How is completion time estimated?

By weighting each question type by typical response time: multiple choice ~12 seconds, open-ended ~45 seconds, scale/Likert/NPS ~8 seconds. These are conservative averages — real times vary by question complexity and respondent engagement.

What's the drop-off model based on?

The model starts at 0% drop-off for the first 4 questions, then adds ~5pp per additional question, plus 20pp for each open-ended question past the first one. It's calibrated against Refiner's 2026 in-app benchmark (27.5% average response, 36.1% mobile in-app) and SurveyMonkey's 2026 finding that ~60% of surveys are completed on mobile.

Are open-ended questions really that costly?

Yes. Open-ended (text) questions take 3-5× longer to answer than multiple choice and trigger significant drop-off, especially on mobile where typing is harder. Limit to one open-ended question per survey if completion matters — and put it last so partial responses are still usable.

Does showing a progress bar help completion?

For short surveys (under 6 questions) progress bars actually reduce completion because they highlight how little progress each click makes early on. For longer surveys (10+) they help — respondents need reassurance that the end is in sight.

Can I reduce drop-off without cutting questions?

Yes. Three high-leverage tactics: (1) put the easiest question first to build commitment, (2) replace open-ended fields with multiple choice plus an optional 'Other' field, (3) use conditional logic to skip irrelevant questions per respondent — apparent length matters more than actual length.

SpaceForms / Tools / Survey Length Calculator

Survey Length Calculator (Free, 2026)

Set your question count and the mix of question types. The calculator estimates expected completion time and drop-off rate using a model calibrated against Refiner's 2026 in-app survey benchmarks (27.5% average response rate, 36.1% mobile in-app).

Total questions: 10

11530

Question type breakdown

Multiple choice

~12 sec each

Scale / NPS / Likert

~8 sec each

Open-ended (text)

~45 sec each

Estimated completion time

3 min

(178 seconds)

Expected drop-off

50%

50% completion rate

Borderline — consider trimming

Time model: 12s (multiple choice) + 8s (scale) + 45s (open-ended) per question. Drop-off model: 0% for first 4 questions, then +5pp per additional question, plus +20pp for each open-ended question past the first one. Capped at 95%. Calibrated against Refiner 2026 in-app benchmarks.

Survey length benchmarks (2026)

Survey type	Ideal length	Typical completion rate
In-app micro-survey (NPS, CSAT)	1-2 questions	36.1% (mobile, Refiner 2026)
Post-purchase / transactional	3-5 questions	25-40%
Customer feedback (email)	6-10 questions	15-25%
Employee engagement	15-25 questions	60-80% (captive audience)
Market research (paid panel)	15-30 questions	70-90% (incentivized)
UX / product research	8-15 questions	20-35%

Sources: Refiner 2026 In-App Survey Benchmark (27.5% average, 36.1% mobile in-app, 26.3% CSAT), SurveyMonkey 2026 State of Surveys (~60% mobile completion), industry compilations.

Why survey length is the #1 lever for completion

Of every controllable variable in survey design — wording, invitation copy, incentives, design polish — survey length has the single largest effect on completion rate. Refiner's 2026 in-app benchmark shows a 36.1% completion rate for one-question micro-surveys, falling roughly 5 percentage points per additional question after question 4 — and crashing further if any of those questions are open-ended.

The mechanism is straightforward: each additional question is an additional opportunity to abandon. With ~60% of surveys now completed on mobile (SurveyMonkey 2026), the friction compounds — small screens, no keyboard, often answering during a 30-second context switch. Designers who optimize for desktop completion consistently overestimate how many questions their mobile audience will tolerate.

This calculator's drop-off model bakes in those realities. It's deliberately pessimistic — better to plan for the floor and be pleasantly surprised than to launch a 15-question survey expecting 50% completion and get 12%.

Six tactics that actually cut drop-off

Lead with the easiest question. A single-click scale question first builds commitment. People rarely abandon after they've already answered something.
One open-ended max — and put it last. Open-ended fields take 3-5× longer and drive 20pp of drop-off each. If you must have one, put it at the end so partial responses are still useful.
Use conditional logic. Show only the questions relevant to each respondent. A 15-question survey that displays 6 per respondent feels like a 6-question survey.
Set length expectations up front. "Takes 90 seconds" beats a vague invitation. Per SurveyMonkey, surveys with stated time take 14% more starts and 8% more completions.
Mobile-first design. Single-column layouts, large tap targets, no horizontal scrolling. ~60% of completions are mobile (SurveyMonkey 2026) — design for that first, desktop second.
Show progress only on long surveys. For surveys under 6 questions, progress bars actually reduce completion. For 10+ questions, they help — respondents need to see the end is near.

When long surveys are okay

Not every survey needs to be 4 questions. Three contexts where longer is fine:

• Captive audiences — employee engagement surveys (15-25 questions) routinely hit 60-80% completion because the audience is paid to participate.
• Incentivized panels — market research panels with cash or points incentives tolerate 20-30 questions at 70-90% completion.
• High-intent audiences — onboarding questionnaires for new SaaS users, where the survey unlocks personalization, sustain longer formats.

For everything else — cold email surveys, in-app prompts, post-purchase feedback — the calculator's guidance holds: ruthlessly cut anything that doesn't change a decision.

Worked example

Scenario

You're designing a post-purchase feedback survey. The draft has 12 questions: 7 multiple choice, 3 scale (1 NPS + 2 CSAT), 2 open-ended ("What did you like?" and "What could be better?"). You're worried 12 is too many. What does the calculator predict?

Calculation

• Time: 7 × 12s + 3 × 8s + 2 × 45s = 84 + 24 + 90 = 198 seconds = 3.3 min
• Base drop-off: (12 − 4) × 5pp = 40pp
• Open-ended penalty: (2 − 1) × 20pp = 20pp
• Total expected drop-off: 60% → completion rate 40%

Recommendation

Cut one open-ended question (saves 20pp). Combine the 3 scales into 1 composite if possible. Result: 10 questions, 2.5 min, ~30% drop-off → ~70% completion. That's a publishable response rate without sacrificing the data you actually need.

Build a shorter survey, free

SpaceForms includes mobile-optimized templates, conditional logic, and live completion-rate analytics. Skip the bloat — AI-suggested question reductions help you cut to the questions that actually change decisions. Unlimited responses on the free tier.

Build a free survey Free survey maker guide

FAQs

How long should a survey be?

In-app and mobile surveys should stay under 2 minutes (4-6 questions). Email surveys with engaged audiences can go to 5 minutes. Above 8-10 minutes, completion crashes below 50% in most contexts.

Are these time estimates accurate?

They're conservative averages. Simple multiple choice ranges 8-18 seconds depending on option count. Open-ended ranges 20-90 seconds depending on prompt complexity. The model uses mid-range defaults — your actual times will vary 20-30%.

Why does the drop-off model only kick in after question 4?

Refiner's 2026 data shows the steepest drop-off happens between the start and question 1 (the decision to engage at all). Once respondents are 3-4 questions in, they've committed — drop-off per additional question becomes more linear.

Does this account for incentives?

No. The model assumes unincentivized surveys (the typical SaaS/product feedback case). Cash, gift card or sweepstakes incentives can reduce drop-off by 30-50% for the same length.

What about matrix / grid questions?

Matrix questions (asking the same scale across 5+ items) read as one question visually but take 4-6× longer to answer. Treat each row as ~6 seconds and consider whether mobile respondents will tolerate the horizontal layout — most won't.