Pharmaceutical companies have made enormous strides in data and analytics investments. In fact, 85% are actively funding data and AI initiatives. And yet, many still struggle to translate that data into the one thing that really provides strategic impact: actionable business intelligence.
According to Jason Smith, chief technology officer of AI & Analytics at Within3, the problem isn’t a lack of data—it’s how that data is used.
“We’re very data rich, but we’re still struggling to have high-value insights,” Smith told attendees at Pharma 2025 in Barcelona. “We find that 43% of companies are still fragmented in how they approach decision-making.”
So why is it still so hard to turn insights into action—and what can organizations do differently?
The real cost of siloed intelligence
Despite the promise of AI and advanced analytics, most pharma companies face major integration roadblocks. Critical information—real-world scientific, commercial, and regulatory, —is scattered across departments, buried in unstructured formats, or trapped in legacy systems.
This siloed environment limits collaboration, slows time to market, and makes true strategic planning nearly impossible. According to Smith, the impact is real and quantifiable: organizations that successfully integrate insights across departments report up to 50% higher innovation capability and 12–28% faster market entry.
Building toward integration: a three-stage journey
How can pharma organizations move from fragmented to fully integrated? Smith outlined a pragmatic approach that begins with achievable wins and can scale over time.
- Start with focused, departmental use cases.
Many medical affairs teams, for example, struggle with data scattered across CRMs, SharePoint, and internal docs. Implementing structured data capture—with natural language processing layered on top—can surface valuable insights from existing information almost immediately.
- Connect insights across functions.
The real breakthrough comes from sharing intelligence across commercial, clinical, and medical domains. Smith cited Pfizer’s breast cancer drug Ibrance as a case in point: when they discovered physicians prescribing it off-label for male patients, traditional trials weren’t an option due to low patient volume. So Pfizer combined clinical trial data with safety and real-world evidence into a unified data lake—generating enough evidence to seek expanded FDA approval without launching new studies.
- Scale to full enterprise integration.
Novartis took this even further. By aligning on FAIR data principles—Findable, Accessible, Interoperable, and Reusable—and building a cross-functional platform to match, they were able to support more than 200 analytical use cases. The results? Project setup times dropped from weeks to a single day, and insights from millions of patients were now accessible to the teams that needed them most.
It’s not just about the tech—it’s about the adoption
Advanced tools and platforms matter—but they’re not the end of the story. According to Smith, the biggest challenge is human, not technical.
“No matter how sophisticated our AI is, if we don’t have organizational adoption, we’re going to have issues.”
To drive success, pharma companies need to:
- Define core business problems and ROI goals upfront
- Secure cross-functional buy-in from the start
- Design with user experience, not just compliance, in mind
Starting small—with the right guardrails and governance—can quickly prove value and open the door to broader transformation.
The bottom line
Data isn’t enough. To gain a true competitive advantage, pharma companies must connect, interpret, and act on that data—across every function. The organizations that build integrated, insight-driven strategies will lead the next generation of drug development, launch success, and patient impact.
Check out the key takeaways from Pharma 2025 here, and then get in touch with our team to talk about building a cross-functional insight strategy that works for your organization’s specific needs.
