If you change the question and if you want to synthesize a new compound, the Gaussian program is very useful. It is possible to calculate the transition state existing in the reaction path between the starting material and the target compound. If you know the chemical structure of the transition state and the height of its potential, the degree of difficulty of synthesis of the target compound will become much more precise. Since the structure of the transition state can not be observed in the experiment, the Gaussian program will provide you with very important information for the target compound. Furthermore, since the chemical shift of NMR of the expected target compound can also be calculated by the Gaussian program, information on whether the synthesized compound is a target compound or not can also be obtained before synthesis.
I usually benchmark methods and basis sets against compounds that are similar and have a lot of experimental data available as well as high quality calculations. After you find a good compromise for the model system its good practice to at least try to see how basis set size affects the properties you care about in your novel system.
If you change the question and if you want to synthesize a new compound, the Gaussian program is very useful. It is possible to calculate the transition state existing in the reaction path between the starting material and the target compound. If you know the chemical structure of the transition state and the height of its potential, the degree of difficulty of synthesis of the target compound will become much more precise. Since the structure of the transition state can not be observed in the experiment, the Gaussian program will provide you with very important information for the target compound. Furthermore, since the chemical shift of NMR of the expected target compound can also be calculated by the Gaussian program, information on whether the synthesized compound is a target compound or not can also be obtained before synthesis.