Millions of data points. Five people worth knowing.

The starting point was the data itself. Claritas PRIZM segments consumers into dozens of distinct groups — each with its own demographic profile, behavioral patterns, and lifestyle characteristics. I went through the full dataset, sorting the segment files and looking for natural groupings that made sense in the context of Priority Health's audience.

The goal wasn't to use every segment — it was to find the clusters that actually reflected who Priority Health was talking to. As I sorted and grouped, patterns started emerging: segments that shared similar attitudes toward healthcare, similar communication preferences, similar barriers to engagement. Those patterns became the foundation for five persona groupings.

This phase was largely independent work — deep in spreadsheets and segment documentation — before bringing any of it to stakeholders for validation.

Data alone tells you what's there — it doesn't tell you what it means. Interviews were how I validated that the groupings I'd built from the PRIZM data matched the reality of who Priority Health was actually serving.

I conducted interviews with both internal stakeholders and real users. The stakeholder sessions helped surface how Priority Health's own teams thought about their audience — where they aligned with the data, and where their assumptions diverged from it. The user sessions were about understanding the people behind the segments: what they cared about, what made them hesitate, what they needed from their health insurance that they weren't getting.

Several of my initial groupings held up. A few needed adjustment. The interviews grounded the personas in something real rather than leaving them as a purely analytical output.

With validated personas in hand, the goal shifted: getting Priority Health's team to internalize the personas well enough to actually use them. I facilitated three distinct workshop formats, each designed to build a different layer of understanding.

Human analogy. I had participants assign each persona a job or occupation — not because the data said so, but because it forced them to think of each persona as a real person with a specific context, not a demographic bracket. The exercise consistently produced better instincts about communication tone and channel than any amount of data presentation had.

Abbreviated empathy mapping. For each persona, we brainstormed what they feel and what influences their decisions — specifically in relation to health insurance. This surfaced the emotional context behind the data: what makes someone avoid engaging, what builds trust, what language lands and what doesn't.

20/20 vision. The final workshop asked a direct question: how does Priority Health want to communicate with each of these people? Not how they currently communicate — how they want to. That distinction produced clear, opinionated communication priorities that went directly into the persona documentation.