Electrochemical CO2 reduction (CO2RR) provides an economically feasible and environmentally sustainable pathway to transform atmospheric carbon dioxide into value-added fuels and chemicals. Copper is one metal that can catalyze this process but suffers from issues related to product diversity, high overpotential requirements, and poor stability. Cu–Ag bimetallic alloys offer an improved stability selectivity performance owing to a synergy between both components. In this study, we elucidate this functionality by mapping the relationship between the composition and coordination of metal sites on a CuAg surface with their preferred product profile. We parametrize a model that predicts the energy of metal atoms within a bimetallic nanoparticle as a function of their local composition and coordination environment. We choose CO and OH binding energies as descriptors for CO2RR and demonstrate their correlation with the site metal binding energies. We further construct plots forecasting the thermodynamic preference for CO2RR products as a function of descriptor energies and overlay predicted CO/OH binding energies on those plots. Through these comparisons, we infer the appropriate site compositions and design principles for electrocatalysts optimal for driving the process toward desirable multicarbon products.