Types of Survey Questions: A Guide for Market Researchers

You send your survey to 500 customers and ask them, "How satisfied are you with our service overall?” Then you receive the result: 72% say “Satisfied.” Sounds great, but now what? What are they satisfied with exactly? Is it speed? Maybe it’s your price? Or maybe your support? In the end, you’re left guessing.

This survey returned a number, but it didn’t give you any answers.

The point is that not all survey questions produce actually usable data. The structure of each question dictates what you can evaluate, how respondents will understand that question, and if the answers you get can help you with your next moves.

At ENC (Expert Network Calls), we created this guide to walk you through the survey question types used in professional research programs.

Overview: main types of survey questions. How to choose the right one.

The table below features common types of survey questions. A thorough description of each type is provided further in the article.

*The information used in the table is based on materials from the American Association for Public Opinion Research (AAPOR), Pew Research Center, 2026 article by Md Taufique, K. R., Md M. Sabbir, and F. K. Rabbanee. “A Guide to Key Decision Criteria for Likert-Scale Use in Survey Research.” Global Business and Organizational Excellence 45, no. 5: 493–508, and the Sage Encyclopedia of Survey Research Methods.

1. Open-ended questions

What are open-ended questions?

Open-ended questions let participants respond in their own words rather than choosing from set of answers you offer them. These can be considered a key tool for qualitative research within a survey.

• Example: “Why did you decide to switch vendors?”

Such questions are often used during exploratory studies, before you’ve come up with a hypothesis, or when you want to hear reasoning. They can also be used as follow-ups to closed questions to give more explanation.

Pros and cons of market research

Pros: Richer insights, respondents can speak freely, and researchers can better understand people’s emotions and motivations and get quotable feedback. They can also help uncover new trends or areas for future research.

Cons: Time-consuming, answers vary and are thus hard to compare, skips and drop-offs can occur (not everyone wants to put more effort into completing the survey), subjective interpretation from researchers, and harder to convert into percentages or charts.

Common research mistake

Sometimes, placing an open-ended question too early in a survey is a no. When there’s a lack of context, respondents tend to give general, ambiguous responses.

How to fix it.

First, ask one or two closed-ended questions to “warm the respondent up,” helping them organize their thoughts. Then, add an open-ended question like "Why?" or "What else should we know?" that could help you get better insights.

2. Closed-ended questions

What are closed-ended questions?

This type of question offers respondents a limited, predefined set of options they can select from. These can ask for concrete facts or data or limit responses to "yes" or "no," says Sage, the global academic publisher of books and journals. They may be either single-select (pick one) or multi-select (select all that apply).

Multiple-choice questions are a popular subtype of closed-ended questions. It offers respondents multiple answer options.

Example:

Which research methods do you use? (Select all that apply):

1. Expert network calls.
2. Online surveys.
3. Focus groups.
4. Other (please specify).

When to use multiple-choice questions

These are useful for collecting quantifiable, measurable data that can be easily evaluated or compared.

1. Use single-select when you only need one answer - job level, primary vendor, or industry.
2. Use multi-select when respondents can potentially do several things at once - channels they monitor, tools they use, methods or strategies they rely on, and the like.

Multiple-choice questions are suitable for keeping track of changes over time, for example.

Common research mistake

Not including the “Other (please specify)” option pushes respondents to select the closest available option, which cannot be accurate and can thus impact your data.

Choice overload is another mistake. Too many options at once can lead to respondents not being willing to read every option carefully. This increases the risk of respondents selecting a random answer.

3. Rating scale questions

What are rating scale questions?

Rating scale questions ask respondents to assign a number (usually on a 1–5, 1–10, or 0–10 scale) to show how strong, satisfied, or likely they feel.

Example:

How likely are you to recommend this vendor to a colleague? Rate from 0 to 10.

• 0 = Not at all likely
• 10 = Extremely likely

Rating scale questions have different subtypes. They may ask you to rate something using numbers, stars, or a slider.

Two of the most popular rating scales are the Net Promoter Score and the Likert scale.

Net Promoter Score (NPS)

The net promoter score estimates customer loyalty. It was introduced by Fred Reichheld at Bain & Company in 2003. NPS uses a 0-10 scale. Respondents are grouped into three segments:

1. Promoters: Responses from 9 to 10.
2. Passives: Responses from 7 to 8.
3. Detractors: Responses from 0 to 6.

% Detractors - % Promoters = loyalty score (promoters > detractors = a good loyalty score)

Likert scale questions

What is a Likert scale?

This method was created by psychologist Rensis Likert in 1932. Respondents are asked to indicate to what level they agree or disagree with a certain statement. For this, the scale uses a set of labeled points, usually five or seven: strongly disagree, disagree, neutral, agree, and strongly agree.

Example:

Our research vendor delivers useful insights.

• Strongly disagree.
• Disagree.
• Neutral.
• Agree.
• Strongly agree.

When is a Likert scale used?

A Likert scale can serve as a tool for analyzing employee engagement, tracking customer satisfaction, studying brand perception, and evaluating suppliers. It is extremely useful since most attitudes, such as trust, satisfaction, and willingness to recommend, cannot be evaluated as simply yes or no.

Central tendency bias - Likert scale’s issue

This common problem occurs when respondents are less inclined to select a strong opinion (strongly agree or strongly disagree) and tend to select the middle point. In this case, understanding their position is more difficult.

