Background: In team science, understanding how external funders shape interdisciplinary collaboration and learning is essential. This is particularly salient as multi-team systems (MTS), networks of interdependent teams pursuing shared outcomes within a broader mission, are increasingly deployed to address and solve grand challenges. This longitudinal study investigates how active funder-initiated influences affect an MTS of interdisciplinary biomedical teams, with a focus on a large project team of teams (Project N) and, in turn, how these influences affect team dynamics, processes, and learning outcomes.

Methods: We conducted a mixed-methods case study of “Project N,†one multi-team project in a multi-project initiative to generate high-quality datasets for machine learning in biomedicine. Data were drawn from quantitative team-dynamics surveys that measured five constructs (psychological safety, teamwork, leadership, shared leadership, and satisfaction) across three organizational levels (6 primary work teams, 1 project team, and consortium) at three time points. We explored survey results using qualitative sources, including archival documents and semi-structured interviews with team members, developing a categorical framework of funder-initiated influences. We coded team responses and outcomes for each influence type.

Results: Survey results suggest that team dynamics were generally strongest within primary work teams and weakest at the consortium level, and that team dynamics declined in the second survey before improving somewhat by the third. Qualitative analysis sought to understand funders' influence on teams' worse-before-better experience. We identified three categories of funder influences: structural (award mechanism, project design), recurrent (changes to deadlines and deliverables), and acute (sudden, unexpected events that disrupted workflows). Teams demonstrated adaptation and resilience by leaning into success, reinforcing team identification around common goals, and enhancing communication. These adaptive responses supported learning and performance in the face of funder-initiated influences.

Conclusions: Active funder influences can significantly shape team learning, coordination, and outcomes in interdisciplinary MTS. Our findings highlight how project teams' adaptive behaviors in response to changes from active funders helped them build resilience. They also underscore, for funders, the importance of considering the timing and nature of their influences to promote optimal team learning and outcomes. This study contributes to our understanding of funder-grantee dynamics and research team effectiveness in MTS over time.