CASE STUDY 01

Keeping the platform, changing its job

As AI changed how customers could use our data, the debate inside the company split two ways: the platform is the destination for everyone, or agents make it obsolete. I made a third call.

Role
Product lead, market access portfolio
Company
MMIT
Timeline
2026
Status
Adopted, now in execution
  • Portfolio strategy
  • Market segmentation
  • AI-era commercialization

The problem

MMIT had spent years moving from selling data to selling a platform: one place for market-access intelligence. Then AI changed the math. Customers could build their own analysis, pull data into their own tools, and point agents at it. If customers no longer need one interface to get value, what's the platform for?

The call

I treated it as a segmentation problem instead of a technology decision. Customers differ in budget, technical capacity, internal AI policy, and appetite to build. Some want a finished product. Some want governed data to use inside their own systems. Some are going all-in on AI workflows. Those are several valid ways to buy the same intelligence.

So I kept the platform and changed its job. It moves from the top tier to a turnkey middle tier: trusted data and opinionated analysis, with no assembly required. AI-first and agent-first offerings sit above it. Data feeds and APIs serve customers who want to build for themselves.

AI didn't eliminate the need for products. It made the preferred form of the product less predictable.
Portfolio tiersStacked translucent glass layers. From top: AI-first and agent-first offerings, the platform as a turnkey middle tier, data feeds and APIs for teams who build their own, all resting on one governed data foundation.AI-first and agent-firstIntelligence inside AI tools and agentsPlatformTurnkey, governed, opinionatedData feeds and APIsFor teams who build their ownOne governed data foundationTrusted data under every tierPortfolio tiersStacked translucent glass layers. From top: AI-first and agent-first offerings, the platform as a turnkey middle tier, data feeds and APIs for teams who build their own, all resting on one governed data foundation.

How I got there

I mapped every way a customer could consume our intelligence, including data feeds, APIs, the platform, MCP, and agentic workflows, and defined the buyer, use case, and value proposition for each. The inputs came from client conversations, conference feedback, competitive signals, and how quickly different customer segments were actually adopting AI.

I worked the model through with product leadership, then presented it to executive leadership and the GM of the market access business.

What changed

Executive leadership adopted it as the direction for the portfolio. Product and Data leadership are now using it to define the new offerings, with SaaS, APIs, and AI experiences each playing a distinct role. It also gave us a filter for the wave of new AI-built solutions pitching for a spot on the platform: does this strengthen a tier, or dilute one?

It set up the next piece of work, rethinking how we price and package for AI-era buyers. (See the API case study.)