Survey questions: 120 examples and how to write your own

What makes a good survey question

A survey is a machine for turning attention into decisions. Every question you add spends a little of someone's goodwill, so each one has to pay for itself. Four rules cover almost everything that goes wrong.

One idea per question. "How would you rate our pricing and support?" cannot be answered by someone who loves one and hates the other. Split it. If you find yourself writing "and" or "or" in the middle of a question, you probably have two questions.

Neutral wording. The question should not tell the respondent which answer you are hoping for. "How helpful was our new dashboard?" assumes it was helpful. "How would you rate the new dashboard?" does not. The same rule applies to answer options: if the positive choices are warmer and more numerous than the negative ones, the scale itself is leading.

Answerable from memory. People cannot reliably tell you how many times they opened an app last month, or what they would do in a situation they have never been in. Ask about the most recent occasion, the last thirty days, or a concrete past behavior, and you get answers that hold up.

Worth asking. Before a question goes in, name the decision it informs and the person who will act on it. If nobody can, cut it. This single test usually removes a third of a draft survey and raises the completion rate of what is left.

The 9 types of survey question

Nine formats cover virtually every survey you will ever write. The skill is not knowing them, it is choosing the narrowest one that still answers your question. If you want a deeper breakdown of each format and the settings behind it, the survey question types guide goes format by format.

1. Open-ended

A free-text box with no options. Use it when you do not yet know the answer space, when you want the respondent's own language for later copywriting, or as a follow-up to explain a score. It is the only question type that can surprise you.

Example: "What almost stopped you from signing up?"

Mistake to avoid: stacking three open-ended questions in a row. Typing is expensive. One, placed after a closed question that has already warmed the respondent up, gets far better answers than three at the start.

2. Multiple choice (select many)

A list where more than one option can be chosen. Use it for questions where several answers are genuinely true at once, such as which channels someone uses or which features they rely on.

Example: "Which of these do you use at least once a week? Select all that apply."

Mistake to avoid: forgetting "None of these" and "Other". Without them, people pick the least-wrong option and your data quietly overstates every item on the list.

3. Single choice (select one)

A list where exactly one option can be chosen. Use it when the options are mutually exclusive and you want clean segmentation: role, plan, primary reason, main channel.

Example: "What is the main reason you are canceling?"

Mistake to avoid: overlapping options. "Too expensive", "Not enough value" and "Budget cut" all read as money to a respondent in a hurry, and the split between them means nothing afterward.

4. Likert scale

A statement plus a symmetric agreement scale, usually five or seven points from strongly disagree to strongly agree. Use it to test a specific claim rather than a vague feeling, which is why it dominates employee and attitudinal research. Our Likert scale guide covers point counts, labeling and the neutral-midpoint argument.

Example: "I have the tools I need to do my job well." Strongly disagree to strongly agree.

Mistake to avoid: an unbalanced scale. If the options run from "Good" to "Excellent", everyone scores well and you have learned nothing. Negative and positive points must mirror each other.

5. Rating scale

A direct judgment on a numbered range, typically 1 to 5 or 1 to 10, or a row of stars. Use it when you want a number you can average and watch over time. It is faster than a Likert scale and easier to read on a phone. See the rating scale questions guide for choosing the range and labeling the ends.

Example: "How would you rate the support you received today? 1 to 5."

Mistake to avoid: changing the range between surveys. A 1 to 5 in March and a 1 to 10 in June are not comparable, and nobody will remember which was which by the time someone asks.

6. Ranking

The respondent puts a short list in order of importance. Use it when everything on the list sounds important in isolation and you need to force a trade-off, such as which three of five roadmap items matter most.

Example: "Rank these four improvements from most to least valuable to you."

Mistake to avoid: ranking more than about five items. Beyond that people rank the top two, guess the rest, and your averages for positions four through eight are noise. The ranking questions guide covers when a max-selection question works better.

7. Matrix

One scale applied to several rows at once, such as rating six aspects of a conference on the same 1 to 5 range. Use it to save time when the items truly share a scale.

Example: "Rate each of the following from 1 to 5: venue, schedule, sessions, catering, networking."

Mistake to avoid: long matrices on mobile, where a wide grid becomes a pinch-and-scroll puzzle and people straight-line down one column to escape. The matrix questions guide explains how to break a matrix into separate screens without losing comparability.

8. Yes / No

A binary. Use it as a gate before branching, or to record a plain fact. It is the cheapest question in a survey and the best opener.

Example: "Did you contact support about this?"

Mistake to avoid: using a binary for something that is really a matter of degree. "Were you satisfied?" throws away everything interesting between yes and no.

9. Demographic

Role, company size, industry, region, age band. Use these only when you plan to cut the results by them, and put them at the end, where they feel like paperwork rather than an interrogation.

Example: "How many people work at your company?"

Mistake to avoid: collecting personal characteristics you have no analysis plan for. It costs you completions, and under GDPR it costs you a justification you may not have. Our guide to GDPR-compliant forms covers what to ask and what to leave out.

