How TDH Used New AI-Powered Research Methods to Reveal Deeper Brand Insights

Key takeaways

• TDH developed a hybrid AI research methodology that combines AI at scale with human strategy to uncover deeper audience insight.  

• One unexpected insight didn't just change the messaging, it redefined the role the brand needed to play.  

• While AI helped us move faster, human conversations validated and made findings more tangible.  

• In six weeks, those insights became robust target personas, messaging guidance and a more human brand strategy.

• The future of AI in brand strategy isn't replacing people. It's helping us understand people in new, more personal ways.  

When we began working with our healthcare client, we faced a challenge that’s familiar to many organizations: how do you deeply understand a complex patient population within real-world constraints on timing, budget, and access?

Our client’s target audience included patients across the continuum of a complex disease state, from stage 1 through end-stage disease. This is a population shaped by a feeling of the unknown, complicated care journeys, emotional fatigue, fragmented healthcare systems, often unstable access to care and resources, and highly individualized experiences.  

We needed insights with enough nuance to create highly distinct patient personas to guide our patient outreach across a wide spectrum of care journeys. The work had to go deeper than just surface-level demographics or generalized patient assumptions.

Traditional research could have helped. Eventually.

But we needed meaningful patient insight in weeks, not months.

So the question became: how do we get to the kind of deep human understanding the strategy required, without the timeline and budget traditional research often demands?

The answer? With a little help from AI.

And a lot of help from actual humans.

Building a Hybrid AI Research Model

We built an AI-assisted research approach designed to combine the scale of quantitative signal analysis with the depth of qualitative human conversation.

The goal wasn’t to automate strategy. In fact it was quite the opposite.  

Our real live, flesh-and-blood strategy team was deeply involved throughout the process, shaping the inputs, guiding the tools, interpreting the findings and refining the output. The AI helped us move faster and see more. The humans made sure what we saw actually meant something.

That human and AI partnership allowed us to accelerate understanding without losing the finely calibrated distinctions between patient personas.

The project combined two sequential AI-based methodologies.
‍

First, we used an AI-powered research platform to analyze patient and provider signals at scale.

Through this first AI methodology, we were able to analyze information across forums, social conversations, healthcare discussions, and other real-world interactions and available online sources. The platform helped identify patterns in behavior, emotional drivers, unmet needs, barriers, and communication preferences across disease stages and audiences.

Rather than focusing only on demographics or disease stage, we built patient narratives and personas grounded in lived experience and behaviors.  

The research revealed how patients made decisions, what shaped trust, and how emotional and practical challenges affected engagement and expectations with care.

From there, we moved into validation.
‍

The second methodology we used was an AI-moderated 1:1 interview technique, interviewing real patients. Real people, face to face.

We conducted AI-moderated interviews with real patients to pressure-test and deepen the initial findings. Our team developed the discussion framework, including objectives, prompts, probes, and thematic focus areas, while the platform handled moderation, transcription, and synthesis across interviews.

This second phase was critical.  

It allowed us to move beyond identifying patterns and begin understanding the emotional context behind them.

The combination proved especially powerful because it paired AI’s ability to synthesize information at scale with direct human conversation and strategist interpretation. We could listen closely to individual patient stories while rapidly identifying themes and nuances across the broader population.

The Insight That Changed the Strategy

One of the most important findings from the research fundamentally reframed how we thought about the audience: Patients were not necessarily disengaged from their care, but they were constrained from engaging in their care.

They felt constrained by layers of physical, emotional, and systemic barriers.  

The original assumption that patients were passive or difficult to engage would have led to a very different messaging strategy. But the research revealed something more human and far more actionable: many patients wanted to engage but felt overwhelmed by the complexity surrounding them.  

That insight became a turning point for how we would think about the brand, moving forward.

It shifted the role of the brand, informed the tone of messaging, and reframed communication priorities. Most importantly, it helped create a more empathetic and supportive positioning strategy grounded in the realities patients were experiencing.
‍

Turning Research into Brand Strategy

The output of the project was not simply a research report. The validated personas became the foundation for a messaging matrix that mapped each audience segment to:

• Emotional barriers

• Communication preferences

• Proof points

• Tone considerations

• Engagement opportunities

• Practical messaging guidance

• Video clips of personas in real-life

The work gave marketing, sales and engagement teams clearer language they could immediately apply in the near term, while also establishing a stronger long-term strategic foundation for broader brand positioning, patient engagement, and experience design initiatives.

Perhaps most importantly, the process helped align stakeholders around a more accurate and real-time understanding of the audience. The personas were no longer abstract profiles.  

And the entire project was completed in six weeks.

Why This Matters

AI research is often framed as a shortcut. And we’ll be the first to call out AI slop when it rears its ugly, misshapen head. But this project demonstrates how it can be harnessed, not to replace human insight, but to uncover it faster, and plum it deeper.  

Used thoughtfully, AI can help teams uncover patterns and accelerate insight across complex audiences. But the technology alone is not the strategy. The real value comes from combining AI-enabled scale with human judgment and validation.

For brands where emotional nuance and lived experience shape nearly every decision, that combination can create a significantly stronger foundation for brand strategy work.

In the case of our healthcare client, the result was more than faster research. It was a clearer understanding of the people behind the data and a brand strategy built to meet them with greater empathy, relevance, and meaningful interest.

‍