The challenge of managing the fuzzy front end of the innovation process is particularly acute for large, multi-brand, research and development (R&D)-intensive firms. Poor performance at generating radical innovations has resulted in many large organisations seeking to innovate how they organise for innovation. This paper presents an inductive, longitudinal study of an organisational experiment that sought to get ‘game-changing, radical ideas’ into the new product development funnel of a top three pharma. The immediate outcomes of a team-based internal innovation tournament included 33 new product ideas, 14 of which were radical. The medium term outcome of the experiment was a reorganisation of how the firm now pursues radical innovation activities. We link these outcomes to team leadership, contrasting innovation processes, including decisions about how to incorporate the ‘voice of the consumer’. The inductive, longitudinal study suggests causal interconnections between innovation team leadership, innovation team processes, and innovation outcomes.