Write a survey that respects attention

Overview

Every question you add to a survey costs the participant something, time, focus, and patience. A survey that respects attention asks for only what it actually needs, and makes each question as easy as possible to answer accurately. This matters because a tired or bored participant gives you worse data, not just a worse experience.

Keep it as short as you can

Before adding a question, ask whether you'd genuinely miss it if it were gone. If the answer isn't a clear yes, cut it. Every extra question increases the chance a participant rushes through the rest of the survey, or drops off before finishing.

Put the most important questions first

Participant attention is highest at the start of a survey and drops as it goes on. Ask your most important questions early, while participants are still reading carefully, and save less critical or exploratory questions for later.

Use the right block for the job

Match your question type to what you're actually asking. A Likert Scale works well for measuring agreement or satisfaction consistently. A Star Rating is quick and intuitive for an overall quality judgment. A Text Response should only be used when you genuinely need an open-ended explanation, since it takes far more effort for a participant to answer than selecting an option. Using a heavier block than necessary makes a survey feel longer than it needs to.

Avoid long, dense question text

If a question needs several sentences of setup before you can even ask it, participants are more likely to skim or misread it. Break complex ideas into two shorter questions instead of one long one.

Watch your list lengths

A long list of answer options in a Multiple Choice or Dropdown block takes real effort to read through carefully. If a list is genuinely long, use a Dropdown rather than showing every option at once, and consider whether Bulk Import of Choices can help you manage it without cluttering the question itself.

Don't ask the same thing twice

If two questions are really measuring the same underlying thing, worded slightly differently, participants notice and may disengage, wondering if they missed something the first time. Every question should be pulling its own weight.

Build in a natural end point

Participants can tell when a survey is dragging on longer than expected. If your survey is genuinely long, consider whether it should be split into sections with a clear sense of progress, or whether it's trying to answer too many research questions at once.

What to do next

See Avoid leading questions for how to word individual questions fairly, once you've decided which ones to keep.