News/Views: Ask the Researchers
Send us your questions about research and marketing and VMI executives will respond by posting responses on this VMI Website. E-mail your questions to info@vincentmccabe.com. If you’d like our CEO Jean Vincent to respond to your question personally, address your request to jean@vincentmccabe.com.
To get more details about Vincent McCabe, Inc., its operations or free offers go to www.vmistudy.com/details
Recent Posts:
Q. We’re a small manufacturer of medical products and want to study a new product we’re developing. What is the smallest sample size we can use and still get good information?
A. This is a tricky question and one that comes up a lot. Much will depend on where you are in your investigation. If you are just doing preliminary evaluation, you may need to use qualitative research. In this case, the sample sizes are quite small – 6 to 10 individuals in four or five groups. Qualitative research is descriptive not predictive. The sampling methodology would also typically be a “judgment” sample of people who are exemplary of the audience you have defined. If, on the other hand, you want to predict how many people will buy your new product, you are looking at quantitative research and larger samples. The sample size is not a function of the number of people in the market but rather a function of how diverse the people in the market are. If the degree of concurrence is very high, your sample size can be smaller. If, for example, your market has 90,000 people (the approximate size of the US market for primary care doctors) and you wanted your survey responses to be 95% confident +/- 5% (that’s industry average), you would need a sample of approximately 383 people, assuming some variability (50%) in the base. If you did a survey with only 35 respondents from that same audience and those people agreed 90% of the time, you will achieve a 95% confidence level +/-10% - this may be adequate for your purposes. If the respondents agree only 50% of the time, you’d only achieve a 95% confidence level +/-17%. That’s a pretty big swing. If you are looking for directional advice, a small sample may be adequate for your purposes. On the other hand, if you’re making a major or expensive decision, you’ll want more precision in your predictions. The bottom line – there are many variables that affect a sample size and sampling method. It’s best to have a professional help you determine what will be optimum for you on a case by case basis. JV
Q. Shouldn’t we be able to make good marketing decisions without a lot of research? I’ve been in the business for years and I think I know what’s best. What will research do for me that I don’t already know?
A. I’m a big believer in “gut” level decision-making. I thought Malcolm Gladwell’s “Blink” was well done and convincing. Too much analysis can be a bad thing. However, as a researcher, I argue for getting a more informed gut so you can make those rapid decisions that are needed in today’s fast-paced competitive environment. After hundreds of Voice of the Customer studies, I can honestly say that something new is always learned through formal research. There are some things that are very difficult for people inside a company to learn. There are lots of psychological reasons that explain this. Self preservation is one big reason. People act in self-interest. I don’t know of many customers that will tell you that they secretly think your prices are too low or that they don’t like your salesman. Next is multidimensional complexity. The greater portion of consumer decision-making is subconscious – they simply don’t consciously know the answer to why they buy. So how can they tell you? It’s only through statistical analysis that those subconscious beliefs can be exposed. Up next is cognitive dissonance. That’s Leon Festinger’s theory that a human mind must be consonant – in agreement with itself– i.e. if I think I’m a person that should read best sellers, I will tell you that I read best sellers, when indeed I may not have taken time in reality. These are just a few of the problems. Research can clear away these situations that cloud the issues. Through the use of good questionnaire design and advanced statistical analysis, researchers can dig deeper and get to the truth of the matter. Take a look at www.volunteer20.com to see how research can assist in discovering motivation and market segmentation. JV
Q. I’m considering investing in one of those do-it-yourself online surveying software tools, do you think this is a good investment?
A. It depends. If you are a professional or near-professional researcher and have studied questionnaire design and sampling techniques, these software tools for non-professionals might be helpful to you to implement some quick polls. On the other hand, if this is your first design and you’re doing this to save money, it may cost you more in the long run or just be a monkey on your back that you won’t use as you thought you would. Actually, in my opinion, “no research” is actually better than “bad research” – at least it presents an unknown rather than incorrect information. Before you sign up for monthly usage fees or outright buy these programs, talk to a researcher to see if there’s a better, maybe less costly way. If it makes sense, we would tell you. Researchers are consultants that want what’s best for you no matter the financial incentive. Hate to sound self-serving but this is a situation akin to the early days of desk-top publishing. A software tool won’t make you an artist overnight – skill and training are still required. Instrument design is harder than it looks. JV