The HCP who no longer needs pharma's front door

Start with who's arriving. Gen Z and millennials are projected to make up around three quarters of the global workforce by 2030, Gen Z alone accounting for close to a third. The HCP of 2030 isn't a distant persona. They're already in medical school, residency, nursing and early-career practice now, and the demographic centre of gravity in healthcare is already moving toward people with fundamentally different expectations of information, technology, trust and convenience. The youngest GPs in the UK qualify at 26. The consultants making decisions on integrated care boards a decade from now are, right now, medical students forming their habits.

The shift is already showing up in how HCPs consume information. Over half of physicians prefer short plain-text summaries to long detailed reports. Close to eight in ten cite lack of time as the barrier to engaging with medical content at all. Younger HCPs now rank podcasts above most traditional channels, and it's not hard to picture personalised, AI-generated podcasts curated automatically from what a clinician actually needs to know, replacing the newsletter and the rep email alike. Attention was never pharma's to assume. It was always going to have to be earned in the first few seconds, and that was true before AI, it's just harder to ignore now.

The old question was which channel pharma should use to reach an HCP. The question now is whether pharma remains discoverable, trusted and useful inside an information journey it doesn't own at all.

More than 80% of physicians already use AI professionally, with the most popular use cases being summaries of medical research and standards of care (39%), creation of discharge instructions, care plans or progress notes (30%) and documentation of billing codes, medical charts or visit notes (28%). Open Evidence in particular has become the go-to source for US physicians in a remarkably short space of time, sitting between the doctor and the manufacturer in a way no single pharma channel ever could. This transformation of the information seeking journey is already reshaping how content teams plan today.

What’s changing isn’t the value of pharma’s web presence, but how it’s being used. Leaders describing this shift aren't watching their web presence become irrelevant, they're watching who visits change. Human traffic to HCP portals and brand sites is falling in some cases dramatically, even as overall reach grows, because content built well enough to be trusted is increasingly read, summarised and cited by an AI system before a human ever clicks through. The web isn't dying. It's becoming something closer to infrastructure: the place where a manufacturer establishes authority, so that the agents doing the reading on a clinician's behalf have somewhere trusted to point to.

That reframes what "engagement" even means, and it's worth being precise about the difference. A visit is activity. A citation an HCP's agent actually relied on is a better indication of value delivered, and value builds trust. Most of pharma's existing measurement still counts the first and calls it the second.

What does this look like from the HCP’s perspective? 

Discovery starts inside LLMs and agent-curated feeds, not a search bar or an inbox. Awareness still runs through short-form content, congress, influencers and peer-to-peer conversation, but real conversion increasingly happens through approved, product-connected sources an HCP has chosen to trust, not a channel pharma chose for them. Content quality and credibility matter more than which destination it came from, because for a growing share of interactions, no destination was involved at all.

The clinical relationship changes alongside it. A data-driven patient record starts to flag prevention, early diagnosis and intervention before the HCP has to ask. Peer validation, long the thing HCPs trusted over anything pharma produced, increasingly means an aggregated model built from thousands of real-world cases rather than a handful of names on a conference stage. And in the more provocative corners of these conversations, some leaders go further still: a patient's own AI health agent flagging a treatment option before the HCP does, changing who actually leads the conversation. Patients are changing at least as fast as HCPs, self-diagnosing, researching and arriving with their own conclusions, changing the nature of the consultation. 

The manufacturer relationship narrows to what genuinely earns trust. Transactional and promotional contact increasingly runs through virtual agents. Live time with MSLs gets reserved for conversations that are actually scientific, and more than one leader has already said aloud what most of the industry is still avoiding: the field force as it exists today may simply not be there in its current shape. Not gone, but orchestrated differently, with engagement increasingly led by HCP preference rather than traditional call-cycle planning.

None of this is a distant forecast. The infrastructure for most of it exists today. What's missing, almost everywhere, is an organisation equipped to work with it effectively.

The organisations pulling ahead are building AI fluency while also rethinking what creates value. Authority is becoming as important as reach, and it is earned through consistent, credible behaviour over time. Barely a fifth of HCPs currently place high trust in pharma sources, against well over half for independent ones, and the gap doesn't close because a brand says trust me. It closes because an agent, or a clinician, infers it from behaviour it has already observed. That's a harder thing to plan for than a campaign, and a far more durable thing to win.

That shows up in very specific choices. Some are already building content for two audiences at once, the human reader and the system summarising them on a clinician's behalf, with dedicated teams whose entire job is understanding how that citation layer actually works. Some have moved away from one-off, agency-led campaign builds entirely, toward a product model: reusable components, shared design systems, and the ability to ship a change once and have it work everywhere, instead of rebuilding the same page eighteen times across eighteen markets. Some are treating their own web presence not as a destination to defend, but as the place their data and evidence live, precisely so the agentic layer above it has something worth pointing to.

We’re seeing a major shift in the value of brand promotional vs scientific and educational. However campaigns, KPIs, agency networks and media partners are incentivised to maintain the brand driven promotional status quo. There is also huge investment at enterprise level to establish the technologies, systems and governance to automate and orchestrate engagements. This is resulting in a change to the vendor landscape away from brand agencies towards a centralised approach investing in content factories, design systems and journey automation enabling consistency at scale.

It also means measuring differently. One leader described a client whose leadership had mandated a tenfold increase in touchpoints, with no measure of quality attached at all: more noise, rewarded as if it were more value. The organisations getting this right are moving deliberately away from reach and frequency toward trust, resonance and quality, treating a genuine medical conversation as worth more than ten scrollable ones, and building shared incentives across every function that touches a customer, not just the ones with a sales target attached.

And it means surviving the pendulum. Leadership attention swings hard, from an all-in AI and GEO (generative engine optimisation) push one quarter to a retreat toward field-force simplicity the next, and back again. The leaders navigating this best are the ones who've built a durable foundation and are clear on what's allowed, what's valued and how content gets made, so that whichever way leadership's attention swings next, the groundwork still compounds instead of resetting to zero.

• Decide what you're the authority on, and rebuild around it. Not the brand story: the trial data, the safety record, the real-world evidence. That's what clinicians, AI agents, and increasingly patients, will actually want to cite.
• Design for two audiences from the outset, not as an afterthought. The system reading your content on someone's behalf is now as important an audience as the human who might eventually click through. Invest in a ‘total search’ strategy, truly understand the changing dynamics and adapt. 
• Make the call on your field force deliberately, ahead of the curve. Whether that means fewer reps, a virtual transactional layer, or a model built entirely around scientific exchange and orchestration, deciding this early as a considered bet is a very different thing to having it forced on you later.
• Change what you measure before you change what you build. Consolidation and quality content only pay off if trust and resonance genuinely outrank reach and frequency in how success gets defined, and in what gets rewarded further up the chain.
• Build the foundation that survives the pendulum. Fund the durable layer, the rules, the values, the shared context that makes every future decision faster, rather than only funding whichever initiative currently has leadership's attention. That's what keeps progress compounding when priorities inevitably swing again.

None of this is guaranteed to play out exactly as described here. But the cost of waiting isn't neutral. Every quarter spent optimising reach is a quarter an HCP's agent spends learning who else to trust instead. 

The organisations shaping the HCP of 2030 are the ones already acting like it's 2028, treating this as an authority problem rather than a reach problem, for a customer who's already further along this path than most transformation roadmaps currently admit.