Scaling GenAI Requires Organizational Change
The main challenge for leaders scaling generative AI (GenAI) is not just technology deployment but managing organizational transformation. Novo Nordisk’s large-scale rollout of Microsoft’s Copilot in 2024 revealed that success depends heavily on how employees adapt and collaborate with AI. While tools evolve rapidly, the human dimension shapes whether GenAI delivers measurable value. For example, employees saved an average of 2.17 hours weekly, but satisfaction correlated more strongly with improved work quality than time saved. This shows that GenAI’s value extends beyond efficiency to enhancing the nature of work itself.
Managing Adoption Phases to Sustain Usage
GenAI adoption unfolds in phases, often starting with enthusiasm that declines before rebounding. Novo Nordisk saw 23% frequent users after one month, but 15% dropped off after three to four months as initial easy wins faded. Productivity gains dipped when users expanded tool use from two to five apps but improved again at six or more. This midcycle dip is common and can lead to abandonment without targeted interventions. Novo Nordisk countered this with timed training, license management, champion networks, tip newsletters, and continuous feedback. These strategies emphasize that ongoing support and training are critical to sustain and deepen GenAI adoption.
Tailoring GenAI Support by Business Function
GenAI’s impact varies widely across business functions. At Novo Nordisk, corporate and commercial teams saw the highest productivity and quality gains, while research and development teams experienced smaller improvements. STEM employees struggled with GenAI’s probabilistic outputs and occasional hallucinations—such as factual inaccuracies and reasoning errors—that conflicted with precision-driven tasks. To address this, Novo Nordisk shifted from a uniform rollout to tailored enablement, offering function-specific onboarding, use-case playbooks, and customized Copilot features. This targeted approach helped employees in diverse roles find meaningful ways to integrate GenAI into their workflows.
Leveraging Experienced Employees as Adoption Champions
Contrary to common assumptions, senior employees at Novo Nordisk outperformed younger colleagues in using GenAI effectively. Their deeper understanding of workflows enabled them to identify high-impact use cases and critically evaluate AI outputs. Surveys showed experienced workers achieved greater productivity and quality improvements. To harness this advantage, Novo Nordisk created a champion network mainly composed of seasoned staff who led peer demos and role-specific training. Internal social platforms facilitated knowledge sharing between senior and junior employees. This strategy highlights that GenAI success depends more on contextual fluency than digital nativity.
Addressing Cultural Resistance and Ethical Concerns
Cultural resistance and AI shaming posed significant barriers at Novo Nordisk. Some employees viewed GenAI as unethical or feared making mistakes with AI-generated content. Concerns about privacy, environmental impact, and workflow disruption fueled reluctance. Novo Nordisk responded with transparent ethical use guidelines, clarified ownership rules, and leadership messaging that framed GenAI as a tool to enhance work quality rather than cut corners. They fostered safe spaces for questions and peer support via platforms like Viva Engage. These efforts helped transform skepticism into engagement, demonstrating that addressing cultural and ethical concerns is essential for sustained adoption.
Six Key Levers to Scale GenAI Adoption Successfully
Based on Novo Nordisk’s experience, six levers are critical for scaling GenAI as an organizational transformation: layered training and onboarding tailored by role, champion networks providing domain expertise, internal communities for peer support, targeted microcommunications, ongoing feedback mechanisms, and leadership commitment to ethical use. For instance, senior-led champion sessions and microlearning modules address initial learning curves, while use-case libraries and peer shadowing reduce resistance. Continuous pulse surveys and usage analytics enable adaptive support. Employing these levers in combination ensures that GenAI initiatives move beyond technology deployment to lasting organizational change.
