Life science industry leaders expected plenty of action in 2025, what with another changing of the guard in D.C. and the ongoing push to include AI in all aspects of everything. Even so, the pace and scale of change over the last 12 months awed even the most battle-hardened marketers and technologists. We asked some of them to weigh in on what we just experienced – and what it all means. Here are their takes on a year that, for whatever else you might say about it, was definitely not boring.
*****
Looking back at 2025, what shift within pharma marketing or health-tech surprised you most, and why?
Meg Rivera, former SVP, U.S. market president, Organon and CMO, head of commercial, Akili Interactive: Perhaps that in an age of increasing focus on technology, we saw humanity re-enter the picture. I wouldn’t say pharma suddenly became dramatically more human in 2025, but we did see subtle signals that the pendulum is starting to swing back – not through big tear-jerker campaigns, but through quieter moments of honesty and clarity. You saw it in obesity care, where companies like Lilly and Novo were pushed to communicate more transparently about access barriers, affordability challenges and what real patients were experiencing. That shift of meeting people in their lived reality rather than idealized journey maps is what “more human” actually looks like.
But it’s delicate progress. At the same time, companies are becoming enamored with AI, and cost pressure is driving operational expense squeezes, downsizing and consolidation, and overall increased scrutiny of profit margins. Those forces could easily pull us back toward impersonal automation, in the form of more content, more speed and less meaning. The opportunity for 2026 is to keep the human moments – clarity, accessibility, lived experience – at the center. AI should enable those things, not erase them.
Molly Walpola, former CX senior manager, Novo Nordisk: We spent two years worrying if AI would take our jobs, but the real disruptor in 2025 wasn’t an algorithm – it was the twin vice of regulatory pressure. The combination of price caps and PBM reform forced the industry to stop experimenting with tech and start deploying it for survival. The surprise wasn’t that AI worked; it was how fast we pivoted from generative toys to agentic workhorses to protect our margins. We stopped asking AI to write emails and started asking it to save the P&L.
Justin Molavi, senior director, AI business products, Genentech: I can’t say that I’m very surprised by what I’m seeing, as I think we’re behind many other industries that have been marketing in channels we’re just beginning to play in (for example, streaming audio or shifting existing channels to more programmatic buys). But I am impressed with the speed of how we’re embracing these digital channels. Pilots are quick and result in sustained change, which is exciting.
Jane Urban, chief data and analytics officer, Improzo and former VP, customer engagement operations, Otsuka: The biggest surprise of 2025 was how fast pharma moved from piloting AI to actually embedding it in day-to-day operations. The industry suddenly realized that committing to any single vendor or platform was a trap. The real lesson of the year, then, was that flexibility beats loyalty. AI is evolving too quickly to bet on one stack, and companies that locked themselves into a single solution learned the hard way that you need an interoperable, future-proof approach if you want to keep pace.
Will McMahon, director, omnichannel strategy, AstraZeneca: The speed at which some pharma companies embraced integrated, direct-to-patient models. Telehealth and fulfillment have grown since COVID, but what’s new is the thoughtful reengineering of siloed resources into intuitive, end-to-end customer experiences.
These offerings shorten journeys from weeks or months to days and extend support beyond traditional access and affordability into integrated patient services or practice support for HCPs. It signals a shift in how marketers are moving from point solutions to orchestrated experiences that meet patients where they are, with measurable improvements in friction, speed and satisfaction.
Scott Miller, president and CEO, Solutions for Enterprise Imaging, GE HealthCare: In 2025, the most striking shift was how quickly health systems prioritized enterprise-wide imaging integration as a strategic imperative rather than a technical upgrade. For years, imaging solutions were largely evaluated at the departmental level. This year, however, organizations increasingly recognized that imaging data is foundational to the entire care pathway – and that it must be accessible, connected and intelligently orchestrated across the enterprise.
What surprised me was not the direction of the shift, but its speed and scale. Providers are now expecting imaging platforms to unify data, streamline workflows and improve collaboration across radiology, cardiology, oncology and other clinical domains. The focus has moved from selling discrete tools to delivering holistic, outcome-driven solutions that reduce clinician burden and enhance the quality and consistency of interpretation.
This momentum underscored a broader truth: Innovation is only meaningful when it improves performance across an entire healthcare system, not just within a single modality or department.
Zach Capetola, oncology omnichannel marketing strategy-HCP engagement and capabilities, Merck: Nothing surprised me, but one thing stood out: How quickly teams jumped to advanced AI use cases without solid foundations in place. We saw the same movie during the digital wave. The difference now is that the stakes, and expectations, are even higher.
