The Milky Way satellite galaxies are part of a vast polar structure (VPOS), a thin plane also consisting of globular clusters and streams of disrupted systems. I will present an updated analysis of the orbital poles of the satellites based on their proper motions, revealing that most satellite galaxies co-orbit within the VPOS. In addition, it was recently discovered that about half of the satellite galaxies of the Andromeda galaxy define and mostly co-orbit in a thin plane. I will illustrate the relative orientations of the dwarf galaxy planes with a 3D-model, revealing a surprising amount of spacial order in the Local Group. The existence of similar, co-rotating phase-space structures in the two only satellite galaxy systems for which 3D positions are known emphasizes the need to develop an understanding of their origin. I will discuss why suggested formation scenarios which interpret the satellite galaxies as tracers of dark matter sub-halos being accreted along cosmic filaments fail to deliver a satisfactory explanation. As an alternative origin I suggest the formation of tidal dwarf galaxies (TDGs) in the phase-space correlated debris of interacting galaxies. However, this has far-reaching consequences. If a number of local dwarf galaxies are ancient TDGs, spuriously interpreting them to trace the dark matter sub-structure gives rise to inconsistencies with cosmological predictions, artificially attenuates the missing satellites problem and therefore seriously affects current near-field cosmology.