120 survey questions you can copy

Below are 120 questions grouped by what you are trying to learn. They are written to be pasted straight into a survey and edited lightly for your product name and tone. Take five or six from a group rather than all fifteen: the point of a long list is choice, not length. If you want these already assembled, the survey templates gallery has ready-made versions you can clone.

Customer satisfaction questions (15)

• How satisfied are you with your most recent purchase?
• How well did the product meet the expectations you had before you bought?
• How easy was it to find what you were looking for?
• How would you rate the quality of the support you received?
• How long did you wait before someone replied to you?
• Did we solve your problem the first time you contacted us?
• Which part of the experience was most frustrating?
• What almost stopped you from buying?
• How would you rate the value you got for what you paid?
• How clear was the pricing before you paid?
• If you could change one thing about the experience, what would it be?
• How likely are you to buy from us again?
• How does this compare with the service you used before?
• What did we do better than you expected?
• Was there anything you expected to find and didn't?

NPS and loyalty questions (10)

• On a scale of 0 to 10, how likely are you to recommend us to a friend or colleague?
• What is the main reason for the score you gave?
• What would we have to change for you to give us a 10?
• Who else do you think would get value from this?
• How long have you been a customer?
• How often do you use the product?
• If we shut down tomorrow, what would you use instead?
• How disappointed would you be if you could no longer use this product?
• Would you renew today without having to think about it?
• Have you recommended us to anyone in the past six months?

The first question is the standard Net Promoter question and should be asked in exactly that form if you want your score to be comparable to anyone else's. The NPS survey template pairs it with the follow-up above, which is where the useful part actually lives.

Product feedback questions (15)

• Which feature do you use most often?
• Which feature have you never used?
• What were you trying to do the last time you got stuck?
• How easy was it to set up your first project?
• What is the one thing we should build next?
• Which task still takes longer than it should?
• What do you do outside the product that you wish you could do inside it?
• How well does the product fit the way your team already works?
• What did you expect this feature to do that it doesn't?
• How would you describe this product to a colleague, in your own words?
• What nearly made you give up during your first week?
• Which other tools did you compare us with?
• How important is this feature to the way you work?
• Where in the product do you most often feel unsure what to do next?
• If you could delete one part of the product, what would it be?

Onboarding questions (10)

• What made you sign up today?
• What are you hoping to get done in your first week?
• How did you first hear about us?
• What are you using today to solve this?
• How many people will use this with you?
• What is your role?
• How much of the setup did you finish without needing help?
• Which setup step was least clear?
• How confident do you feel using the product on your own?
• What would have made the first ten minutes easier?

Churn and exit questions (10)

• What made you decide to cancel?
• Which of these best describes your reason for leaving?
• Was there a specific moment when you decided to stop?
• What were you hoping the product would do that it didn't?
• Are you moving to another tool, and if so which one?
• What would bring you back?
• How would you rate the value you got for what you paid?
• Did you contact support before you decided to cancel?
• How easy was it to cancel?
• May we email you if the thing you were missing ships?

Employee engagement questions (15)

• How likely are you to recommend this company as a place to work?
• I have the tools and information I need to do my job well.
• I am clear about what is expected of me this quarter.
• When did you last receive feedback on your work that you found useful?
• My manager supports my development.
• I feel comfortable raising a concern with my manager.
• How would you rate your workload over the past month?
• How often do you work outside your normal hours?
• The work I do is recognized.
• Teams here share information well with each other.
• I can see a path to grow in my role here.
• How would you describe the culture here in one sentence?
• What is slowing your team down the most right now?
• What is one thing leadership should stop doing?
• Do you expect to still be working here in twelve months?

Event feedback questions (10)

• How would you rate the event overall?
• Which session was the most useful to you, and why?
• Which session would you cut next time?
• How would you rate the venue and the facilities?
• Was the schedule too short, about right, or too long?
• How easy were registration and check-in?
• How useful were the networking opportunities?
• What topic should we cover at the next event?
• How likely are you to attend again?
• What is the one thing we should change next time?

Course and training feedback questions (10)

• How would you rate this course overall?
• How clearly were the objectives explained at the start?
• Was the pace too slow, about right, or too fast?
• How useful were the exercises compared with the lectures?
• How confident do you feel applying what you learned?
• Which module was hardest to follow, and what made it hard?
• How helpful was the feedback you received on your work?
• Were the materials easy to access and use?
• How likely are you to recommend this course to a colleague?
• If the course had one more hour, what should we add?

Market research questions (10)

• How do you solve this problem today?
• How often does this problem come up?
• What does it cost you when it goes wrong?
• Who decides which tools your team buys?
• What budget range would a tool like this fall into for you?
• Which of these three descriptions sounds most like what you need?
• Where do you look first when you research a new tool?
• What would make you switch from what you use today?
• How many people at your company would use something like this?
• What would stop you from buying this today?

Website and UX questions (10)

• What were you trying to do on this page?
• Were you able to finish it?
• If not, what got in the way?
• How easy was it to find what you needed?
• What information did you expect to find here that wasn't here?
• In your own words, what does this company do?
• What made you hesitate before clicking?
• Was anything on this page confusing or unclear?
• How did you get to this page?
• What would you like to see on this page that isn't here yet?

