Questions to ask in a survey: how to write good ones

Ten rules for writing survey questions people can answer honestly, seven before-and-after rewrites, and more than 40 questions to ask in a survey for customers, apps, websites, employees and events.

By Ethan Brooks, Software Engineer at Florm

Published November 10, 2026 · Editorial policy

Quick answer

A good survey question asks one thing, uses plain words, stays neutral, names a specific time period and offers balanced, complete answer options. Start from the decision you need to make, ask only what you will act on, put easy questions first and test the survey on a phone before you send it.

Start with an objective, not a list of questions

Before you write any question, finish this sentence: “After this survey, I will decide whether to ____.” The blank could be “change the checkout page”, “keep the weekly team meeting” or “add a feature”. Every question on the survey should help make that decision. A question that only satisfies curiosity makes the survey longer and lowers the number of people who finish it.

Then pick the type of question that fits each piece of information: a rating scale to measure how much, a choice question to find out which, a yes/no question for a fact, and an open question for why. Wording comes next.

10 rules for writing good survey questions

Ten rules for writing survey questions with reasons and examples
Rule	Why it matters	Example
Ask one thing at a time	A double-barreled question has two answers hidden in one.	Bad: “How satisfied are you with our prices and delivery?” Better: ask about prices and about delivery separately.
Use plain words	Jargon, acronyms and long sentences slow people down and cause misreadings.	Bad: “How would you rate the UX of our onboarding flow?” Better: “How easy was it to get started?”
Stay neutral	A leading question suggests the “right” answer and inflates scores.	Bad: “How great was your experience?” Better: “How was your experience?” with a balanced scale.
Be specific about time and place	Vague periods like “often” or “recently” mean different things to different people.	Bad: “Do you shop with us often?” Better: “How many times did you buy from us in the last 3 months?”
Balance the answer options	A scale with more positive than negative choices pushes answers upward.	Use equal numbers of positive and negative points, with a neutral midpoint when neutral is a real answer.
Make options complete and non-overlapping	People get stuck when two options fit, or none does.	Bad: 18–25 and 25–34. Better: 18–24 and 25–34, plus “None of these” or “Other” where needed.
Avoid absolutes	Words like always, never and all invite “no” answers.	Bad: “Do you always read the instructions?” Better: “How often do you read the instructions?”
Ask about behavior, not intentions	What people did is more reliable than what they say they will do.	Bad: “Would you use a mobile app?” Better: “Which of these have you used in the past month?”
Put questions in a sensible order	Early questions color later answers, and hard questions cause drop-off.	Start with easy closed questions, then the key ratings, then optional open questions, then demographics.
Keep it short	Every extra question costs completions.	Ask only what you will act on. If you can’t name the decision a question feeds, remove it.

Most of these rules come from one idea: the respondent should understand the question the way you meant it and feel free to give any answer. When a question hints at an answer, hides two questions in one or uses a word that means different things to different people, the data stops describing your respondents and starts describing your wording.

Bad survey questions and how to fix them

These seven rewrites show the most common problems. Most can be fixed by splitting a question, removing an adjective or replacing a vague word with a number.

Examples of poorly worded survey questions and better versions
Problem	Weak question	Better version
Double-barreled	How satisfied are you with our product and support?	Ask two questions: “How satisfied are you with the product?” and “How satisfied are you with our support?”
Leading	How much did you enjoy our excellent service?	How would you rate our service? (1 Very poor – 5 Excellent)
Loaded	Where do you waste the most time at work?	Which of these takes up most of your working day? (Single choice list of tasks)
Vague	Do you use our app regularly?	How often did you use our app in the past 30 days? (Never / 1–3 times / Weekly / Almost daily)
Absolute	Do you ever have problems logging in?	How often do you have problems logging in? (Never / Rarely / Sometimes / Often)
Overlapping options	How many employees? 1–10, 10–50, 50–100	How many employees? 1–10, 11–50, 51–100, more than 100
Jargon	Does our UI meet your UX expectations?	How easy is our app to use? (1 Very difficult – 5 Very easy)

