5 Market Research Questions Every Executive Must Ask!

The Problem With Market Research

Market research should be a required course for everyone in business school, not just marketing students. Many decision makers across functions use it but few have a clue as to whether the information they’ve been given is valid. As a result, despite good intentions, hard work and financial cost, executives often make important decisions based on poor information.

To put this in perspective, market research is core to providing the basis for developing an annual marketing plan and adapting marketing strategies throughout the year.  As result, I examine and analyze a lot of research in the course of the year. While I’m surprised at the number of articles and blog posts that don’t offer further insights beyond rehashing the basic results reported in the related press release, many market research studies have some form of bias or error built into the data.

5 Issues with most market research

As a starting point, every senior manager responsible for fielding a market research survey should ask these five questions. You shouldn’t allow the survey to be fielded until you receive sufficient answers to them.

  1. Is your survey group statistically valid? Are your respondents a representative group? Did you select your participants from a larger population using a random selection process? Putting a quick survey on your website doesn’t qualify. If you have trouble getting a target audience, this is a strong indicator that your research and related marketing will be challenged. In other words, if you can’t target survey participants, how will you target your marketing?
  2. Is your questionnaire clear and unbiased? Do your questions make underlying assumptions about respondents and their answers? Does the survey give respondents a full array of options? For example, the question shouldn’t offer only positive responses such as do you agree, somewhat agree or strongly agree because these answers don’t allow participants to disagree.
  3. Are the categories in your questions mutually exclusive? The goal is to ensure that all of your respondents understand the references in your questionnaire. In addition, since each additional question has an impact on whether people will respond (as well as the cost), determine whether the information it obtains will help you make your decision? For example, a former client fielded research by a top research firm that asked if the respondent was an executive assistant, secretary, administrative assistant or other. While these job titles may vary by company, this question should have been reworded to describe specific job functions to yield usable results.
  4. Are your answers completely exhaustive for each question? Have you covered all possible responses to your question? If not, what’s missing. Not sure, include “other” as an option and ask respondents to fill in what they mean. To help avoid this issue, have someone outside of your department or preferably your organization proof your survey. Don’t assume that your market research supplier will ensure there’re no issues.
  5. Is the survey information gathered in an  unbiased way? If you’re fielding a live survey or a phone survey have you trained survey takers? If not, they may just fill in the forms with whatever or interpret participants’ responses based on their own leanings.

Market research requires valid, unbiased input from a statistically valid sample of participants who’ve been chosen at random. Further, questions and answers must be clear leaving no room for respondent interpretation or your results will be skewed.

Are there any other questions regarding market research you’d add to this list? Have you had any market research issues you’d like to share? If so, please include your recommendations in the comment section.

Happy marketing,
Heidi Cohen


Here are some related articles you  may find of interest.

Photo credit: http://www.flickr.com/photos/marcobellucci/3534516458/

 

 

 

Tags . Bookmark the permalink.
  • Krithika

    What do you feel about targeting respondents through Facebook ads? Does that add bias to the sample? If not, why not?

  • Larry Burns

    As a 30+ year marketing researcher I applaud the notion of truly comprehending the biases and potential error in ANY data one consumes.

    To however suggest that “Market research requires valid, unbiased input from a statistically valid sample of participants who’ve been chosen at random” is to live in a world of text books. If you are truly doing a MARKET study – i.e. trying to size a market or really needing a random sample I’d agree – but aside from political polling (which is killing us in other ways) very, very, few research programs can truly do random anymore. Why? Humans simply are unwilling, in ever increasing numbers, to offer their opinions in ways that really allows true ‘random’ samples to be drawn – even if you set out to do so. So, aside from not wanting to toss aside anything that is not based on pure random samples I think you make terrific points – questionnaire design is in fact an art, and really it is far to easy to create “bad” questions that you do not realize were bad until after the fact. Take you time in data collection instrument creation, you will be far happier when the results come back! So, can we agree that it is possible to learn valuable things in a non-random way – you just need to know that a particular piece of data is or is not effectively ‘project-able’ to anything ….

    However, my point was not to call your notion of taking great care to ask about representation into question. To me, the key point you make that I am in 1000% agreement with is to KNOW what you are reading, know what the data is, know how the data is being turned into information or even with some luck and skill into knowledge. It is far too easy to quote “statistics” and percentages to make a point while sounding wise and sage but, upon even superficial probing you are often left with a total ‘foundation of feathers’ upon which someones ego is attempting to construct a temple. We desperately need good marketing research, we need context to help us understand the impact of decisions we are considering – but most importantly we all need to THINK.

    Education used to be about learning how to critically question – we could all use a refresher course and your post does a great job of reminding us !! It is far too easy to be ‘algorithm-ed’ into submission – and nothing replaces the stuff between our ears, question, dig, and use the research you execute as a merely one piece of your context and understanding. Now climbing down off his soap-box …. thanks.

  • http://www.c2hm3.com Christina McCale

    Lovely List, Heidi!

    I would add “What do you really want to know?” and “how are you going to use the data?”

    I have found that by formulating the research objective in the form of a question (ie: “Why do customers never come back to purchase again?”) helps narrow down what it is that we are researching.

    Second: I would add “Can you answer the question in any other way other than doing original research?” There are some firms who want to go out and poll or study everything. And while that’s a nice idea, sometimes we just don’t have the time or money to invest in original research every time we want to. So it’s our fiduciary responsibility to the firm to make sure we are using our marketing dollars wisely. If we can get a good answer to the question through internal data, existing customer records, true-d up sales figures, install records, etc., then we should do so. If a solid environmental scan by using existing market data on the web will give us reasonable answers, then use it.

    Great columns -keep it up!

  • http://www.samplequestionnaire.com/market-research-questionnaire.html Market Research Questionnaire

    We do the market research to know the exact opinion of product or service. If question is not clear we can’t expect useful data… so questionnaire is more important in market research.