The weather enterprise must continue to innovate and expand to keep people safe from disruptive weather, and to keep up with the demand for weather intelligence for economic development. Despite considerable scientific and technical advances, underinvestment in services, coupled with poor regulation and business practices, prevents the enterprise from fully protecting developing economies. New approaches are needed. The lean startup focuses on development with a minimum level of investment. While the build-measure-learn approach is used extensively in the private sector, we consider its application to NMHSs and how this might be used to prototype new capabilities quickly with modest investment in different components of the meteorological value chain. The process introduces new ways to manage projects under extreme uncertainty and benefits traditional institutions to strengthen the way they operate.