partly read ~ use ROI as a lens to decide which feature to build ~ problems in the mirror are larger than they appear ~ innovation is based on understanding the problem space better than anyone else