Blog

Three Ways to Make Segmentation Work in Innovation Research

By John Burns

Marketing departments the world over have too many dusty binders on out-of-the-way shelves containing segmentation research nobody ever used. This is a shame. Segmentation helps you target the right consumers with the right product innovations. The good news is that segmentation research is more than 50 years old, ample time to have developed best practices you can follow to help take segmentation off the shelf and put it into the hands of marketers.

CONDUCT THE RIGHT KIND OF SEGMENTATION

Making Segmentation Work in Innovation ResearchSeveral varieties of segmentation exist. These include needs-based segmentation which can encompass occasions, as well as, attitudinal (or psychographic) and behavioral segmentation. All are effective, but for innovation, needs-based is typically the better choice. Done correctly, innovation research identifies and prioritizes a comprehensive list of consumer needs. Therefore, segmenting consumers on these needs is logical and has several benefits:

  • Identifying similarities and differences in the needs that are important to each segment
  • Establishing the business value or opportunity of each needs-based segment
  • Guiding your innovation efforts to target needs relevant to segments with the most opportunity

AVOID THE PERILS OF A-PRIORI SEGMENTATION

Companies often conduct a-priori segmentation, where they select the characteristics they will use to segment customers prior to conducting the research. These characteristics typically include demographics and consumer life stages. Consumer needs, however, are far more complex than demographics or life stage can explain. Consumer needs emerge from a complex interplay between real-world situations and human psychology. For innovation research, only needs-based segmentations, arrived at using appropriate multivariate statistical techniques, can identify real segments (e.g., Convergent Cluster Analysis, Latent Class, etc.). These segments frequently skew more heavily to one or another demographic or life stage segment, but rarely are segments totally or even mostly defined by these characteristics.

BUILD DATABASE BRIDGES

Unfortunately, your marketing or retailer databases won’t contain the survey data necessary to classify all your consumers into one of the needs-based segments created in your survey. Fortunately, a solution exists for bridging the survey and databases.

  1. Prior to launching your survey, look at your database and identify demographic, life stage and any other data you deem relevant for characterizing consumers. Use logic and judgment here.
  2. Turn these items into questions and include them in your survey. Ask these questions in a format that exactly matches the way they are in your marketing or retailer databases.
  3. Once your needs-based segmentation is complete, use the questions that match back to your databases in a multivariate analysis to predict which segment each individual belongs to. In statistical terms, the questions that match back to your databases are independent variables; the segments you created from your survey are dependent variables. The analysis will produce an algorithm, commonly referred to as a typing tool, which classifies (i.e. “types”) respondents into one of your segments. Since this algorithm only contains questions that exactly match your database, you have built a bridge between your segmentation and your database. Prediction accuracy of 70% or higher is common.
  4. Run this typing tool on your marketing database to classify consumers or shoppers into your needs-based segments. In addition, you can use your typing tool to classify prospects for your sales force, visitors to your website and potential respondents in research projects.

Three Ways to Make Segmentation Work in Innovation ResearchYou might ask: If we shouldn’t use demographics and life stages to create needs-based segments, how can we use them to predict these segments? The answer is that you’ve already created your needs-based segments using appropriate statistical techniques. Rather than creating segments using demographics and life stages, you’re using them to predict the relatively small number of needs-based segments you’ve already identified.

MAKING SEGMENTATION WORK FOR INNOVATION RESEARCH

Segmentation can help you target the right consumers with the right product innovations. Address these three issues, on your own or with a qualified research partner, and it will make segmentation a useful complement to your innovation research.

Market Segmentation

Share this: