Some of us working on StarLink 6-8 years ago had spreadsheets with basic cost assumptions of launch, sats and ground terminals, as well as revenue projections. The general consensus was, wow, the ROI is really quick, lets build and launch as many of these as fast as we can