Your insights team in 3 years

By Hannah Mann, Founding Partner, Day One Strategy

From producer to orchestrator: the next job in pharma insights

AI is changing pharma insights work, and not always in the ways people expect. Our new AI in Pharma State of Play report set out to understand exactly how, through twenty interviews with insights and analytics leaders across ten pharma organisations.

Of the four insights we uncovered, one stood out: the next step is people directing teams of AI agents, rather than producing the work themselves. This article looks at that finding in more depth, alongside our own view of where it leads.

“In three years’ time what a team of five used to do, now it’s going to be one or two people, with a bunch of agents.” That’s a VP of Global Strategic Insights describing where their own team is heading in the future.

Today, most AI use in insights is narrow and productivity-focused: drafting emails and reports, summarising documents that would once have taken hours to review, or pulling the right file out of a crowded archive in seconds. Where agents exist, they tend to do one job well, such as screening RFPs, checking compliance or monitoring competitor activity. They are trusted precisely because their scope is small and their output is easy to check.

But that is the foundation – moving to the next step is much, much harder.

From one chatbot to a team of agents

Across our conversations, one theme kept surfacing: a connected ecosystem of specialised agents, each doing a distinct job, working together under one person’s direction.

One market research team’s internal target is for every team member to have an AI-augmented “copy of themselves” running in the background by 2028. In practice, that means a tool that works behind the scenes, gathering intelligence and flagging what matters before anyone has to ask for it.

But most teams aren’t there yet, and here’s why. AI can save real time, but someone still has to check its work, and right now that checking takes almost as long as doing the task yourself. Ask the same question to two different AI tools and you can get two different answers. Until people trust what AI gives them without checking every line, that bigger vision stays exactly that: a vision.

That gap between ambition and reality came through clearly in our interviews. One insights lead at a global pharma company summed up where most teams sit today: “everyone sees the vision, but a genuinely tangible, day-to-day working relationship with agents hasn’t arrived yet”.

Multiple agents talking to each other, and to the team, is still something people want, not something they have.

A wider pattern, not just a pharma one

This pattern in pharma insights fits a wider change in how businesses are starting to organise themselves. The traditional business hierarchy was built for a world where information was scarce and decisions moved slowly, passed down through layer after layer of management. A growing body of thinking on intelligence-native organisations describes something much flatter: information flows freely, and AI agents work around the clock, spotting patterns and running scenarios while people sleep.

In this model, people don’t disappear. They set direction and decide what actually matters. Agents handle the signal. People handle the meaning. That’s exactly what our interviewees described, without any prompting from us, when they talked about having a copy of themselves working in the background, freeing them up to focus on the decisions that genuinely need a person in the room.

It’s already showing up outside insights, and outside pharma. In May 2025, Moderna merged its HR and IT functions under one leader, Tracey Franklin, whose title changed to Chief People and Digital Technology Officer. Around the same time, ServiceNow made a similar move, renaming Jacqui Canney’s role to Chief People and AI Enablement Officer. Neither change was cosmetic. Both reflect the same idea driving what we heard in pharma insights: managing people and managing AI capability can no longer sit in separate boxes.

What AI still can’t do

None of the leaders we spoke to expect AI to make the final call. One senior insights lead put it plainly: AI helps with efficiency, but it doesn’t understand emotion, and it won’t replace the human connection that real innovation depends on. If a team wants to think differently, not just faster, someone still has to do the thinking.

That’s the part AI can’t do. Deciding what a change in the data actually means for a specific brand, in a specific market, with specific regulatory rules to follow. Deciding whether a change in sentiment is worth acting on, or just noise. That judgement comes from experience, from having sat across the table from the people whose decisions it’s meant to help.

From producing the work to directing it

So what does this mean for the people doing the job today? The insights professionals who do best over the next few years won’t be the ones producing every chart and every summary themselves. They’ll be the ones directing a growing team of agents towards a clear business outcome: framing the right question, setting the agents to work, and deciding what the answer actually means for the brand.

This is what we mean by hybrid thinking: the power of technology and human judgement combined. Not technology replacing judgement, and not judgement working alone without the benefit of what technology can now do. The two working together, with a person firmly in charge of the direction.

At Day One, this is the thinking behind our Precision Intelligence approach and the engine that powers it, InsightBrain. Built for pharma, designed so people spend less time producing and checking outputs, and more time on the decisions that need their judgement. Industry knowledge and market research expertise become more valuable in this model, not less. They’re what makes the direction trustworthy in the first place.

Where this leaves insights teams now

Different companies are in different places. The teams furthest ahead have already stopped treating “more AI” as the goal. They’re building specific, well-governed uses of it that create real value, and using the time it frees up to do more of the work only people can do.

This is one of four insights from our AI in Pharma State of Play report, drawn from twenty conversations with insights and analytics leaders across ten pharma organisations. If you want the full picture, including where ROI, data and governance stand across the industry right now, you can read it here: dayonestrategy.com/resources/ai-in-pharma-state-of-play-report-2026