Every marketer does research of some sort or other on a frequent basis. Whether it’s formal research designed to understand customer satisfaction or new market needs, or less formalized tasks such as gathering input from customer advisor councils, one of the key aspects of the marketing function is gathering information about the outside world to better understand and react to it.

Unfortunately all too often too many of us simply plunge in and start asking questions (or designing surveys) without spending as much time as we should thinking about how the information is going to be used, or even what the questions really are. Instead, we can take a page from our physicist and biologist friends and apply a little bit of the scientific method. Specifically I am talking about forming and testing hypotheses to guide our research.

In case you started zoning out during that last sentence all I really mean is taking an educated guess at what we think we are going to find at the end of the research. The hypotheses (or educated guesses) should be based on the best of your current knowledge, and should frame up the areas of inquiry you want to dig into.

For example, let’s say you are designing a customer satisfaction survey, and you believe that your customers are generally happy but there are one or two areas in which you believe your customer experience is lacking, let’s say in product reliability and service/support. You could form hypotheses along the lines of “while we are doing fairly well our lack of product reliability is causing a great deal of customer churn,” “customers don’t place much value on service/support so lack of performance there is not a problem,” or “the customers who are unhappiest with our service/support are our least profitable customers anyway so it makes sense for not to invest too heavily in it.”

Each of these hypotheses will lead to different aspects to dig into in the research and different questions to answer. For example, the first hypothesis connects customer satisfaction with certain service components to customer churn and will cause you to think about things like interviewing customers who recently left and understanding the reasons for their defection. The third hypothesis on the other hand may lead you to link the results of your research with information from your customer database to correlate the satisfaction findings with customer profitability metrics.

Applying this approach is likely to inform your research methodology and questions to be addressed. The results are likely to be more valuable to your organization, and more likely to touch upon critical business issues you need to resolve.

Your hypotheses don’t have to be complicated, or even terribly “unusual” or “insightful”. The key is that it forces you to think about critical business and marketing issues and how your research will be applied to answering them. And by doing it up front, you can construct your research to ensure it addresses the questions you really need answers to, so you don’t get stuck with the “oh, I wish I had asked that” syndrome quite so often.