How to conduct better market research surveys

Do you find that sometimes your marketing research survey results can be confusing and invaluable?

Are you looking guidelines on how to improve how you capture data to drive your business decisions?

Well the first, and most important step is ensuring that you spend the time to plan and test your survey first. Are you structuring the questions in a way that will get you the information that you need?

This will ensure that you have data that is valid and actionable.

Too many time a study can start off with good intentions and because of deadlines, end up becoming a waste of money and valuable time. So if you’re this is your first time conducting a study or you are looking for tips to improve, this guide is ideal for you.

In this article I’ll go over:

    1. What type of Market Research survey should you conduct?
    2. What kinds of questions should you include in your Market Research survey?
    3. Marketing Research survey Do’s and Don’ts
    4. How long should you make your Market Research survey?
    5. Then we’ll work through an example survey built by someone on my team

What type of Market Research survey should you conduct?

Before you make any decision to conduct research you should think about what information you already have access to and what information you are lacking.
This will help you to clearly set an objective for your research and really guide how you should proceed.

If the information that you are looking for is an opinion, about understanding the user experience, or about getting to know your customers sentiments, feelings, or needs, you might want to consider conducting interview or a survey with mostly open field questions.

This will give you the most insight into how the participant feels in their own words. It will give you rich information straight from the consumer’s mouth and allow you to add follow-up questions on an interesting topic, in the case of an interview.

But, if what you are looking for is hard figures to use in decision-making, ratings of your product or feature, or a better understand of using buying habits, and demographics, I would recommend using a survey with multiple choice and rating scales answer choices.

In the end, your market research survey can be a combination of many different types of questions. Before starting you should be clear on what is your ultimate goal and what information will really help you reach that goal.

What kinds of question should you include in your Market Research Survey?

The two main question type are quantitative and qualitative questions. Quantitative questions supply an answer choice that you wish the user to select. Qualitative questions provide an open field where the user can provide a response of their own design.

  • Use quantitative questions when you want to have measurable data.
    • There are many ways to quantify data, you can use scales, multiple choice responses, data capture, check boxes. These questions are useful when you want to know how many people feel the same way, how many people have or will take the same actions, and as a basis for segmentation.
    • As a rule of thumb your answer choices should be mutually exclusive and collectively exhaustive. Maybe you’ve heard it before, but if not, when you are setting up the answer choices make sure that you include all possibilities a user could have (within reason) and make sure there is no overlaps in the choices.
  • Use qualitative questions when you want to understand underlying emotions
    • Qualitative question are a great way to understand how your target feels, in their own language. If you are asking for an opinion on how to improve, or what emotional reaction a person gets it is best to use this type of question.
    • The responses to these questions can become particularly important when you are creating marketing or other communications for this segment. You now know their pains and gains and can user their own words to get to the heart of this problem and show how your product or service offers the perfect solution.

Of course, it is possible to have both types of questions in a survey. If you find yourself with too many open field questions than maybe an interview might be a better solution than fielding a survey.


Marketing Research Survey Do’s & Don’ts

DON’T use complex working

Ask yourself, “could a 6th grader understand my questions?” It is a rule of thumb in advertising that copy should be written at a 6th grade level to ensure that it is understandable for everyone.

This is a good tool that you can also use to ensure that you are writing clearly and simply.

There are even tools made for Microsoft Word to measure the level of English. I used them when I was working on healthcare advertising and many of the concepts were extremely complex and need to be further simplified.

But there’s no need to get so complicated, you just think of the KISS principles, that, fun fact, was invented for the US Navy.“Keep it simple, stupid.”

Take out any industry jargon and over complicated terms, they aren’t adding any value. In fact they might be taking it away.

Ensure that you are stating your questions in the most basic way and that any language you use is universal.

DON’T use too many open questions

As mentioned previously, only use open questions when you really need to understand how your participant is feeling in their own words.

These questions are important because they don’t assume that the only answer choices are the one that you and your team have come up with on your own.

