Brands often develop personas as a way to keep customer characteristics top-of-mind as they consider communication and innovation efforts. In some situations, these personas are developed via qualitative research or internal knowledge/hypotheses about customers. When personas are identified in this way, one question we are often asked is ‘Now that we’ve defined these personas, can we build a typing tool to find and size them among our target audience?’

A key deliverable of any quantitative segmentation is a typing tool, which allows the user to identify segment membership. We are able to create typing tools because for every person in our sample, we know their segment and have all of their data. Given this data, various statistical techniques are used to determine an efficient subset of questions that most accurately assigns people into a segment.

For personas defined based on qualitative research or internal knowledge/hypotheses about customers, we don’t typically have the necessary data to apply these same statistical techniques. Personas may have some inherent quantitative characteristics (e.g., Group A is older and has fewer pain points, Group B is younger and experiences more pain points), yet they are often not as specifically defined as they would be in quantitative research (e.g., how much older, what specific pain points?).  And, they are not necessarily defined to account for all potential customers (e.g., where would someone who is older and has more pain points be classified?).  

To develop a reliable typing tool, a quantitative dataset is needed. In our experience, there are two ways to approach this request to develop a typing tool for personas via follow-up quantitative research:

1. Use the components of each persona to develop traditional ratings/bipolar scale questions that would be used in a quantitative segmentation survey, conduct a typical segmentation analysis, and see how the results align/diverge from the original personas. It’s important to be prepared to see segments that do not align exactly with the original personas. This occurs with some frequency as personas are not typically designed to identify mutually exclusive groups, but account for different aspects of attitudes/motivations. The good news is that persona work is typically a great foundation for segmentation survey development. In the end, the typing tool delivered will predict membership in the segments uncovered (which may not be the same as the original personas).
2. Develop a summary of each persona, show those in the survey, and have people self-select into one. It’s important to be prepared for certain personas to have small base sizes. This can occur because there are often universal truths that nearly everyone agrees with/feels are important and then there are niche items that differentiate between people – if there are personas aligning with those universal truths, respondents may self-select into those instead of a niche item. To avoid this we can ask respondents to first select all personas that apply and then have them rank the order. We then assign anyone who ranked a smaller/niche persona in their top two as being part of that persona. In terms of the typing tool, the self-select question asked in the survey IS the typing tool.

Using either of the approaches described above, the client team can feel confident that they have a method to identify their personas in future research.