Organic-inorganic lead-based halide perovskite compounds have recently been utilized in solar cells with power conversion efficiency (PCE) of >23%. However, replacing the lead with less-toxic elements while maintaining high device performance remains a challenge. For this reason, significant effort has been directed toward the development of Pb-free compounds, including bismuth (Bi)-based methylammonium bismuth iodide (MA3 Bi2I9), but such systems continue to severely underperform when compared to the prototypical Pb-based methylammonium lead iodide (MAPbI3) perovskite. For the latter, and other Pb-based systems, it is known that Pb2+ complexes with polar solvents, such as DMSO and DMF, and solution-processed MAPbI3 and PbI2, form intermediate ordered precursor phases that incorporate these solvents into co-crystals during solution processing. Here, we compare and contrast the solidification and growth behaviors of Bi and Pb precursor inks based on the same solvents using multi-probe characterization methods. In both instances, we see evidence of a sol-gel process whereby solvent-metal complexes form and lead to a colloidal solution which solidifies. We show that the Bi-based compound crystallizes directly and rapidly into a textured polycrystalline microstructure from a precursor solution without evolving through intermediate crystalline solvated phases, in contrast to MAPbI3. This solidification process produces isolated crystals and challenges the growth of continuous and crystalline films required for application in solar cells. We reveal that solvent engineering in combination with antisolvent dripping are necessary to address this limitation and enable the formation of continuous polycrystalline films of Pb-free MA3Bi2I9 and functional solar cells. The work provides valuable insights linking the solid-state microstructure of the film and solar cell performance to the ink formulation and the solidification pathway.