Mariano Bendersky, former head of digital healthcare and innovation, U.S. autoimmune T1 diabetes, Sanofi: 2025 marked a clear inflection point in pharma’s move toward more direct, patient-centric engagement. We saw a growing number of companies meaningfully adopt direct-to-patient strategies and integrate telehealth more deeply into care and engagement pathways.
What surprised me most was the pace of this shift, despite early-year caution driven by regulatory scrutiny and Congressional letters. As the year progressed, organizations became more confident navigating compliance while still innovating, recognizing that patient expectations, access challenges and digital care models were evolving faster than traditional pharma operating models. Importantly, the shift reflected not just new channels, but a greater comfort level with experimentation within regulatory guardrails.
What does the agency-side consolidation we saw in 2025 mean for pharma companies? How could some of those relationships evolve in its wake?
Walpola: Consolidation in 2025 effectively killed the vendor model and ushered in the platform era. It’s a double-edged sword: We finally gained a unified view of the customer through connected data, but we lost the nimble reactivity of the boutique. We bought scale at the cost of speed, so now our job is to engineer the agility back into the system.
Consequently, the old retainer model is dead. With agencies now pitching themselves as data-rich powerhouses, clients are demanding they put skin in the game. We’re seeing a rapid shift toward performance-based contracting where payment is tied to outcomes – script lift, adherence or enrollment, rather than hours. If an agency claims their AI stack is revolutionary, the contract now says, “Prove it, and we’ll pay you. If not, we won’t.”
McMahon: This one is a mixed bag, because I know a lot of talented, hard-working people impacted by consolidation this year. Agencies play several important roles for their clients, few more important than acting as a catalyst for fresh thinking and a guard against stale ways of working. Disruption to teams and priorities can blunt that competitive advantage.
I expect new models and capabilities will emerge, but constant team reshuffling and ambiguity impose a real tax on momentum. Client expectations for value must be delivered in parallel with disruption via faster onboarding, continuity in strategy and creative, measurable efficiency gains and clearer accountability. Otherwise, clients will rebalance with alternative partners and insource critical capabilities. It will be interesting to see early results – perhaps more modular engagements, outcome-based contracts, embedded hybrid teams and tighter integration with client data and tech stacks to maintain continuity amid change.
Rivera: The consolidation we saw in 2025 should be a wake-up call, not because of who bought whom but because it spotlighted a deeper problem: We’re cutting costs without changing how the work actually gets done. You can rearrange head count, merge networks and streamline P&Ls, but if the underlying operating model stays the same – which often includes siloed teams, bloated processes and opaque scopes – clients will suffer.
And they did in 2025. The old AOR model is already over, but the answer isn’t what we’re seeing play out. The answer is much more fundamental: Clients need transparency, honesty and real partnership. They need agencies that put shared performance standards on the table and that are willing to work differently, not just cheaper. If networks can’t make that shift, I think we’ll see an exodus – not just to independents, but toward a completely reimagined agency partnership model. This year made it clear that while resourcing can change overnight, trust and value cannot.
What’s one AI use case in pharma marketing or health-tech that proved more impactful than people anticipated and/or one that underdelivered?
McMahon: Underdelivered: This might be a bit contrarian, but I admittedly had high expectations for GenAI and content. We’ve seen more content produced more cheaply, but not a true leap in end-to-end speed from ideation through MLR to publication. The individual parts are real and quite impressive (image/video gen, contextual targeting, agency tools, publisher formats, agentic workflows), but they aren’t yet unified into a scalable, compliant production engine for all but the largest enterprises. Many of the small and mid-sized firms best suited to benefit from content at scale are still struggling to adopt and scale this technology without the human (and machine) resources needed.
Overdelivered: Embedded AI in foundational channels like email and SMS. Companies like Ostro showed that integrating AI into existing touchpoints to enable two-way communication can drive personalization and responsiveness without creating new walled gardens. The trend I hope this foretells is fewer sandboxes or platforms in search of an audience and more embedded solutions that can unify creation, governance and activation into a single pipeline designed for pharma’s complexity and scale.
Bendersky: AI-driven patient and HCP identification proved more impactful than many anticipated in 2025. What had lived in pilots for some time began to show tangible value, enabling teams to focus on the right patients and engage the most relevant HCPs with greater precision and efficiency. We also saw meaningful progress in AI-enabled omnichannel orchestration. Rather than optimizing individual channels in isolation, brands increasingly used AI to determine the right channel, message and timing. This improved engagement quality, not just volume.
