How to Design a Short Survey People Will Finish

Think about the last survey you abandoned. Chances are it opened with a wall of text, asked you the same thing three ways, and had a progress bar that barely moved. If you want to design a short survey that people actually finish, you need to treat respondents’ time as a limited resource and spend it carefully. This guide sets out a practical process: decide what matters, cut ruthlessly, order the questions kindly, and test on a real phone.

Why short surveys work better

Every question asks something of the respondent. Attention fades, thumbs get tired, and the temptation to close the tab grows. Long surveys tend to lose people part way through, and the people who stay are not always typical. Those who do finish may rush, picking the same option repeatedly, which damages your data without any obvious warning sign.

A short survey is not a lazy survey. It is a focused one, built around a decision you need to make. Fewer questions mean cleaner analysis, faster reporting and respondents who are happier to help next time.

Step 1: Write the decision first

Complete this sentence: “I will use these answers to decide whether to ______.” Every question must then pass a simple test: if the answer would not change that decision, delete the question.

Examples of passing questions for a decision about a new class timetable:

• Which evening works best for you?
• How would you usually travel to the class?

Examples of failing questions:

• How did you hear about us? (interesting, but not for this decision)
• What is your job title? (collected out of habit)

Keep a separate “nice to know” list for a future survey. Resist the urge to smuggle those questions in.

Step 2: Count what you can afford

There is no magic number, but a useful discipline is to set a time budget before you write anything. Many general-audience surveys feel comfortable at a few minutes. A rough rule of thumb is that a simple tick-box question takes seconds, while an open-text question can take a minute or more.

Work out a budget like this:

1. Decide the total time you are willing to ask for.
2. Allow most of that for your core questions.
3. Reserve a small slice for any demographics or screening.
4. Count open-text questions at a much higher cost than closed ones.

If your draft exceeds the budget, cut. Then time yourself, and time a colleague who has not seen it before. People who wrote the questions always complete them faster than newcomers.

Step 3: Choose question types that are quick

Closed questions are faster to answer and to analyse.

• Single choice for categories and simple facts.
• Short scales (for example five points, with labelled ends) for agreement or satisfaction.
• Yes/No for plain facts, though it hides nuance.
• One optional open question at the end, such as “Is there anything else we should know?”

Avoid large grids where respondents rate ten items on the same scale. They look efficient to the author but are tedious on a phone and encourage straight-lining.

Make questions optional unless an answer is essential. Forcing responses leads to random clicking or abandonment.

Step 4: Put the questions in a kind order

Order affects both completion and honesty.

1. Start with an easy, relevant question that signals the survey is quick. An engaging opener keeps people moving.
2. Group related questions together so the respondent is not mentally jumping around.
3. Place sensitive questions later, after trust has been built, and offer “prefer not to say”.
4. Put demographics at the end unless you need them to route people.
5. Avoid priming. A question about a specific problem early on can colour later answers about overall satisfaction. Ask general questions before specific ones.

Use skip logic so people only see questions that apply to them. If someone has never used a service, do not ask them to rate it. Test every branch to ensure nobody hits a dead end.

Step 5: Write an opening that earns the click

Your introduction should be three or four lines at most. Include:

• Who you are and why you are asking.
• An honest estimate of how long it takes.
• What will happen with the answers.
• Whether responses are anonymous, and how you will treat personal data.

Do not promise “two minutes” if testing shows five. Misleading estimates breed resentment and reduce trust. If you collect personal information, follow data protection rules and keep the privacy statement brief and clear.

Step 6: Design for phones first

A large share of people will open your survey on a phone, often while doing something else.

• Use one question per screen if your tool allows, or a short scrolling page with big tap targets.
• Avoid dropdown lists with many items, which are fiddly on a touchscreen.
• Keep text short, and avoid long matrices.
• Make sure the page loads quickly and does not need zooming.
• Check it in both light and dark mode, and with larger text sizes, for accessibility.

A progress indicator helps only if it is honest. A bar that jumps from 20% to 90% after one question feels like a trick.

Step 7: Test, then trim again

Before launching, run a pilot with a handful of people who resemble your audience. Watch where they hesitate. Ask them:

• Was any question unclear or repetitive?
• Did the length match what I said at the start?
• Was there a point where you felt like stopping?

Most pilots reveal at least one question that can go. Remove it. Then check that results are recorded correctly, including on mobile.

Step 8: Make the ask easier to say yes to

Design is only part of completion. A few other choices help:

• Send from a recognisable name rather than a generic address.
• Pick a sensible time. Mid-morning on a weekday suits many work audiences, but test what suits yours.
• Send one polite reminder to non-responders after a few days.
• Tell people when you will close the survey.
• Share what you did with previous results.

If you offer an incentive, keep it modest and transparent, and be clear about terms.

A quick checklist before you press publish

• [ ] There is a written decision the survey informs.
• [ ] Every question would change that decision.
• [ ] Time estimate tested with newcomers.
• [ ] Most questions are closed, with at most one or two open.
• [ ] Skip logic tested for every branch.
• [ ] Opening text is short and honest.
• [ ] Works comfortably on a phone.
• [ ] Only necessary personal data collected.

How to Write Unbiased Survey Questions
Online Surveys Explained: How They Work and How to Run One

Frequently asked questions

How many questions should a short survey have?

It depends on question types and audience, not a fixed number. Set a time budget, then fit questions into it. Ten simple closed questions can be quicker than three open-ended ones.

Should I show a progress bar?

Only if it is accurate. A bar that moves evenly reassures people, while one that stalls or jumps can cause drop-off.

Is it better to use one question per page?

On phones it often feels quicker, but it can add taps. Test both layouts with a few people and see which they prefer.

What if I genuinely need a long survey?

Split it into stages, offer it only to those who agree to take part, or consider interviews for deeper topics. Be open about the length and consider acknowledging people’s time.

Conclusion

To design a short survey people will finish, start from the decision, cut every question that does not serve it, favour quick closed questions, order them kindly and test on a phone. Be honest about the time it takes, and respondents will repay you with complete, thoughtful answers.