One of the ways to solve this (unless the midpoint is important for your survey) is to avoid a neutral point, leaving just 4 steps, to push respondents toward a clear answer.

4. Dichotomous questions (Yes/No)

What are dichotomous questions?

These questions have only two answer options, for example, Yes/No, True/False, or Agree/Disagree.

Market researchers use dichotomous questions for:

• Screening respondents (Example: "Have you purchased a laptop over the last 12 months? Yes/No.")
• Classifying respondents (Example: "Are you currently using any products of Brand X? Yes/No")
• Analyzing behaviors or facts (Example: "Did you visit our website this month? Yes/No")
• Reducing survey length and effort.

When do dichotomous questions fail?

They can tell you what people think or do, but don’t explain why.

For example, "Would you buy this product? Yes/No" - this sounds useful. But does not reveal why they would or wouldn’t buy it, what features matter most, or whether they have any concerns.

5. Ranking questions

What are ranking questions?

Respondents are asked to provide a list of answers in their preferred order (from the most to the least preferred).

Example:

Rank the following factors in your vendor selection, with 1 being most important:

• Speed of response.
• Expert quality.
• Price.
• Ease of use.

When should you use ranking questions?

Use them to determine priority levels among limited choices when deciding on compromises concerning resource distribution, for example. Such questions can help you unveil the options that respondents consider the most important.

Limitation

Respondents must be aware of all of the items on the list. Otherwise, the survey turns into a guessing game. Too many answer options can make participants feel overwhelmed.

6. Matrix/grid questions

What are matrix questions?

A matrix question is a group of related questions presented in a grid format, with individual questions or statements arranged as rows and predefined answer options as columns. Answer options are usually presented on a scale (however, non-scaled options like Yes/No or categorical choices are also possible).

When should the matrix questions be used?

These work best when you need to gather answers to a number of related questions using the same set of answer options. Most common uses include:

• Customer satisfaction surveys (respondents rate multiple aspects of an experience).
• Sub-topics within a larger questionnaire (a group of related questions can be brought together for a better understanding).
• Multiple rating-scale questions (this saves you space and makes the survey clearer).

Example:

How would you rate the following features of our mobile app?

Common research mistake

There are cases of straight-lining when respondents select the same answer option across several rows of a matrix question without actually reading each statement. The issue often occurs when the matrix question is viewed on mobile devices, since large matrices are harder to navigate.

How to fix it.

• Keep matrix questions short (around 5-7 rows).
• Perform trials on mobile devices before launch.

7. Demographic questions

What are demographic questions?

In market research, demographic questions are used to gather background information about the respondent or group of respondents, which can be useful for market segmentation and targeting. This can include information on job title, seniority, company size, industry, and region.

Example:

What is your employment status?

• Employed full-time.
• Employed part-time.
• Self-employed.
• Unemployed.
• Student.
• Retired.
• Other.

When are demographic questions used?

Data collected from the results of demographic questions can help you develop a product and service offer, decide on pricing, and organize customers according to their needs and interests.

Tips:

• Put demographic questions near the conclusion of the survey, as starting off with personal questions can limit engagement. An exception is when you need to screen respondents - that’s where these should be used earlier.
• Include a “Prefer not to answer” - option, which is required for sensitive topics like income or ethnicity. This makes respondents feel more comfortable and thus helps you avoid survey abandonment.
• Use just the right number of questions to get the best possible responses.

Quantitative vs qualitative survey questions

One counts, the other one listens.

• Quantitative questions are closed-ended questions like multiple choice, rating, and ranking (help you learn how many, how much, how often, etc.). They generate numbers you can compare.
• Qualitative questions are mostly open-ended and provide you with valuable context, motivation, and reasoning (help you learn why, how, and what else).

Professional research programs may apply both types of questions.

Frequently asked questions (FAQ)

What are the types of survey questions?

There are two main types of survey questions: Closed-ended and open-ended questions that are divided into several other types:

Closed-ended: Multiple choice, dichotomous (yes/no), rating scale, ranking questions, and matrix questions.

Open-ended: Unstructured (blank text box), sentence completion, word association.

What is the difference between a Likert scale and a rating scale?

"Rating scale" is a broader term. This scale uses a numerical range to measure a degree or intensity. The Likert scale is a particular type of rating scale. It examines opinions, attitudes, or behavior within a range (for example, between “Strongly Disagree” and “Strongly Agree”, when the former is 1 and the latter is 5).

How many questions should a market research survey have?

There's no universal rule. Some say it should be under 10, others mention 15-20 questions, which means that usually a survey may have between 5 and 20 questions. If there are very few questions, you may have important gaps in your data analysis, while an abundance of questions can lead to drop-offs. Focus on questions that could bring a lot of useful insights for you.

The importance of survey questions in market research

Survey questions help companies collect quantitative and qualitative data about markets and opportunities, customers’ needs and preferences, product ideas, trend tracking, and competitors.

Surveys gather data from large groups of customers or the general market.

However, respondents may lack deep industry knowledge. It is here that researchers turn to the Expert Networks.

Expert Network Calls (ENC) is a marketplace that unites 22 trusted Expert Networks in one platform, giving clients combined access to 3M+ experts with niche knowledge.

Expert calls deliver deep, forward-looking insights from knowledgeable insiders (executives, practitioners, experts in niche fields) who offer insights that surveys might miss.

ENC provides free transcripts and AI-generated summaries for every call, available in 8 languages.