Back to top

By Bob Relihan

Last year I discussed why it may be wise to employ qualitative methods after a quantitative investigation. This approach, of course, reverses the traditional order of such things. Many of us were taught to use in-depth interviews or focus groups to generate hypotheses for quantitative testing. But, as product development cycles get shortened it is tempting to short-circuit the initial qualitative phase of a research program and move straight to quantification. Marketers, after all, know their category and their consumers. Right?

Here are 3 reasons why early qualitative research will make the design and analysis of quantitative projects smoother and more precise.

  • Consumers really do define words in unexpected ways. I recently completed a series of individual interviews of customers who purchased a particular product. This product was available as either a "custom" item or "off-the-shelf." I recruited customers of each product type for my interviews. To my surprise, many of those I thought would be "custom" buyers turned out to purchase non-custom products. What happened? It turned out that there were so many varieties of off-the-shelf products, customers often printed the names of the items they purchased. The product was just what they needed. It was "custom." If I had used my screener to fill quotas of "custom" and "off-the-shelf" users for a large scale survey of product attitudes and usage, my results would have been incomprehensible or, worse yet, clear but wrong.

  • Consumers really do view categories differently. I was exploring product usage in a category that the brand managers divided into four groups --premium brands, value brands, bargain brands, and low-price brands. Each of these categories represented a different price point and value proposition. Consumers just didn't see the category that way. Any product in the category was either a "name brand" or a "generic" --each of which had very different price expectations. If we looked at the results of a brand image survey based upon the four slices of the category marketers saw, we would probably have confused results that would be more clear viewed through the two-segment lens consumers used.

  • Consumers really classify products in unusual ways. From the perspective of brand managers, a category is often defined by attributes of production and distribution that may be irrelevant to consumers. Designing a quantitative study based on these assumptions can be misleading. For example, "salty snacks" is a category that makes sense to marketers because it corresponds to how products are distributed and displayed in stores. It makes some sense to consumers, but it is not perfect. When I have explored salty snack qualitatively with consumers, I always make sure that we consider candy, cookies, frozen single-serve pizzas, apples -- anything that can is consumed at a non-meal occasion. The brand manager may say these aren't in my category. But what happens in the interview? We often discover that some salty snacks --pretzels, for example -- have as much in common with chocolates as they do with other salty snacks. And, these associations reveal a range of relevant attributes we would not have seen if we had limited our exploration to just "salty snacks."


So, if you want to have maximum confidence in you quantitative instrument and in the results it produces, a qualitative first step is invaluable. You really do need to "measure twice and cut once." Moreover, it is much more efficient if your qualitative consultant actually works with your quantitative consultant, discussing the issues in depth rather than simply passing on a report of qualitative findings.