What I’m Still Thinking About After FiercePharma 2026

Fierce Pharma always leaves you with more ideas than you know what to do with, and this year was no exception. I sat in on sessions covering everything from agentic AI to patient trust to reputation management. Here's what stuck with me the most this past week in Philly:

Agentic AI IS the operating model.

The framing I can't stop thinking about came from Novartis's Kevin Ryan: pharma marketing is shifting from "player to conductor." Agility isn't a value anymore, it's the operating system with faster decisions, clearer decision rights, guardrails instead of gates. Merck's Lori Holland-Hancock took it a step further, describing agentic AI already taking over omnichannel execution work outright, not just assisting with it. The conversation has moved past "should we pilot this" and into "how do we restructure around it."

Brand teams are proving the AI efficiency case themselves

One of the most candid moments of the week came from Pfizer's Kimberly Coleman Clotman, who shared that some of her rare disease teams, working without dedicated budget, are now building full omnichannel content and engagement assets themselves using AI. This signaled that the efficiency gains agencies have been promising are real, achievable and here. We've leaned into this at Spectrum Science, using AI-supported workflows across every discipline globally. Agencies showing their own math on speed and efficiency will stay indispensable.

One genuinely inspiring case study on the "doing more with less" front: Dr. Nkiru Enendu's work on a rare disease launch showed what's possible by workshopping a "choose your own journey" experience that adapted content as patients moved through it. That resourcefulness is exactly the successful brand teams are employing.

Orchestration, identity and measurement and the same unsolved problem

Every brand is being pitched an omnichannel media platform, yet very few of those platforms talk to each other. ViiV's Ruth Woodrow pointed out that HCPs still land on the same content across disconnected systems that never carries them forward. Underneath that is a data problem: identity has to be the cornerstone for any new media platform. When everything starts with identity records, measurement becomes easier, too. Veracyte's Molly Cornbleet made a sharp point that resonated with us: net promoter score and brand lift are useful numbers, but they aren't omnichannel measurements. Real progress means starting with audience insight and audience action, then scaling only after small tests earn buy-in. BMS's Gaetan Akinrolabu echoed a similar spirit with the "ABT" (always be testing) philosophy, pushing back on the idea that reach alone tells you whether omnichannel is actually working. Orchestration without the right measurement lens underneath is just more disconnected channels, faster.

The room is quietly worried that AI scale will affect trust

AI makes it easy to reach more people, more often. But Lundbeck's Alice Harmon shared research that showed that past a certain frequency threshold, promotions don't always equate to more scripts (they drive opt-outs). ViiV's Meaghan Rivera named the shift directly: technology is table stakes now, so the real differentiator is investing in people, not just pushing more messages through more channels.

Where does this leave us?

Every one of these threads points to the same underlying shift: the tools are getting more powerful, but the judgment calls of where to invest, what to measure, how much is too much are getting harder, not easier. That's the work we love. If you're wrestling with any of this inside your own organization, we'd love to compare notes. Reach out to the Spectrum Science team, and let's figure it out together.