Personas and Market Segmentation – Do They Really Have to Conflict?


Those who work in research may have encountered a struggle between two different views of the customer. The first, market segmentation, groups people into distinct categories based on particular attributes or psychographic characteristics to determine market sizing or potential and is typically developed through quantitative methods like cluster analysis. The second, personas, are a set of customer archetypes that demonstrate typical motivations, behaviors and needs of various customer types and is typically developed through qualitative research methods.

While both provide an important view of the customer, they often conflict with one another—either literally containing conflicting viewpoints or simply causing conflicts between stakeholders who may prefer one over the other. This typically results in the usual corporate turf war and a lot of talking past one another—which benefits no one and certainly not the customer.

Many argue that segmentations and personas serve distinct purposes and thus should be kept separately. While this may be true in an ideal sense, in practice, in my experience, it typically leads to confusion about the customer among stakeholders and wars over who the customer really is what he/she really wants. In practical terms, it is better to have one overarching viewpoint on the customer and the groups of customers that exist for your product or service. If the customer truly is the same, this should, of course, be possible.  

In our practice, to get over the conflicts between personas and segmentation that we have witnessed all too often, we bring together the best of both worlds to create a hybrid approach. As is typical in the creation of personas, we begin with qualitative research to develop a set of mutually exclusive categories that define differences in customer types. From these attributes, we then build out personas, just as you typically would. However, we take things a step further by entering the mutually exclusive categories of data into a survey that can then be used to quantify how large the persona groupings actually are. Oftentimes people do not fit neatly into persona categories given that these are ideals, however, you can create cutoffs with the data in order to find a best fit.

By approaching persona building and segmentation this way, you can assure that both are speaking the same language (since they are built from the same categories) and you can strengthen both through the interaction of each. There are also side benefits such as the ability to develop a screener from the quantitative data that can used to recruit future research respondents who fit particular persona categories.

Greg VanderPol