A practical guide to the 12 main survey question types, with clear examples and advice on when to use multiple choice, rating scales, NPS, open-text, and more.
Choosing the right question type is half the battle in survey design. The wrong format produces messy data, frustrates respondents, and buries the insight you were after. This guide covers the twelve question types you'll actually use, with concrete examples and guidance on when each one earns its place.
Contents
- Closed vs. open questions
- Multiple choice and checkboxes
- Rating scales and Likert
- NPS and CSAT questions
- Ranking and matrix questions
- Open-text and special formats
- How to choose the right type
- Frequently Asked Questions
Closed vs. open questions
Every question type falls into one of two camps. Closed questions give respondents a fixed set of options and produce clean, countable, comparable data. Open questions let people write freely and produce rich, surprising detail that is harder to analyze at scale. A good survey leans heavily on closed questions and uses open ones sparingly, where the depth is worth the analysis effort.
The trade-off is effort versus nuance. Closed questions are fast for respondents and fast to analyze, but they only capture answers you anticipated. Open questions capture the unexpected but cost respondents more effort and cost you analysis time. Keep this tension in mind as you read through the types below — most of them are closed, and that's by design.
Multiple choice and checkboxes
1. Single-select multiple choice. The respondent picks exactly one option from a list. Use it for mutually exclusive answers: "Which plan are you on?" or "How did you hear about us?" Keep the options exhaustive and non-overlapping, and add an "Other" field when you can't list everything.
2. Multiple-select (checkboxes). The respondent picks any number of options. Use it when several answers can be true at once: "Which features do you use?" Be cautious comparing checkbox results across respondents, since people select different numbers of items, which can distort percentages.
3. Dropdown. Functionally a single-select, but collapsed to save space. Reserve it for long, familiar lists like country or state. Avoid dropdowns for short lists — radio buttons are faster because every option is visible — and avoid them on mobile when you can, since scrolling a long dropdown with a thumb is tedious.
4. Yes/No (dichotomous). A two-option question that's fast and unambiguous. Great for screening and qualifying ("Have you purchased in the last 30 days?") and for clear binary facts. Don't use it for attitudes that have shades of gray — "Did you like it? Yes/No" loses far more than a rating scale would capture.
Rating scales and Likert
5. Rating scale. Respondents rate something on a numeric scale, often 1 to 5 or 1 to 10. Use it to measure intensity — satisfaction, quality, likelihood. The key decision is the number of points. Five points is easy and fast; ten points gives finer resolution but adds cognitive load. Whatever you choose, label the endpoints clearly ("1 = very poor, 5 = excellent") and stay consistent across your survey.
6. Likert scale. A specific kind of rating where respondents state agreement with a statement: "The checkout was easy" with options from "Strongly disagree" to "Strongly agree." Likert is excellent for measuring attitudes and is the backbone of many employee engagement surveys. Use an odd number of points if you want to allow a neutral middle, or an even number if you want to force a lean one way or the other.
7. Star rating. A visual rating, usually one to five stars. It's intuitive and familiar from review sites, which makes it a low-effort choice for quick feedback. The downside is that stars carry cultural baggage (people anchor on "five stars = perfect"), so they cluster at the top. Use them when you want speed and recognition more than statistical precision.
NPS and CSAT questions
8. Net Promoter Score (NPS). A single, standardized question: "How likely are you to recommend us to a friend or colleague?" on a 0 to 10 scale. Responses split into promoters (9-10), passives (7-8), and detractors (0-6), and the score is the percentage of promoters minus the percentage of detractors. Its strength is comparability — you can benchmark over time and against others. Learn the full method in our guide to the NPS survey, and always pair the score with a follow-up "why?" question to make it actionable.
9. Customer Satisfaction (CSAT). A short rating of satisfaction with a specific interaction: "How satisfied were you with your support experience?" on a scale like 1-5. CSAT is best measured right after a transaction while it's fresh, which makes it ideal for support tickets, deliveries, and purchases. See how it's structured in our CSAT survey overview. The distinction matters: NPS measures overall loyalty, CSAT measures a single moment.
Ranking and matrix questions
10. Ranking. Respondents order a list of items by preference. Ranking is powerful when you need to know priorities — "Rank these features by importance" — because it forces trade-offs that rating scales don't. The catch is that ranking is high-effort, especially on mobile, so keep the list short (four to six items) and never ask people to rank more than they can hold in their head.
11. Matrix (grid). A table where rows are items and columns are a shared scale, letting you ask the same rating question about many items at once. It's compact and efficient on desktop, which is why it's common in detailed market research surveys. But matrices are a known cause of "straight-lining," where bored respondents pick the same column down the row, and they're painful on phones. Use them sparingly, keep them small, and avoid them entirely on mobile-heavy audiences.
Open-text and special formats
12. Open-text. A free-form box where respondents write whatever they want. Open-text is where you find the things you didn't think to ask about — the actual words customers use, unexpected pain points, ideas you'd never have listed. Its cost is analysis: free text doesn't tally itself, and large volumes require manual coding or text-analysis tools. Use it deliberately, usually as one well-placed follow-up rather than a survey full of empty boxes.
A few special formats round out the toolkit:
- Slider. A draggable scale, useful for continuous values like budget or percentage. Engaging but imprecise — people rarely land on an exact number.
- Date/time picker. For collecting specific dates cleanly, such as a purchase date or appointment.
- File upload. For screenshots, receipts, or photos when context requires it.
- Image choice. Multiple choice where the options are pictures — handy for design, packaging, or brand preference tests.
How to choose the right type
Work backward from the answer you need, not forward from the question you want to ask. Start with the decision the data will inform, then pick the simplest format that captures it. A short decision checklist:
- Do you need to count and compare? Use closed questions — multiple choice, scales, NPS, CSAT.
- Is the answer one option or several? Single-select for mutually exclusive, checkboxes for "select all."
- Are you measuring intensity or attitude? Use a rating or Likert scale, not Yes/No.
- Do you need priorities and trade-offs? Use ranking, kept short.
- Do you want to discover the unexpected? Add one open-text question, sparingly.
Above all, match the format to your respondents' device and patience. The most elegant matrix question is worthless if half your audience is on a phone and abandons it. When in doubt, choose the simpler format — clean data from an easy question beats messy data from a clever one.
Want to put these question types to work? SurveyMaker supports every format above and helps you pick the right one as you build.
Create a survey free or browse templates that already use the best question type for the job.
Frequently Asked Questions
What is the difference between a rating scale and a Likert scale?
A rating scale asks respondents to score something on a number line, like rating quality from 1 to 5. A Likert scale is a specific type of rating where people state how much they agree with a statement, from strongly disagree to strongly agree. Likert is best for measuring attitudes and opinions.
How many points should a rating scale have?
Five points is a reliable default — it's easy to answer and analyze. Use ten points, as NPS does, when you need finer resolution and plan to benchmark over time. Always label the endpoints clearly and keep the same scale length consistent throughout your survey.
When should I use an open-text question?
Use open-text when you want to discover things you didn't think to ask, such as the exact words customers use or unexpected pain points. Use it sparingly, usually as one follow-up after a rating, because free-form answers take real effort to analyze at scale.
Should I avoid matrix questions on mobile?
Generally yes. Matrix or grid questions are compact on desktop but hard to read and tap on phones, and they encourage straight-lining where bored respondents pick the same column repeatedly. For mobile-heavy audiences, break the matrix into separate single questions.