Both industrial and academic research produce innovations that can be used to add value to products and services. Many labs seek ways to produce innovations that 1) have higher value, 2) are faster to market, and 3) are produced as a consistent stream. We have found that by treating the innovation process itself like a chemical process, the flowrate of valuable innovations can be increased. In this talk we analyze the innovation process according to streams (e.g., desired outputs, inputs, recycles, purges) and unit operations (e.g., idea generation âreactorsâ and idea evaluation âseparatorsâ). This perspective enables the analysis and measurement of R&D, providing a coherent philosophy for innovation processes.