Abstract It could soon be technologically and economically feasible for anyone on Earth to have 2-way access to the global river of data that is modern civilization. Within the next 5 years, new low earth orbit (LEO) satellite constellations providing high-speed broadband may exceed the number of operational satellites currently in orbit. A fundamental question facing the new satellite constellation builders is whether the changes that have occurred in the 2 intervening decades are sufficient to alter the failed results of the 1990s-era constellations. The most likely answer is yes, but this depends on both political actions by governments and the business acumen of firms. Constellations must effectively target the demand segments, coordinate their spectrum usage with each other and the next generation of wireless networks, and avoid rampant orbital debris. As a side benefit, these new LEO constellations may provide a commercial setting for early use of reusable launcher capabilities.