Surveys 101: how to choose
the right questions

In this article you’ll learn which questions to use, which to discard, and how you can tailor your questions to your audience for authentic, actionable feedback.

Keep reading to explore.

Or if you’re already itching to make a survey, just go ahead.

Open-ended questions provide depth

A Freudian favorite, the open-ended question (How do you feel about …) invites people to share and go deeper than yes or no. Open-ended questions get you spontaneous, organic responses with one caveat—answers might snowball off topic.

Guide the audience with multiple choice questions

In this popular question type you give people a fixed number of options to choose from and, depending on what you want to know, the survey taker picks one or several of them.

Multiple choice is a great way to test what the audience remembers or prefers out of the given options, but it’s an ineffective way to explore ideas and thoughts.


So what’s the right way to ask a multiple choice question? You make sure it’s the question itself, and not the answers, that provides information and context. The answers have only one job, and that is to differ from each other.

Here’s how not to do it:

Regarding citizenship, which of these applies to you? Choose one or several.

  • I am a European citizen
  • I am an African citizen
  • I am an Asian citizen
  • I am a North American citizen
  • I am a South or Central American citizen

And here’s how to do it:

Which citizenship(s) do you hold?

  • European
  • African
  • Asian
  • North American
  • South or Central American

Structure your question so clearly that the survey taker will understand it the first time they read it. If your audience gets confused or annoyed by repetition, it’s likely they will ditch your survey.

Do this and your multiple choice question will move people toward the finish line, not drag them along.

Use scaled question types for attitudes and opinions

The audience might not trust you enough to say what they really think. If so, scaled questions can make the more sensitive ones easier to digest.

These types of questions also have subcategories, including 1) ordinal scale, 2) ratio scale, and 3) interval scale.

1) Ordinal scale ranks things by quality, such as Boring ←→ Fun or Sick ←→ Healthy. This is a great way to measure basic attitudes because people are often in tune with them without giving it much thought.

2) Ratio scales measure degrees of difference where the numbers on the scale are tied to a unit with proportions. For example, if a 6 on your scale measures dollars, length or weight, a 3 represents half the actual value.

3) Interval scale questions also rank degrees of difference, but in this case proportions have no meaning. On a scale of 15, where a 1 means disagreement and a 5 means agreement, a 4 doesn’t really mean “twice as much agreement” as a 2.


Always label the extreme ends of your scales clearly enough to rule out misinterpretations. Imagine taking a survey. You’re asked how much you like dinosaurs on a scale of 110. If a 10 means they’re “great” and a 0 means they’re “ok”, where does dislike or hate of dinosaurs fit in?

Also beware of cultural contexts. In poker, is the ace of spades higher than the ace of hearts? In the UK, hearts is often higher than spades, but not in the US. Keep this in mind, be extra clear, and save people from headaches.

Check out Typeform’s survey templates for inspiration. We have templates for customer satisfaction, market research, brand awareness and many more.

Avoid constant sum questions

They may look scientific and professional, but constant sums can seem complicated and should be avoided—especially if you don’t know your audience well. Here’s an example: “Distribute 100 points across A, B, and C based on what you value most in a friend; A) honesty, B) loyalty, C) shared history.

Take it as a rule of thumb that people won’t like doing math. Constant sum questions are typically inappropriate for paper surveys, but they work better online. Go with simple.

Reducing dropout rates

Some surveys (hopefully not yours) don’t give enough options. Sometimes the options aren’t appropriate. It leaves people feeling like they have to either lie or just drop out of the survey. To prevent this, include an option for Other or None of the above when appropriate.

Survey makers often ask for information that has no value and that wastes attention. If you ask which brands of dog food someone recognizes, a response choice like Anything with puppies! won’t help you understand brand awareness.

Closing thoughts

Let’s face it. Taking a survey is probably not the highlight of someone’s day. Respect people’s time and attention. How? Try to consider your target audience as well as what you want from them.

  • How much time and attention can the audience spare for this?
  • How motivated are they?
  • Which questions get me the most nuanced and honest answers for the least effort?
  • Will my questions give me information I need to know, or information I just want to hear?

Remember that the point of every survey is truth, not approval or comfort.

Creating the perfect questions is key. But it’s just the beginning. To start a conversation and get something out of it—and not just collect data—a one-trick pony wouldn’t make the cut.

That’s why Typeform is the place that takes you full circle. Create and design with no coding or design skills. Drag and drop your questions to structure it in minutes. Connect it to Google Sheets, Trello, Gmail, and more to make things even easier. With the smartest logic feature on the market, your survey can change behavior or outcome based on how they answer. What’s more, it can address people by name and will always asks one question at a time, like any dialogue.

Once you’re done creating your survey, you need to reach your audience. Just embed your Typeform survey in an email, website, or blog. Once you’ve reached them and have everything you need, we help you gather and export all responses in a comprehensive, actionable report.

Typeform isn’t just here to help you ask the best questions, but to engage in conversation. We’re here from hello to see you soon, and way after that.