Hand holding a printed page of charts above a notebook, a pen and a calculator on a desk while planning a survey

How to write objective survey questions

Objective survey questions are worded so that two people with opposite opinions would both feel the question is fair. Four habits get you there:

Remove evaluative words. “Great”, “excellent”, “waste” and “annoying” tell people what you think. Put judgment into the answer scale instead (“Very poor … Excellent”).
Mirror the scale. Offer as many negative options as positive ones, in the same strength. “Very satisfied / Satisfied / Neutral / Dissatisfied / Very dissatisfied” is balanced; “Excellent / Very good / Good / Fair” is not.
Prefer facts over opinions where you can. “How many times did you contact support?” is easier to answer truthfully than “Is our support too slow?”.
Don’t reveal the sponsor’s hopes. Avoid intros like “We’ve worked hard to improve…” before the rating.

An intro that is neutral and short helps: say who is asking, how long the survey takes and what the answers will be used for. For closed questions, the guide to closed-ended questions shows how to build complete, non-overlapping answer lists.

Questions to ask in a survey, by purpose

If you are staring at a blank form and asking “what survey questions to ask”, start from the purpose below and cut anything you won’t act on. The type of each question is in brackets.

Customer survey: what to ask after a purchase or support contact

Pair one satisfaction rating with questions that explain it.

Overall, how satisfied are you with your experience? (Rating scale, 1–5)
What was the main reason for your visit? (Single choice)
Did you find what you were looking for? (Yes / No)
How easy was it to get what you needed? (Rating scale, 1–5)
How likely are you to recommend us to a friend or colleague? (NPS, 0–10)
What could we do better? (Long text, optional)

App survey questions (in-app or after a session)

Ask about one recent action, not the whole product, and keep it to three or four questions.

How easy was it to complete [task] in the app? (Rating scale, 1 Very difficult – 5 Very easy)
How often do you use the app? (Single choice: Daily / Weekly / Monthly / Rarely)
Which feature do you use most? (Single choice list of features)
Did anything not work the way you expected? (Yes / No)
If yes, what happened? (Long text)
What one thing should we improve first? (Short text)
How likely are you to recommend the app to a friend or colleague? (NPS, 0–10)
Which device are you using? (Single choice: Phone / Tablet / Computer)

Website visitor survey questions

Short and easy to skip works best for visitors who did not ask to be surveyed.

What brought you to our site today? (Single choice)
Did you find what you were looking for? (Yes / No)
How easy was it to find information on this site? (Rating scale, 1–5)
What information was missing, if any? (Long text, optional)
How did you hear about us? (Single choice)

Employee survey questions

Use statements people can agree or disagree with, and protect people’s privacy.

I have the tools and resources I need to do my job well. (Likert, 1–5)
My manager gives me useful feedback. (Likert, 1–5)
I see a clear path to grow here. (Likert, 1–5)
How likely are you to recommend this company as a place to work? (NPS, 0–10)
What is one thing we should change? (Long text, optional)

Event and workshop survey questions

Send it the same day, while the event is fresh.

Overall, how would you rate the event? (Rating scale, 1–5)
Which session was the most useful to you? (Single choice)
Was the length of the event about right? (Single choice: Too short / About right / Too long)
Which topics would you like next time? (Multiple choice)
Would you attend a future event? (Yes / No)
What is one thing we should improve? (Long text, optional)

For more free-text prompts, see the open-ended question examples. For demographic questions, use the demographic survey questions guide.

Question order and survey length

Open with an easy, relevant question. A simple closed question builds momentum.
Ask your key rating before related detail. If you ask about support, delivery and price first, those topics color an overall satisfaction rating asked afterwards.
Group related questions together and keep the same scale for the whole group.
Put open-ended questions near the end and keep them optional. They take the most effort.
End with demographics and contact details. Ask only what you need, and make sensitive questions optional.