Demographic questions (5)

• Which age range do you fall into?
• Where are you based?
• Which of these best describes your role?
• How many people work at your company?
• Which industry do you work in?

Seven ways survey questions go wrong

Bad data rarely announces itself. A biased survey question still returns a tidy chart, which is exactly why these seven failures survive so long. Each one below comes with the rewrite.

1. Leading questions

The wording carries the answer inside it. People are agreeable, especially to the company that just served them, and they will follow the direction you point.

Bad: "How much did you enjoy our redesigned checkout?"
Better: "How would you rate the checkout experience?"

2. Double-barreled questions

Two questions wearing one coat. Whatever number comes back, you cannot tell which half it refers to.

Bad: "How satisfied are you with the speed and accuracy of our delivery?"
Better: Two questions, one about speed and one about accuracy.

3. Loaded questions

The question smuggles in an assumption the respondent never agreed to, so any answer concedes the premise.

Bad: "What do you like most about our new pricing?"
Better: "How do you feel about our new pricing?" followed by an open question asking why.

4. Absolutes

Words like always, never, all and every force people into a position they do not hold, and they answer for the exception rather than the pattern.

Bad: "Do you always read our product emails?"
Better: "How many of our product emails did you open in the past month? None, some, most, all."

5. Jargon and internal language

Your feature names, acronyms and internal team words mean nothing outside the building. People guess, and their guess becomes your data.

Bad: "How useful is the unified activity layer?"
Better: "How useful is the page that shows everything your team did this week?"

6. Too many options

A list of eighteen choices gets read for about five seconds. People pick from the top, which is why order in a long list quietly becomes the strongest predictor of the result.

Bad: A single-choice question with eighteen industries listed alphabetically.
Better: Six or seven grouped options plus "Other", or a searchable dropdown when the full list genuinely matters.

7. Asking what you will not act on

The most expensive mistake, because it costs you the response and the relationship. If you ask what someone wants built and nothing changes for two years, they stop answering, and so does everyone they told.

Bad: "What features would you like to see?" asked quarterly with no follow-up.
Better: Ask it once, publish what you decided and why, and ask again only when there is something new to decide.

How many questions should a survey have?

There is no magic number, but there is a real mechanism: every question is a chance to quit. A respondent decides to continue at each screen, and that decision gets harder as the perceived remaining effort grows. So the honest answer is not a count, it is a budget. Decide how much of someone's time you have earned, then spend it.

For most teams the working ranges look like this.

Two things matter more than the count. The first is perceived effort: five open-text boxes feel longer than fifteen single-choice questions, so mix formats and front-load the easy ones. The second is relevance, which is what conditional logic buys you. A twenty-question survey where each person only sees the eight that apply to them behaves like an eight-question survey, because it is one.

Then stop guessing. Upform's form analytics shows a per-question drop-off funnel next to your completion rate and average completion time, so you can see the exact question where people leave. In practice it is almost never "too many questions" in general. It is one specific question, usually a required open-text field or a demographic question asked too early. Cut that one and the rest of the survey recovers. Our guide to form abandonment goes through the common culprits.

Order matters as much as wording

The same set of questions performs differently depending on the sequence. Three rules carry most of the benefit.

Easy first. Open with something a respondent can answer in one tap and without thinking: a yes/no, a rating, a single choice. The first answer is the commitment. Once it is given, people finish far more often, which is why a demographic question in position one is such a costly habit.

Sensitive last. Salary, age, company size, contact details and anything personal belong at the end. By then the respondent has invested effort and understands what the survey is for, so the ask feels reasonable rather than intrusive. It also means that if they abandon at that point, you still keep everything they already answered.

Group by topic, and change subject deliberately. Jumping between support, pricing and product and back again makes a survey feel twice as long. Keep related questions together and put an explicit sentence at each shift, so people know they have crossed into a new section rather than being asked the same thing again.

One more effect worth knowing: earlier questions prime later ones. Ask five detailed questions about problems with your support team and then ask for an overall rating, and the overall rating will be lower than if you had asked it first. If you care about a headline score, ask it before the diagnosis, not after.

What changes when you ask one question per screen

Every rule above applies to a page-long form. But a conversational form, where each question gets its own screen, changes what you can practically ask. Long lists of choices become readable instead of cramped. A single open-text question in the middle of a sequence gets a proper answer rather than three words, because it is the only thing on the screen. And because you see exactly one question at a time, you never see the intimidating wall that makes people close the tab before they start.

It also changes the arithmetic of branching. When questions arrive one at a time, skipping the irrelevant ones is invisible: nobody notices the eleven questions they did not get. That is what makes a longer, better-targeted survey possible at all.

That is the format Upform is built around, and the free plan includes all 21 question types, conditional logic and unlimited responses, so the survey you design is the survey you can send. The how to make a survey guide walks through building one end to end.

Survey question FAQs