Remember the final objective of your research and think about the time necessary to analyze the results of the survey, before deciding to add open fielded questions.

DO use Collectively Exhaustive…but categorized answer choices

Do some of your answers naturally fall into the same bucket? Do you get the same sentiment from a group of responses?

Try to be as concise as possible, but think of the many ways your participant might respond.

If you have some previously knowledge about your target and would like to test if your assumptions are correct, you can put these assumptions as the answer choices.

Remember when you give a Yes or No questions to also have a middle choice. And when you have a list to include an “other.”

We don’t want to assume that we have all of the answers.

Don’t use leading question

This might seem like an obvious don’t, but you can’t imagine even after knowing it, how many people inadvertently use leading questions in their surveys.

I’ve seen it in surveys coming from even major brands, so it is worth taking some time to reiterate it.

Often we want to get a specific response in our survey because we believe our experience points us in that direction.

But, remember, the point of research is to uncover truths that we may not have been thinking about previously. Re-read and reword all of your questions to ensure you are not leading the participant to the response that you want them to give. An example of a leading question might be “How do you feel about the slow customer service from competitor brand A?” Instead try this more neutral question – “What is your opinion of the speed of customer service from competitor brand A?” or “Could you rate the customer service of competitor brand A on a scale of 1 to 10, with 1 being very slow service and 10 being rapid service? Additionally, you could take speed out of the question here to learn more of overall opinions of the customer service level provided.

DON’T include double barrel questions

A double barrel question is when you lump two thoughts or actions together in your question. One example would be “Do your current suppliers deliver on-time and offer discounts?” – This question asks about two topics you are keen to get information about, but by not separating them you will miss valuable information if the participant only answers one of the questions or gives an ambiguous answer. 

DO think about how you will use the data to formulate how to structure the answer choice

If you would like to know how often a customer received late order you would not want to say “Do you often receive late orders?” A better question would be – “How many times have you received late order in the last month?” This question will get you quantitative data that could be easily acted upon later.

DO offer bucketed responses when possible

A bucket response prevents the interviewee from having to remember exact details that might frustrate him during an interview. In our previous point you might want to ask “How often do you receive late order? Never, once a month, once per quarter, once per year?” The easier you make the questions to answer, the more likely you will get a valid response.

DO use visual aids to guide the response

Having worked in many developing countries the concept of a percentage is often difficult to explain. Building in visual aids into your questions can help define hard concepts and get you the measurable results you are seeking. Bring 10 plastic chips with you to help explain percentages visually. Show images of brand of products and ask customers to group them. There is nothing that says a questionnaire or interview cannot be interactive.


How long should you make your Market Research Survey?

The length of the survey really depends on your objective and what you are willing to offer in return for the response. If you want more in-depth research or an in person / virtual interview that might take a participant half and hour or more a monetary benefit is usually warranted.

Try a gift card from Amazon. It’s something that you don’t need to buy in advanced, you don’t pay fees for and you can send virtually.

I have both conducted and participated in surveys offering gift cards from $20 to $100. When you get to the higher end the participant understands that they are expected to do more.

Make your budget in advanced and think about how many participant you need to get data enough data to make informed decisions.

When building personas I only wanted 20 in person interviews, but when conducting quantitative research to build our product roadmap I needed hundreds of respondents. In the second case it would be too costly for a small company to offer a monetary benefits to everyone, so we offered our product for a limited time to respondents.

If you have a product or service, this is a good alternative to spending extra money on your research and you might end up with even more loyal users. If you don’t have anything to give away and you are hoping for the good will of your respondents you should try to make the survey as short as possible.

But in this case it is most likely that only your most loyal followers will be willing to participate. So be clear on who your audience is and what is the objective of your research.

Working through an example

I find that it is easiest to learn when you have an example to work with. I was recently instructing a young colleague of mine in creating his first survey. The object of the survey was to get an understanding of the feelings of the potential consumer in their own words. To understand their pain points and barriers to following a diet.