For length, a short survey that people finish beats a long one that most abandon. For feedback after a single interaction, three to six questions is usually enough. Add questions only when you can name the decision they feed.

Test your survey before you send it

Ask two or three people who were not involved in writing it to complete the survey on a phone while they think aloud. Watch for questions they re-read, answer hesitantly or interpret differently from what you meant. Then fix the wording, send a small first batch, and check the results before sending to everyone. Use this checklist for a last pass:

Every question connects to a decision I will make.
Each question asks about one thing.
The wording is plain and has no jargon or acronyms.
No question hints at the “right” answer.
Time periods are specific (“in the last 3 months”).
Answer options are balanced, complete and don’t overlap.
Scales use the same labels every time.
Optional questions are marked optional; required ones are truly needed.
The order runs from easy to hard, with demographics last.
At least two people outside the project have tested it on a phone.

How to write and run your questions in Florm

Florm shows one question at a time, so each step holds a single idea, which helps you follow the “ask one thing” rule. Florm offers 12 question types from Yes / No to NPS.

Build the form in the visual editor and add one question per step. Edit titles, descriptions and answer options inline; what you see is what respondents get.
Choose the type that fits the answer: Rating scale for how much, Single or Multiple choice for which, Yes / No for facts and Long text for why.
Mark only essential questions as required, and keep open and sensitive questions optional.
Preview on your phone, then share by link or QR code. Pass context such as campaign or customer ID with hidden fields so you don’t have to ask for it.
After the first responses, check completion and drop-off in analytics. Partial responses show which question people stop at; rewrite or remove it.

Not sure where to start? Copy a ready-made set from a free template, such as the customer satisfaction survey, the feedback form or the online survey maker page for use cases.

Start free More survey guides

How to write survey questions FAQ

What questions should I ask in a survey?

Ask only questions tied to a decision you will make. A typical short survey has one key rating (for example satisfaction or NPS), one or two questions that explain it, one optional open question such as “What could we do better?”, and, if needed, a few demographic questions at the end. The question banks on this page list examples for customers, apps, websites, employees and events.

How do you write a good survey question?

Ask about one thing, use plain words, stay neutral, be specific about time and place, and give balanced, complete answer options that don’t overlap. Then test the survey with a few people on a phone to catch confusing wording before you send it.

What is an objective survey question?

An objective survey question is worded neutrally, so it does not suggest an answer or reveal what the author hopes to hear. It uses balanced answer options, avoids emotional or evaluative words such as “great” or “waste”, and asks about observable facts and behavior where possible.

What are good app survey questions?

Good app survey questions ask about one recent action or feature, not the whole product: how easy a task was, which feature is used most, whether anything did not work as expected, and what to improve first. Keep it to three or four questions and offer a way to skip.

How many questions should a survey have?

As few as will answer your objective. Most customer surveys work well with three to six questions, and every extra question lowers completion. Check where people stop with partial-response and drop-off analytics and cut questions that nobody finishes.

What is a double-barreled question?

A double-barreled question asks about two things at once, for example “How satisfied are you with our prices and delivery?”. A person who likes one and dislikes the other can’t answer it accurately, and you can’t tell which one the answer is about. Split it into two questions.

Should survey questions be required?

Make a question required only if the survey is useless without it. Optional open questions and sensitive demographic questions should stay optional; forcing an answer causes drop-off and low-quality answers.

Can I copy these survey questions?

Yes. Copy and adapt any question on this page for your own surveys, in Florm or any other tool.

Types of survey questionsTen kinds of question, with examples of each.Online survey makerCreate, share and analyze surveys with Florm.Feedback form templateA short, friendly form for bugs, ideas and opinions.

Good questions in. Clear answers out.

Create your first form in minutes with the free Early plan. No credit card required.

Start free Browse templates