Pharma Organization Change Management – Driving System Thinking in Pharma
Why Organizational Change is the Real Challenge
Many organizations in Life Sciences have already invested in data platforms, AI, and increasingly, knowledge graphs (KGs). The promise is clear: break silos, connect knowledge, enable AI to achieve better decisions (aka “Decision Intelligence”). And yet, despite working technology, value often remains elusive.
The reason is simple, but uncomfortable: system thinking and graph-based ways of working are not tools they are ways of thinking hence working. In the AI era, system-thinkers with a well-working imagination, thinking (IQ) and feeling (EQ) outside of the (AI) box, will thrive.
System-thinking challenge habits, processes, incentives, and even identity. It requires moving from:
-
- isolated data → connected knowledge
- local optimization → systemic coordination
- reporting → transparent decision enablement
This is not a deployment problem. It is an organizational change problem and to make this shift tangible, it helps to think in terms of a mindset topology (see https://www.biomedima.org/a-mindset-topology-of-it-and-business-in-pharma/) Organization Change Management (OCM) must evolve on the four complementary axes of the OCM compass:
-
- Top-down (leadership and direction)
- Bottom-up (adoption and practice)
- Left-right (business → IT understanding)
- Right-left (IT → business understanding)
If one axis is missing, the system does not hold.
1. Top-Down: Creating Direction, Ownership, and Legitimacy
Every transformation starts or fails at the top. System thinking initiatives are rare and often struggle because they are perceived as “digital projects,” delegated to IT or innovation teams without clear business ownership. Without strong leadership alignment, they remain optional, experimental, and ultimately expendable.
To succeed, three elements are critical.
-
- First, management engagement must be explicit and sustained. Not just initial approval, but continuous involvement. Leaders must not only sponsor the initiative, they must adopt it and show by the example”, use its outputs to deliver value, ask questions differently, and communicate relentlessly that this way of working matters.
- Second, there must be single-thread ownership. System thinking cuts across functions: R&D, regulatory, commercial, IT. Without clear accountability, it dissolves into fragmentation. One accountable leader, not a committee, must own the outcome.
- Third, sponsorship must be sustainable, not opportunistic. Many initiatives start with enthusiasm and fade when priorities shift. System thinking requires time. It must be positioned not as a pilot, but as a strategic capability.
2. Bottom-Up: Making It Real for Those Who Do the Work
Top-down direction creates legitimacy. Bottom-up engagement creates reality.
System thinking fails when it remains conceptual i.e. when it is presented as architecture, frameworks, or abstract models disconnected from daily work.
Adoption happens with doers. Clinical scientists, data managers, regulatory experts in different organizational departments are the people who interact with data every day. If they do not see value, nothing changes.
The most effective approach is simple: bring the work to the data, and the data to the work. Instead of generic demonstrations, use real, messy, familiar datasets. Organize workshops (BYOD = Bring Your Own Data workshops) where teams bring their own data and questions:
-
- “Why is it so hard to reuse this information?”
- “Where are we losing time reconciling data?”
- “What do we wish we could see across studies?”
Then, show concretely:
-
- how data can be linked across silos
- how context is preserved
- how questions can be answered faster
This creates a shift from abstraction to experience.
Educational sessions are important, but insufficient. What matters is hands-on realization:
-
- seeing connections emerge
- understanding how meaning is preserved
- experiencing how decisions can be supported
At that point, system thinking stops being “a new concept” and becomes a better way to work.
3. Left-Right: When Business Looks at IT
One of the deepest barriers lies in how business perceives IT (see https://www.biomedima.org/a-mindset-topology-of-it-and-business-in-pharma/).
From a business perspective, IT is often seen as:
-
- focused on structure rather than meaning
- slow to adapt to evolving needs
- disconnected from decision-making realities
This perception is not entirely wrong but it is incomplete.
Business professionals operate in a world of uncertainty, where decisions must be made without full information. They seek answers like:
-
- “Should we move forward?”
- “What is the risk?”
- “What does this mean for our strategy?”
When interacting with IT systems, data platforms, knowledge graphs, they often encounter complexity:
-
- unfamiliar representations
- too much information
- not enough direction
The reaction is immediate: “How does this help me in my daily routines?”
To bridge this, business must evolve as well. It must recognize that:
-
- structure is not bureaucracy, it is what makes knowledge reusable
- consistency is not rigidity, it is what enables scaling decisions
- data models are not technical artifacts, they are representations of reality
Business needs to develop a minimal level of data and system literacy. Not to become technical, but to understand:
-
- how meaning is encoded
- how relationships are represented
- how insights can be derived
Without this, expectations remain disconnected from what systems can realistically deliver.
4. Right-Left: When IT Looks at Business
The mirror image is equally important and often more difficult.
From the IT perspective, business can appear:
-
- ambiguous
- inconsistent
- driven by narratives rather than logic
Requests seem unclear: “We need better insights”, “We need to understand the data better”. Without structure, these are difficult to operationalize.
IT professionals are trained to seek clarity, define schemas, and ensure consistency. Their instinct is to ask: “What is the model?” But business operates differently.
Meaning is contextual. Identity is fluid. Decisions are probabilistic. A dataset is not just data it is:
-
- evidence for regulators
- a story for investors
- a signal for internal decisions
For IT to enable system thinking, it must move beyond structure alone. It must embrace:
-
- ambiguity as a starting point, not a problem
- interpretation as part of the system, not outside of it
- narrative as a valid layer of meaning
This does not mean abandoning rigor. It means extending it. The role of IT shifts from:
-
- building systems → enabling understanding
- managing data → supporting decisions
And this requires a cognitive shift as much as a technical one.
Aligning the organization Cognitive Topology
System thinking and graph-based ways of working combine with AI promise something powerful: the ability to connect data, meaning, and decisions across an organization. But this promise cannot be delivered by technology alone.
It requires alignment across four dimensions:
-
- leadership that legitimizes and sustains
- practitioners who adopt and experience
- business that understands structure
- IT that understands meaning
This is the true topology of transformation. When these axes align, something changes fundamentally:
-
- data becomes knowledge
- knowledge becomes actionable
- systems become trusted
When they don’t, even the best technology remains unused. Because in the end, the challenge is not building systems that connect data. It is building organizations composed of humans and machines that can think in systems. And that is a much harder problem.