CASE STUDY 02

A payer intelligence solution built for the market

A client request for more payer data turned into a new product, because what clients really needed was help making sense of the data they already had.

Role
Product lead, research through GA
Company
MMIT
Timeline
2025–2026
Status
GA June 2026, about three months in market
  • Product strategy
  • Willingness-to-pay research
  • Applied AI
  • Commercialization

The problem

A client asked for payer detail our old database couldn't support. MMIT's new enterprise data model could, and it opened the door to much more. The easy move was bolting the new data onto the existing product.

But I'd watched how clients used Landscape. They weren't short on data. They needed to know what changed, why it mattered, and where to look next, and the existing product left that work to them.

The call

Build a new experience around those questions instead of extending the old one. That meant more investment, so I had to prove it was worth it.

How I got there

I priced it before we built it. An earlier product taught me this the hard way: I'd validated it by asking customers about requirements and came away with a great feature list and no idea whether anyone would pay. For iQ, every research conversation ended with “Would you buy this, and for how much?” Blinded market research on a prototype showed stronger demand for the bigger experience, and that made the investment case.

“Would you buy this, and for how much?”

I designed around how clients think. The new data model was far more flexible, and far more complex. The job was to expose the flexibility without the complexity. In external beta sessions we watched how clients navigated, listened to the words they used, and reshaped the product until it matched how they actually research and plan. An internal beta with client-facing teams gave us a second feedback loop.

I let the math decide and AI explain. I didn't want a model deciding what changed in a payer's coverage. I designed a deterministic framework that scores the magnitude and significance of each change and identifies its likely drivers. AI only writes the plain-language explanation on top. That kept the insights accurate and got AI into production fast.

Let the math decide, AI explainA payer change goes through deterministic scoring for magnitude, significance, and drivers. AI then writes a plain-language explanation, and the client gets an insight about what changed, why, and where to look.Math decidesAI explainsPayer changeA coverage orrestriction shiftsScoringMagnitudeSignificanceLikely driversSummaryPlain-languagesummary of thescored resultClient insightWhat changedWhy it mattersWhere to lookThe model never decides what changed. It only puts the result into words.Let the math decide, AI explainA payer change goes through deterministic scoring for magnitude, significance, and drivers. AI then writes a plain-language explanation, and the client gets an insight about what changed, why, and where to look.Math decidesAI explainsPayer changeA coverage orrestriction shiftsScoringMagnitudeSignificanceLikely driversSummaryPlain-languagesummary of thescored resultClient insightWhat changedWhy it mattersWhere to lookThe model never decides what changed.It only puts the result into words.

We built it in parallel. The enterprise data transformation was still underway, so the product and the data model evolved together, with beta evidence feeding both.

What changed

Landscape iQ beat its first-year bookings goal before year-end and signed its first clients before launch, at the price we'd validated. The internal beta did some selling of its own: client-facing teams who'd used it started bringing it to customers early.

It went GA in June 2026, so the story is still being written. The early numbers say the bigger bet was the right one.