You cannot always rely on the standard market segments as presented before. Sometimes, you will need to find ways of identifying distinctive segments among a large population. Rather than dictate the terms of your segmentation strategy, you would want the facts to speak for itself. This is where a statistically-based method of segmenting the market can be utilized, something that large organizations use.
Step 1. Conduct a wide survey. Using a survey methodology, have a large group of respondents (in the thousands) identify the different product attributes and their importance ratings for a selected industry; brand awareness and brand ratings; product usage patterns; attitudes toward the product category; and demographic, psychographic, and mediagraphic (i.e., media, such as television or newspapers, that in respondents most commonly view or avail of) profiles of the market.
Step 2. Process the data. Using the statistical tool of cluster analysis, where respondents are grouped together based on similarities of their answers, the respondents are grouped into a given number of clusters. This is done automatically by statistical software. The research& only needs to specify how many clusters to sort the respondents into. This is the easy part.
Step 3. Profiling. This is the hard part. While the software can classify and categorize respondents into clusters, it generally cannot describe what each cluster is composed of or why respondents were clustered together the way they were. This is where the analytical skills of the researcher come in. By scanning the list of respondents who were placed under a particular cluster and by scanning their answers, the researcher can make qualitative conclusions as to what the commonalities are among the respondents in this cluster and then gives it a name. For instance, upon realizing that a significant number of the respondents in the cluster happen to listen to talk radio while driving to work and that most of them are in middle management, the researcher might name this cluster the “Talky Managers” (note: you probably can do a better job of naming this group!). The clusters, as profiled, officially become market segments.
For smaller firms, a segmentation study such as the above may be too expensive to undertake. In that case, the traditional segmentation variables become useful, particularly if they are mixed and matched. For example, a firm can look into the two variables of lifecycle stage and social class to establish the possible segments of “young mass market,” “young AB,” “teen mass market,” and so on.
A popular way of segmenting the market is through benefit segmentation, which is how many fast-moving consumer goods are segmented. In benefit segmentation, attention is focused on the needs that people may have for a particular product rather than the types of people who may be targeted. For example, for toothpastes, certain people may have a need for cavity prevention while others need fresher breath and still others have a need for great taste. Thus, segmentation becomes focused on the benefit rather than on the consumers’ demographic (and other) profiles. The advantage here is that benefits may appeal to people of different ages, social classes, etc., thereby giving the product the potential to have a larger market base than otherwise.