The original survey was conducted in Spanish, but I have translated it here for you.


What are the problems with this first question?

Take a moment and think about it before reading my response.

First. I don’t like the introduction to the survey. We are not giving them an idea of why we are asking for their help, not telling them how much time this will take them. We are missing information. We could have increased the number of respondent by simply adding a sentence of explanation.

Secondly, if you said that this is a leading question you are correct. We are putting an idea into their head about what we think the perfect diet is before giving them a change to make their own unaided response.

There are however some good things here. The question is clearly stated. We want to know about your diet to lose weight. If your objective is not weight loss, you won’t answer it and affect my survey results.

The survey is also short, with just 5 questions and we can see this from the progress bar on the bottom. Although we are creating a few extra steps for the responded on having to click through, I understand that we might not want to influence their response to this questions with the following questions.

Here’s the second of the five questions:


In this question we have bigger problems than the first.

I don’t like the idea of including examples in the question and this is even worse than the first because it gives multiple example that really influenced the results.

If you look at the responses many people just selected from this short list as their answer and moved on.

We miss the richness of having the response in their own words by putting these examples in.

How could we fix this question?

If you have a question where you want to know a finite amount of information or have some ideas to test, this question is best served as multiple choice.

We’ve reached question 3.


This is another question where I see BIG problems.

The first being with the use of capitalization. We are leading the responded to an implication that they have a big problem to resolve and the response needs to be extravagant. Which may not be the case for this person.

And again the example in the question are leading and we have many respondents just choosing these and moving on.

And lastly, having three open field questions in a row can make the respondent tired and less likely to want to continue with the survey. Try to use variety of questions if you can achieve the same result.

On to question 4…


Take a minute more to think about what really makes this question difficult, besides the inclusion of example, which in this case are okay and not overly leading.

If you said that the goal of the question is actually to get quantitative financial / numerical data and this is qualitative question, you are correct.

When I first saw this question, knowing that it is an open field doesn’t lead me to want to write a concise answer or even think about numbers.

And the question would be shorter without the example that shows that you want numerical data. Shorter surveys almost always get more responses, especially when you are not offering a benefit to answer the survey.

We know that they are a limited number of things that we really want to know about with this question, so let’s just ask them. How much do you spend on the gym? How much do you spend on a nutritionist? Give them the option to leave the question blank if it doesn’t apply to them and put a 0 if they have a service and is offered for free.

In this question we don’t lose anything by putting the words here for them. And if there are many other choices that they might want to put here, we can always give them the option to select “other” and fill in what is the other expense.


If you’re thinking, there can be anything wrong here, we’re just asking for their name and email address.

Think again.

We just captured a nice amount of qualitative data on this person, but really we know nothing about them. Nothing useful to segment them. Nothing useful to understand a bit more about this person’s life.

If you are going to ask demographic questions at the end of the survey, make sure that they also have an objective.

In this survey it might have been nice to know the gender of this person, their age, their profession. This would really help when you are starting to work on your targeting.

If we are making a product that is pricey and would be better geared toward 50 – 65-year-old men but the respondents of our survey are 18 – 24-year-old women, maybe we are using the wrong language to speak to this target.

But, we don’t know that from this survey.

That survey could even tell us, well you thought 50 – 65-year-old men were interested in this product, but really you should rethink that targeting, there is a whole new group of people who you should be reaching.


I hope that this article was helpful in giving you ideas to improve your market research surveys. There is some much more rich information on this topic that I can provide, so I’ll update this post periodically to flesh out some more details on improving how you capture data and use it in your business.

2 thoughts on “How to conduct better market research surveys

  1. Pingback: Semana Summary (12 February – 18 February) | Repeat One, Repeat All

  2. Pingback: How to write a marketing positioning statement? | Repeat One, Repeat All

Leave a Reply

This site uses Akismet to reduce spam. Learn how your comment data is processed.