Read Time: 2 minutes

From Hypothesis About Segmentation Options

Use your qualitative research devise various ways in which we might segment users. In the end we should have a variety of candidates to analyze.

Collect Data Based on Segmentation Options Through Research

For each segmentation option, there are particular questions that we’ll need to address through using site traffic analysis. Maybe asking questions about how often and for how long does the user use the existing solution? The goal at this point is to gather more data for the next step.

Segment Users Based on Statistical Cluster Analysis

An algorithm takes more of an active role in guiding the segmentation model, rather than just testing the existing assumptions. You could end up with a number of clusters and any number of attributes as key differences between each cluster. This step in the personas process is significantly different than other approaches because the segmentation is data driven as well as human driven.

Create a Persona for Each Segment

When the cluster analysis finishes the segments, we need to take the data and make it real, like in the qualitative research process.

The Pros of Using a Qualitative Research Process

• Quantitative techniques, in combination with human intelligence, generate the personas. A human bias influences this approach less than the others. To a skeptical client, this scientific process is more likely to dispel any doubts they might have about said persona as a valid research tool.
• The iterative nature finds the best solution. Using the cluster data, we might find surprising patterns in the data that no one else may have considered. This could lead to better insights about their users and better customer satisfaction.
• More variables can be examined.

The Cons of Using a Qualitative Research Process

• The quantitative research requires more effort on our part.
• We’ll need to have more specialized skillsets. I’ll need to bring statistical data analysis to the table and working with automated analysis processes try and sort out all of these segments to make sense of it all.
• The results could introduce new ways of thinking that the client may be uncomfortable with.

If you had the time and money to invest in this type of research, more power to you. I would suggest this route if you were only doing usability testing for a client or the client request to review our data that you collected about the solution and audience. You could take more time to explore multiple segmentation models to find the right one and our personas might be driven by multiple variables. This would be good if you didn’t understand which segmentation model is the most important to the client and or solution.

Learn more about quantitative research at these links:

• https://www.simplypsychology.org/qualitative-quantitative.html
• https://libguides.usc.edu/writingguide/quantitative