The ability of measuring surface ocean chlorophyll concentrations from space has revolutionized studies of large-scale ocean biology and ecology. Yet despite decades of efforts from the research community, small discrepancies between different sensors in their estimated chlorophyll at global and regional scales do exist, often in the order of 5-10%. These artifacts create difficulties in interpreting inter-annual changes and long-term trends. Here I demonstrate how two coastal events led to a new algorithm concept, which removes these discrepancies significantly so cross-sensor consistency is much improved. In addition, quality of individual images is also improved in terms of reduced noise and more coherent features. The new algorithm has been implemented by NASA to process data collected by all ocean color satellite missions for most of the global ocean. The study provides a clear example of how thinking out of the box may lead to surprises.