary, stating the rounded-up number keeps expectations honest.
Yes. It is free, runs fully in your browser, and asks for no account, so your survey structure never leaves your device. “Copy my results” grabs the rounded-up time estimate, the completion-rate forecast and the question-by-question breakdown in one go; “Share results” sends the same summary to your team or into the invite text.
Add This Calculator to Your Website
Embed this calculator on your own site with the code below. Pick a size to preview it live, then copy the code.
Powered by Calculator Zilo
References
- Pew Research Center. How We Keep Our Online Surveys From Running Too Long (15-minute cap on panel surveys, question point system, open-ended questions taking minutes versus about 10 seconds for an ordinary question). Accessed October 24, 2026.
- Zoho Survey. How Long Should a Survey Be? (abandonment spikes at 7-8 minutes, 5-12 minute sweet spot, Kantar data on 25-minute surveys losing three times the respondents). Accessed October 24, 2026.
- Dynata. How Do I Calculate My Survey Duration? (3-4 questions per minute estimation method, each grid row counted as an individual question). Accessed October 24, 2026.
- AskYourTeam. How Do We Calculate Time to Complete a Survey? (30 seconds per free-text question, 8.5 seconds for all other questions). Accessed October 24, 2026.
- Workday. Concept: Approximate Survey Completion Time (average of approximately 15 seconds per question across the Peakon customer base, rounded to the nearest minute). Accessed October 24, 2026.
- Henricks Consulting. How Many Items Should I Choose for My Survey? (fixed-response and multiple-choice questions about 12 seconds on average, commonly as little as half that; qualitative and open-ended answers most commonly 45 seconds to 1 minute). Accessed October 24, 2026.
- Drag’n Survey. What Is the Maximum Number of Questions to Ask in a Survey? (multiple-choice questions 7 to 10 seconds, rating scales 5 to 7 seconds, open-ended questions 20 to 30 seconds). Accessed October 24, 2026.
- Melanie Revilla and Carlos Ochoa. Forum: Ideal and Maximum Length for a Web Survey, International Journal of Market Research (2017) (respondents’ ideal length at a median of 10 minutes, stated maximum at 20 minutes). Accessed October 24, 2026.
- HubSpot. How Long Should a Survey Be? The Ideal Survey Length (Qualtrics data: surveys longer than 12 minutes, and 9 minutes on mobile, see substantial break-off; keep surveys under 15 minutes). Accessed October 24, 2026.