While there have been rapid advances in assessments of biodiversity and ecosystem services (BES), a critical remaining challenge is how to move from scientific knowledge to real-world decision making. We offer 6 lessons from our experiences applying new approaches and tools for quantifying BES in 20 pilot demonstrations: (1) Applying a BES approach is most effective in leading to policy change as part of an iterative science-policy process; (2) simple ecological production function models have been useful in a diverse set of decision contexts, across a broad range of biophysical, social, and governance systems. Key limitations of simple models arise at very small scales, and in predicting specific future BES values; (3) training local experts in the approaches and tools is important for building local capacity, ownership, trust, and long-term success; (4) decision makers and stakeholders prefer to use a variety of BES value metrics, not only monetary values; (5) an important science gap exists in linking changes in BES to changes in livelihoods, health, cultural values, and other metrics of human wellbeing; and (6) communicating uncertainty in useful and transparent ways remains challenging. •We evaluated applications of ecosystem service information in 20 decision contexts.•Simple production function models can help inform a diversity of decisions.•Ecosystem service values requested go beyond monetary metrics.•Training in use of simple tools helps build capacity for further process iteration.•Gaps exist in human wellbeing metrics and treatment of uncertainty.