Miller: One AI use case that exceeded expectations in 2025 was AI-driven workflow orchestration within enterprise imaging. Many initially saw AI’s value in imaging as tied to reconstruction, detection or automation of individual tasks. But its most impactful contributions came from coordinating the full diagnostic workflow: intelligently prioritizing studies, surfacing critical findings, harmonizing data across locations and supporting clinicians in managing rising imaging volumes.
This evolution had far-reaching effects. By reducing manual steps and helping distribute workloads more effectively, AI improved both operational efficiency and clinician well-being. It strengthened consistency across the enterprise and reinforced the role of imaging as a connected, strategic asset rather than a standalone function. The lesson was clear: When AI is applied to enterprise-level challenges rather than isolated tasks, its impact is significantly amplified.
Capetola: Contextual data overdelivered. Pulling together fragmented information to build integrated HCP profiles gave us sharper segmentation and more meaningful engagement. Predictive analytics underdelivered, because the promise exceeds the current reality. AI is powerful, but people (and clinical decision-making) aren’t that linear. We need to temper our expectations and let nuance lead the way.
Walpola: If you look at the 2025 scorecard, the unsexy operational stuff won. AI in manufacturing quietly saved the industry billions by optimizing yield and maintenance. But in marketing, the “enterprise brain” concept underdelivered. We thought we could feed decades of fragmented CRM data and field notes into a model and get magic customer insights, but we underestimated the mess. As it turns out, unstructured commercial data doesn’t become gold just because you run it through a better algorithm. “Garbage in, garbage out” applies to AI just as much as it does to Excel.
If you had to identify one lesson the industry learned in 2025, whether through success or failure or lived experience, what would it be?
Capetola: That you can’t build a smart, connected ecosystem on a crooked foundation.
In 2025, we saw teams racing ahead with AI pilots and shiny new tech while their data infrastructure looked like a DIY basement with uneven footings and a leaky sump pump. The lesson? A smart fridge and a trendy color palate won’t save you if your foundation can’t hold weight.
Walpola: The biggest lesson marketing leaders learned in 2025 is that governance beats capability. We spent years chasing flashier generative AI tools, thinking that “more power” meant “better campaigns.” But the reality check hit hard: The best AI is useless if it feeds on bad data.
2025 proved that the most innovative marketing teams weren’t the ones using the flashier tech, but the ones using better-validated, better-managed systems. It matters because we realized that you can’t automate personalization until you fix your data foundation. The winners this year stopped buying new toys and started cleaning their rooms.
Rivera: The biggest lesson of 2025 was a simple one: If you don’t fix the experience, nothing else matters. We chased efficiency – consolidation, automation, more tech, fewer people – but efficiency doesn’t solve access, affordability or trust. The brands that made progress were the ones that stopped polishing the story and started addressing the friction. That’s the shift the industry has to make: From optimizing the machine to improving the human experience it’s supposed to serve.
Bendersky: The key lesson from 2025 is that incremental change is no longer sufficient. Life sciences organizations are being pushed to evolve by converging forces like regulatory shifts, changing HCP expectations, digital-first communication norms and the rapid evolution and larger use of large language models.
Innovation cycles are shortening, and organizations that fail to adapt their operating models, decision-making speed and digital capabilities risk falling behind. 2025 reinforced that digital and AI transformation is no longer a future aspiration; it is a near-term business imperative tied directly to relevance, efficiency and patient impact. Importantly, transformation is as much about governance, talent and operating model as it is about technology.
McMahon: Everything comes back to data and how usable it is for any given team. Some pharma companies have made integrated two-way communication, voice of customer and the development of more robust customer feedback loops a priority, but the value is constrained without a strong data strategy. Governance, identity, consent, enrichment and activation across channels and measurement need clear ownership and operating models to realize the potential of these insights.
With the advent and scale of these conversational AI tools, we are going to see a massive amount of first-party customer intelligence created and captured over the next few years. Organizations that treat conversational data as a durable asset, applying ethics, user consent and security, will set the pace in trust-based customer engagement. This should create faster learning cycles, better personalization and demonstrably better outcomes.
Molavi: One key lesson has been that we own our campaigns, end to end. We can’t outsource our understanding of what went wrong or right, and without these lessons we’ll continue to be fragmented as marketing and analytics leaders. The more we punt these decisions and this understanding (we’ve done this historically for fair reasons, as we’ve been used to our existing agency/company relationships), the less we’ll grow collectively.
Want to contribute to an upcoming story on industry trends in the year ahead? Drop us a note at hello@kinara.co and subscribe on the website to receive Kinara content.
