I had the fortune of meeting Jay Andersen, one of the pioneers in pharmaceutical decision-making. Below, you’ll find a few takeaways I had from our conversation that I thought every pharmaceutical leader should take to heart.
Jay Andersen, a University of Minnesota PhD statistician, transformed pharmaceutical decision-making during his 28-year tenure at Eli Lilly. Starting in 1993, he developed groundbreaking approaches to portfolio management and probability assessment that remain industry standards today.
In the early 1990s, Lilly embraced decision analysis through collaboration with the Strategic Decisions Group in Menlo Park. Championed by the support of Group VP Leigh Thompson, a small group developed portfolio processes that evolved into a cornerstone of Lilly’s development strategy. It survived multiple reorganizations and becoming a model for making decisions at other pharmaceutical companies.
3 critical success factors in pharmaceutical decision-making
i. Assessment vs calculation
The distinction between assessing and calculating probability of success is crucial. While modern approaches often seek to compute probabilities through data analysis, Lilly’s success came from structured expert assessment. Their Probability Assessment Group (PAG) brought together cross-functional experts who calibrated team estimates across the portfolio, ensuring both technical accuracy and organizational buy-in. This human-centered approach provided explainable, defendable assessments that executives could trust.
ii. Cultural considerations
Culture inevitably shapes decision-making, for better and worse. In hierarchical organizations, implementation can be swift but may lack transparency. Consensus-driven cultures require more stakeholder engagement but often yield more robust decisions. While this cultural influence can preserve institutional knowledge, it can also perpetuate biases that lead to overlooking promising rare disease treatments.
iii. Risk-centric analysis
Understanding specific risks, rather than abstract probabilities, provides the clearest path to good decisions. Whether evaluating toxicology, clinical endpoints, or market potential, granular risk assessment offers concrete factors that teams can analyze and address.
The AI opportunity
Artificial intelligence presents compelling possibilities for modernizing portfolio management and pharmaceutical decision-making. While current AI approaches often struggle with limited datasets and company-specific development hurdles, they offer potential advantages in speed, scalability, and objectivity. The most promising near-term applications may be in business development, where political considerations are reduced and external assessment is more readily accepted.
The future of pharmaceutical portfolio management likely lies in combining the structured human expertise pioneered by Andersen with the analytical power of AI. Success will require maintaining the explainability and cultural sensitivity of traditional approaches while leveraging technology to enhance the speed and consistency of pharmaceutical decision-making. As the industry evolves, the fundamental principles of good decision-making—expert judgment, cultural awareness, and rigorous risk assessment—will remain essential.
