Turning research patterns into behavioral archetypes

CHALLENGE

The Centers for Medicare & Medicaid Services’ Quality Payment Program serves a wide range of healthcare professionals and organizations. The program already had personas that described differences between organizations, but they didn’t fully explain the people working within them.

​

Two people in similar roles or practice settings could approach the program very differently. One might plan months ahead and feel confident making decisions, while another might be overwhelmed, short on time, and focused on getting through the next requirement.

​

We needed a better way to talk about those differences and what they meant for the experience.

​

PROJECT TIMELINE

March - April 2025

​

MY ROLE

I led the research and development of the archetype framework. I synthesized findings from prior studies, used AI-assisted analysis to help surface patterns across the research, and then reviewed and refined those patterns using my own knowledge of the participants and research context. I also worked with stakeholders to make the framework clearer and more useful.​

​

RESEARCH METHODS

Cross-study synthesis, thematic analysis, behavioral pattern mapping, stakeholder collaboration, and framework development​

​

 

BACKGROUND

The Quality Payment Program already had organization-level personas based on characteristics such as practice size, reporting structure, and available resources.

​

They were useful, but they couldn’t always explain why people working in similar environments behaved so differently.

​

Across our research, I kept seeing those differences. Some participants were strategic and highly engaged. Some knew the program well but simply didn’t have much time for it. Others depended heavily on colleagues or outside vendors. And some questioned the value of participating at all.

​

Rather than treating those differences as noise from one study to the next, I wanted to see what would happen if we looked at the research as a whole.

​

The archetypes project grew out of that question.

​

 

RESEARCH QUESTIONS

• What behavioral patterns showed up repeatedly across our research?

• What seemed to shape the way people approached participation?

• Which differences meaningfully changed what people needed from the program?

• How could we describe those patterns without turning them into rigid personas?

• Could the framework give teams a more useful way to talk about participant needs?

 

 

APPROACH

Reviewing the existing research

​

I started by revisiting research from across the program rather than treating each study as a separate body of work.

​

I wasn’t especially interested in grouping people by job title, practice type, or demographics. Instead, I looked for differences in things like motivation, confidence, available time, reliance on outside support, willingness to ask for help, and how much planning people did.

​

That shift made it easier to see connections between participants who might look very different on paper but approached the program in similar ways.

​

Using AI to surface patterns

Because I was working across a large amount of existing qualitative research, I used Dovetail’s AI features to help with an early round of synthesis.

​

It was useful for the work AI is particularly good at: sorting information, grouping related quotes, and suggesting patterns I might want to investigate.

But those groupings were a starting point, not the answer.

​

I compared them against my own notes and knowledge of the research, moved quotes when the AI misunderstood their context, challenged categories that felt too broad, and renamed or reshaped patterns when the underlying behavior told a different story.

​

The process helped me move through a large body of research more efficiently while still keeping interpretation and judgment firmly in my hands.

 

Building the framework

From there, I grouped related behaviors, needs, and attitudes and looked for combinations that appeared consistently enough to represent recognizable approaches to participation.

​

The goal wasn’t to decide that someone was one archetype.

​

I wanted the framework to help explain why two people in similar circumstances might need very different kinds of support.

​

That process initially produced nine archetypes spanning different levels of engagement, confidence, capacity, and experience.

​

For each archetype, I documented things like what motivated the person, how they tended to make decisions, where they struggled, what kind of support they relied on, and what might help them move forward.

​

I also grouped related archetypes together so the framework would be easier for teams to understand and use without having to memorize nine separate profiles.

​

 

Refining the work with stakeholders

I shared the emerging framework with colleagues and stakeholders and asked a pretty simple question: Do you recognize these people?

​

Those conversations helped me see where archetypes felt meaningfully different, where the boundaries were fuzzy, and where our language needed work.

​

They also shifted the project from an interesting research artifact into something teams could actually use when talking about content, tools, and participant support.

 

REPRESENTATIVE ARCHETYPES

For this case study, I’m showing three simplified examples from the larger framework.

The time-sensitive participant

Wants to meet program requirements but has limited time and capacity. Often focuses on the most immediate task and may depend heavily on reminders, colleagues, or outside support.

​

The experienced strategist

Understands the program well and approaches participation deliberately. Looks beyond basic compliance and uses available information to make informed, longer-term decisions.
 

The reluctant participant

Feels disconnected from the value of the program or questions whether participation is worth the effort. May engage only when required and is less likely to seek out additional guidance.

 

Archetypes and visuals have been simplified and recreated for this case study. No participant data or original client materials are included.

 

 

OUTCOME

The archetypes gave the team a more human way to talk about the people participating in the program.

​

Instead of assuming that people in similar roles or organizations would behave alike, we could talk about the very real differences in confidence, motivation, capacity, and support that shaped their experience.

​

The framework added another layer to the existing organization-level personas and gave teams shared language for patterns we had been seeing across studies. It also opened up better conversations about whether the same guidance, tools, or support would work equally well for everyone.

​

Just as importantly, the project turned research that already existed into something new. Looking across studies allowed us to see patterns that would have been difficult to notice when each project was considered on its own.

 

 

REFLECTION

What I like most about this project is that the interesting part wasn’t collecting new data. It was looking at research we already had with fresh eyes.

​

Individual studies had given us answers to individual questions. Putting them next to one another told us something bigger about the people using the program.

​

It also gave me a useful early lesson in working with AI as part of research synthesis. AI helped me get through a large amount of qualitative material and gave me patterns to react to, but reacting to them was the important part. I still had to know when a quote didn’t belong, when a category was flattening an important difference, or when something technically similar meant something very different in context.

​

And I became even more cautious about treating archetypes as boxes. Someone could be highly strategic during one part of the program and completely overwhelmed during another. The value of the framework wasn’t in labeling people. It was in giving us a better way to ask, “What might this person need from us right now?”