Start with the decision, not the question format
A survey is most useful when you know what you want to learn before writing questions. Are you trying to discover problems you have not anticipated, compare experiences across groups, or decide which part of a service to improve? The answer helps determine whether respondents should choose from prepared options or explain something in their own words.
Closed-ended questions offer a defined set of responses, such as a rating scale, a yes-or-no choice, or a list of features. Open-ended questions ask for a written answer without prescribing its content. Neither format is inherently better. Each produces a different kind of evidence and places a different demand on the person answering.
Use closed-ended questions when comparison matters
Closed-ended questions work well when you know the likely answers and need to compare responses consistently. For example, “Which of these ways did you contact support?” is useful if the options cover the available channels. A question with the same response choices for everyone makes it easier to count selections and compare groups.
The quality of the answer depends on the quality of the choices. Options should be clear, distinct, and appropriate to the question. If someone could reasonably choose two answers, either allow multiple selections or ask for the single most important one. Consider adding “Other” with a space to specify an answer when the list may be incomplete. Where appropriate, let respondents say that a question does not apply rather than forcing an inaccurate choice.
Be especially careful with rating scales. Label the ends clearly, keep the direction consistent throughout the survey, and ask about one idea at a time. “How satisfied were you with the price and delivery?” is hard to interpret if someone liked one but not the other.
Use open-ended questions to uncover meaning
Open-ended questions are helpful when you do not yet know which answers to expect. “What was the hardest part of getting started?” gives respondents room to describe obstacles in their own terms. Their wording may reveal a confusing step, an unexpected need, or a reason that a prepared list would have missed.
Written responses can also explain a closed-ended result. If someone rates an experience poorly, a follow-up asking “What led to that rating?” may show whether the issue was speed, clarity, reliability, or something else. The explanation matters because the same rating can reflect very different experiences.
Open-ended questions take more effort to answer and more care to review. Some people may skip them or give brief replies. Use them selectively, keep prompts focused, and avoid asking for a long account when a short example would serve the purpose. When you analyze responses, look for recurring themes while preserving unusual answers that may point to a problem worth investigating.
Combine both formats in a purposeful sequence
A practical survey often uses closed-ended questions to establish patterns and a small number of open-ended questions to explore why those patterns appear. Suppose you want to improve an online checkout. You might ask whether a respondent completed a purchase, which step was most difficult from a list of steps, and then, “What made that step difficult?” Each question adds something the previous one cannot provide.
Do not attach a text box to every rating simply because it is possible. Choose follow-ups that could change how you interpret an answer or what you do next. You can also use an early round of open-ended questions to learn the language customers use, then draw on that language when creating response options for a later survey. Review those options before use; a handful of comments may not cover every respondent’s experience.
Check questions against the answers you need
Before sending a survey, read each question as if you were encountering it for the first time. Is the timeframe clear? Could someone answer accurately from memory? Does the wording assume an experience they may not have had? Could two people interpret a response option differently? A short test with people who resemble the intended audience can expose confusion before you collect a larger set of answers.
Finally, plan how you will use each answer. Closed-ended results can show how responses are distributed, but they do not explain every motive. Open-ended comments add context, but a vivid comment is not necessarily a common view. Use both with those limits in mind: choose the format that fits the decision, make it easy to answer honestly, and keep enough context to interpret the result.
