How are segments formed

By Rob Couvillon
Vice President of Research and Analytics, Twenty-Ten Inc.

How are consumer segments created? What information do we use? How do we find the right similarities – and differences – to define the segments?

Research behind the segments

Before we can create segments, we first need to understand how consumers think, feel, and act. To do this, we are guided both by our own experience in the health finance sector and by interviews with key industry insiders.

Assisted by that information, we compile a mutually exclusive, collectively exhaustive list of all the needs, attitudes and normative behaviors that define the “key drivers” of brand choice and product usage in the health and finance industries.

This so-called “NAB” list becomes the source of the statements that we will submit to respondents in a national survey of a representative sample of 18-85-year-old health-insured consumers.

Our 2019 Persona^tm research

Next we conduct a survey which for our most recent Persona^tm research was answered by over 8000 consumers with health insurance.

The survey included detailed questions about respondents’ current health coverage, satisfaction with their current plan, how they manage their medical costs, and how they think and feel about their health finances. We also probed consumers’ attitudes and experiences in general finance, health, well being, overall lifestyle and values.

Cluster Analysis

We built our segments by analyzing the survey responses using a statistical technique called cluster analysis. Cluster analysis creates groups of people called segments that are based on how similar and different they are to one another on a specific set of criteria. When using cluster analysis, all the members of a given segment will be like one another on the chosen criteria, but different from members of other segments.

Criteria selection

Choosing the right criteria for segmentation is critical. We are trying to reveal the most extreme differences we can find between segments. For instance, a criterion that differentiates consumers who want precise, day-to-day control of their retirement funds vs. those who want to “set it and forget it” is useful for building financial wellness strategies for a group of employees.

Specifically, in the formation of the Persona^tm segments, many criteria were tested to optimize the result of the cluster analysis. These criteria were drawn from the NAB list which contained all the possible criteria that drive behavior in health and finance (“key drivers”).

Through a series of runs of the clustering analysis, several segmentation solutions (i.e. differing number of segments, varying choice of criteria) were identified in what is called a “champion-challenger” process.

The ultimate Persona^tm solution had seven segments that we are satisfied best represent the attitude and behavior drivers in the health and finance industries.  

Sign up for our mailing list to learn more

Ready to learn more? We are adding research insights and case studies to our blog. To receive them, please sign up for our mailing list here or reach out to us here