Fuel surrogates are mixtures of simple compounds that emulate the combustion characteristics of more complex fuels, with the primary objective to enable detailed combustion modeling of very complex real fuels. Current efforts in surrogate development aim at optimizing the compositions of pure hydrocarbons to emulate multiple combustion related properties. In doing so, weights are assigned when defining optimization problem to reflect the importance of each property. In this study, we report on the relative importance of species selection and their weights on the overall performance of the optimized surrogate. Using experimental data of a reference jet fuel as target, we designed a study using a surrogate optimizer that imposes orthogonal perturbations on the surrogate components and weights and analyzed their impact on the optimized surrogate mixtures. Results from 3600 cases show that perturbations of surrogate components, rather than weights, induce far greater variability in the optimized composition and target property agreement. While the Derived Cetane Number (DCN) agreement shows a greater variability from the weight perturbation, the main reason for such high sensitivity is due to the wide range of values for pure component DCN of the individual components, which is also a result of the surrogate component selection. Further, the results show that the selection of surrogate components nearly predefines the overall shape of the distillation curves regardless of the weight values. The current study quantitatively supports the idea that appropriate selection of surrogate components that capture the physical and chemical characteristics of actual constituents of target fuel will increase the possibility of successful surrogate formulation and will mitigate the impact from arbitrary weight assignment.