If LCDM is correct, then all dark matter halos hosting galaxies, from those hosting dwarfs to those hosting giant clusters, are filled with abundant substructure down to very low mass scales (<< 10^9 Msun). Specifically, even the dark matter halos of Local Group field dwarfs should be filled with subhalos, many of which should be fairly massive (~ 10^8 Msun), and thus are potential targets for hosting small (ultrafaint) galaxies. Here we make predictions for the existence of ultrafaint satellites of dwarf galaxies using the highest resolution cosmological dwarf simulations yet run (Mgas~ 250 Msun). We simulate four halos -- two each at the mass of classical dwarf galaxies (Mvir ~10^10 Msun) and ultrafaint galaxies (Mvir ~ 10^9 Msun) -- down to z=0 using the GIZMO (Hopkins 2014) code. This code relies on state-of-the-art MFM hydrodynamics and implements the FIRE (Feedback in Realistic Environments) recipes (Hopkins et al. 2014) for converting gas into stars and capturing the energy fed back from those stars into the surrounding medium. We predict that ultrafaint galaxies (M* ~ 3,000 Msun) should exist as satellites around more massive dwarf galaxies (M* ~ 10^6 Msun) in the Local Group. These tiny satellites, as well as the two isolated ultrafaints, have uniformly ancient stellar population (> 10 Gyr) owing to reionization-related quenching. The more massive systems, in contrast, all have late-time star formation. Our results suggest that Mhalo ~ 5 x10^9 Msun is a probable dividing line between halos hosting reionization "fossils" and those hosting dwarfs that can continue to form stars in isolation after reionization. Importantly, we show that the extended ~50 kpc regions around Local Group "field" dwarfs may provide efficient search locations for discovering new ultrafaint dwarf galaxies, and discuss the prospects for their discovery in light of the new generation of large surveys and giant telescopes. If these tiny satellites are observed, this would provide evidence that dark matter substructure is truly hierarchical, as predicted in the standard paradigm.