Supernova experiments to characterize dark energy require a well designed low redshift program; we consider this for both ongoing/near term (e.g. Supernova Legacy Survey) and comprehensive future (e.g., SNAP) experiments. The derived criteria are: a supernova sample centered near z=0.05 comprising 150-500 (in the former case) and 300-900 (in the latter case) well measured supernovae. Low redshift Type Ia supernovae play two important roles for cosmological use of the supernova distance-redshift relation: as an anchor for the Hubble diagram and as an indicator of possible systematics. An innate degeneracy in cosmological distances implies that 300 nearby supernovae nearly saturate their cosmological leverage for the first use, and their optimum central redshift is z=0.05. This conclusion is strengthened upon including velocity flow and magnitude offset systematics. Limiting cosmological parameter bias due to supernova population drift (evolution) systematics plausibly increases the requirement for the second use to less than about 900 supernovae.