Presented 12 May, 2023 at the Center for Open and Reproducible Science (CORES) Annual Meeting 2023:Open Science: Intersections with Diversity, Inclusion and Equity. In this talk, I argue that community-driven methods development provides necessary structure for creating robust and reliable inferences in experimental science. I consider the role of open-source scientific software to date, highlighting features of community-driven software that encourage its relative success. I then draw on two software projects from my own research as case studies in the benefits of focused community engagement, while underscoring the challenges in pursuing this work under misaligned incentive structures. Overall, I encourage researchers to carefully consider the role that software development plays in modern experimental science and the interaction of these technical advances with social challenges.