Context. The link between the fate of the most massive stars and the resulting supernova (SN) explosion is still a matter of debate, in major part because of the ambiguity among light-curve powering mechanisms. When stars explode as SNe, the light-curve luminosity is typically sustained by a central engine (radioactive decay, magnetar spin-down, or fallback accretion). However, since massive stars eject considerable amounts of material during their evolution, there may be a significant contribution coming from interactions with the previously ejected circumstellar medium (CSM). Reconstructing the progenitor configuration at the time of explosion requires a detailed analysis of the long-term photometric and spectroscopic evolution of the related transient. Aims. In this paper, we present the results of our follow-up campaign of SN 2020faa. Given the high luminosity and peculiar slow light curve, it is purported to have a massive progenitor. We present the spectro-photometric dataset and investigate di fferent options to explain the unusual observed properties that support this assumption. Methods. We computed the bolometric luminosity of the supernova and the evolution of its temperature, radius, and expansion velocity. We also fit the observed light curve with a multi-component model to infer information on the progenitor and the explosion mechanism. Results. Reasonable parameters are inferred for SN 2020faa with a magnetar of energy, Ep = 1.5(-0.2)(+0.5) x 10(50) erg, and spin-down time, t(spin) = 15 +/- 1 d, a shell mass, M-shell = 2.4(-0.4)(+0.5) M-circle dot, and kinetic energy, E-kin(shell) = 0.9(-0.3)(+0.5) x10(51) erg, and a core with M-core = 21.5(-0.7)(+1.4) M-circle dot and E(kin()core) = 3.9(-0.4)(+0.1) x 10(51) erg. In addition, we need an extra source to power the luminosity of the second peak. We find that a hidden interaction with either a CSM disc or several delayed and choked jets is a viable mechanism for supplying the required energy to achieve this e ffect.