Special theories like planar N=4 SYM have an underlying positive Grassmannian geometry that has allowed us to write expressions for tree-level amplitudes and loop singularities of arbitrary multiplicity. The extent to which such positive geometric structure extends to other QFTs is still not understood. We focus on a possible extension to N=8 SUGRA. We draw inspiration from the BCFW recursive construction of tree amplitudes in order to find "good" building blocks that evince hints of an underlying structure.