The use of heterobimetallic metal complexes as molecular single-source precursors is a promising strategy for the targeted synthesis of phase-pure stoichiometric ternary metal oxide nanocrystals. However, the design and synthesis of these precursors is not trivial and can require considerable effort. Using spinel metal ferrite nanocrystals of formula MFe2O4 (M = Fe2+, Co2+, Ni2+, Cu2+, and Zn2+) as a model system, this paper evaluates the efficacy of the single-source precursor approach by comparing directly nanocrystals synthesized from the solvothermal reaction of heterobimetallic trinuclear oxo-bridged clusters of formula MIIFeIII2(μ3-O)(μ2-O2CR)6(H2O)3, R = CF3 to nanocrystals synthesized from the solvothermal reaction of stoichiometric mixtures of multi-source precursors, such as metal acetate or nitrate salts. For each M explored here, the clusters form phase-pure MFe2O4 nanocrystals with significantly narrower size distributions than nanocrystals synthesized from multi-source-precursors. In the case of M = Cu, the multi-source metal salt precursors produce a mixture of CuO and CuFe2O4. Additionally, changing the electronic nature of the R-group on the bridging carboxylate ligand from electron withdrawing (CF3) to electron donating (CH3 or C(CH3)3) decreases the average diameter of the resulting nanocrystals by a factor of two. The cluster molecules therefore offer superior control over both morphology and composition for transition metal ferrite nanocrystals. We hypothesize that this remarkable performance arises from the presence of pre-formed M2+-O-Fe3+ moieties in the cluster molecules that enable direct nucleation of MFe2O4 and preclude nucleation of binary oxide impurities.