Market Researchers Should Try to be More Like Galileo

If you have this feeling that there’s a better way to set up your research, think about how you might try this out:  treat each study like it’s scientific inquiry.  This shouldn’t be hard; market research is supposed to be scientific inquiry.  It’s a perfect fit.  However, after 20 years in the business, I’ve come to the obvious, in hindsight, conclusion that too much research is done “just because” with no direction.

Some Frameworks for Inspiration

The proposed approach is inspired by classic scientific method and the CRISP-DM framework used in machine learning:

scientific method.png

Cross Industry Standard Process for Data Mining, commonly known by its acronym CRISP-DM

CRISP-DM.jpg

While these two processes serve different purposes in different domains, they do have some important similarities.  A key one is that both include preparation before any “work” is done.  Another is the feedback – continuously learning from work that was already done.  Market research needs more of that.

The Market Research Process

I recommend the following approach:

1.      Identify the Business Issue the research is supposed to address – based on operational data, tracking surveys, previous research, sales, competitor actions, economic events. For example: Prices may be too high or low, new product launch sales are disappointing, channels require fewer SKUs, existing campaign isn’t “working” –  need a new one

2.      Develop Hypotheses that are testable, with action plans as logical outcomes

a.      must have prior action standards

b.      for example: “what will be changed if result X, Y, or Z?”

3.      Do Appropriate Experiment(s)

a.      qual or quant or mix

b.      sample (rep or some other criterion) including size

c.      SHOULD CREATE REPORT STRUCTURE (less data and recommendations) BEFORE RESEARCH BEGINS – this is a post of its own

4.      Reporting with Recommended Actions

5.      Explicit Feedback to use for informing business issues

MR+process.jpg

I hope I was able to get you to think about the overall market research process and on the way, whet your appetite for more details (to come) on each of the five steps.  If you have other ideas or want to build on these, please comment below.

Brett MathesonComment