The study of galaxy formation and evolution using data from large-scale surveys relies to a large extent on an automated and robust application of spectral synthesis to derive physical properties of galaxies. However, most of the state-of-the-art codes used to decompose the galaxy optical spectrum into its elementary spectral components adopt mainly stellar ingredients. With this in mind, in this presentation we will investigate the impact of emission produced by an active galactic nucleus (AGN) and/or excited gas on the estimation of fundamental stellar properties of galaxies with currently available codes as a function, among others, cosmic time and star formation history. Subsequently, we will also see how the neglect of these non-stellar spectral components in the modelling can lead to an erroneous understanding of the formation and evolution of star-forming and/or active galaxies.