Correct me if I’m wrong, but does the market research community seem to be obsessed with the “how” while virtually ignoring the “what”? Look at any list of topics at a conference, or the contributions to Twitter and LinkedIn, and the focus is clearly on methodology and process. There is rarely a mention of the specific market to which the methodology may or may not apply. Yes, there may be case studies, edited with the benefit of 20/20 hindsight, but not much about the nuances of specific markets beyond the usual advice to define goals before selecting a methodology.
Why is this? I’m guessing that contributors to conferences and social media want to appeal to as large an audience as possible. And by focusing on methodologies which have broad applications, they are maximizing their potential audience. Similarly, large agencies want to sell their wares to clients in a number of diverse industries rather than chasing the long tail of niche markets. Better to talk generally about approaches that can be applied widely rather than tailor the message to the market.
I think this obsession with process may be at the root of some clients’ dissatisfaction with suppliers. I was a client myself for a number of years and still have candid discussions with client-side researchers. What seems to be their #1 peeve is that suppliers often lack deep knowledge of the client’s business. Admittedly it takes time and effort to become familiar with a specific market, but clients are more than willing to share their knowledge, even to the point of providing access to earlier research if that leads to better, more appropriate research in the future.
Let me give you a specific example of when research was driven by the “what” rather than the “how”.
I began my career in research with Quaker. Many of my projects were the simple two-dimensional affairs common to CPG (consumer packaged goods) companies with large brands. Concept tests, taste tests, packaging tests, in-home usage tests, etc. All of them were pretty easy to design – A vs B, B vs C, A vs. C – usually a limited number of options – be they brands, colors, flankers, whatever. Frankly, I found these pretty boring to design, but enjoyed playing with the results as an analyst.
In the 70s, several of us left Quaker and other big CPG companies to join Sears. Sears was building a market research department from scratch, hiring in the most experienced people from outside the retail industry. Back then, market research was CPG dominated and pretty much all training was done on the job rather than in academic institutions. You learned the business at P&G, General Mills, Quaker, General Foods, and other large companies, and this training was a stepping stone to more opportunities. Suppliers were still quite small and nothing like the conglomerates they have become today.
And so, along with many other researchers, I moved from a CPG environment, where research design was standardized to fit a world of limited choices, to a world where I had to completely rethink the “how”. And to do this, I first had to understand the “what”.
One project comes to mind of the hundreds I conducted at Sears. I was tasked with helping the buyers in the Home Fashions departments decide what to include in their lines for the upcoming season. One category was bedspreads. Now, there is no way I could run a conventional product test. The issue was far too complex. Of 300 possible choices of potential bedspreads to include in the catalog and the stores, what 100 or so should make the cut. Sears was the largest retailer at the time and the largest employer in the US. Millions of dollars were at stake, and the findings would determine which factories in which towns (still mostly in the US!) would get the orders. And the buyers were counting on the research to help them make career changing decisions.
Obviously, a new type of study was needed. First, I conducted focus groups to understand the process women went through when picking a bedspread. I was looking to develop a hierarchy of decision points to guide in designing a survey. From this I learned that when narrowing choices, first came the end use (master, girl’s room, boy’s room, guest), then price, then the type (quilted, chenille, etc), then color and style. We are now very comfortable with the concept of winnowing down choices from a large pool (think automobile shopping sites), but this was long before the internet.
I wanted the test to replicate as closely as possible the shopping experience for bedspreads. For this, I went to five malls across the country and rented empty stores in each of them. (Though since Sears owned the real estate “rented” meant free.) In each store, we set up displays of over 300 bedspreads using photographs and large swatches. The displays were organized into sections, reflecting the decision hierarchy we uncovered in the groups. Respondents were intercepted, screened for intention to buy, then free to shop in the appropriate section or sections. For instance, someone who indicated purchase interest in a bedspread for the master bedroom was allocated to a price group based on her willingness to pay (low, medium and high), then free to make a selection. There was more to it than that, and the results were weighted. But the point is the study design was flexible and the focus was on the characteristics of the market rather than the methodology.
The buyers used the results to make their buys, and that year bedspread sales increased 29%. This approach was adopted with similar results for other markets, including clothing and other soft home fashions, where there was a need to optimize an interrelated set rather than a single entity like a brand. Old school buyers who had previously been reluctant, even scornful of research came on board. Happy ending, though not for Sears. They never recovered after I left!
Ok, so this was a long time ago and now we have more sophisticated tools at our disposal. But I still believe that any research should begin with a clear understanding of the specific product involved. Start there, do your homework, get to know the market, define the problem, and only then think about methodology. And if something hasn’t been tried before, be innovative. In the end, no one really cares how you did it as long as you provide information that is actionable and beneficial.
Image is from the cover of a free ebook: What Matters Now.