Aims. The TOPGöt project studies a sample of 86 high-mass star-forming regions in different evolutionary stages from starless cores to ultra compact HII regions. The aim of the survey is to analyze different molecular species in a statistically significant sample to study the chemical evolution in high-mass star-forming regions, and identify chemical tracers of the different phases. Methods. The sources have been observed with the IRAM 30 m telescope in different spectral windows at 1, 2, and 3 mm. In this first paper, we present the sample and analyze the spectral energy distributions (SEDs) of the TOPGöt sources to derive physical parameters such as the dust temperature, Tdust, the total column density, NH2, the mass, M, the luminosity, L, and the luminosity-to-mass ratio, L∕M, which is an indicator of the evolutionary stage of the sources. We use the MADCUBA software to analyze the emission of methyl cyanide (CH3CN), a well-known tracer of high-mass star formation. Results. We built the spectral energy distributions for ~80% of the sample and derived Tdust and NH2 values which range between 9−36 K and ~3 × 1021−7 × 1023 cm−2, respectively. The luminosity of the sources spans over four orders of magnitude from 30 to 3 × 105 L⊙, masses vary between ~30 and 8 × 103 M⊙, and the luminosity-to-mass ratio L∕M covers three orders of magnitude from 6 × 10−2 to 3 × 102 L⊙∕M⊙. The emission of the CH3CN(5K-4K) K-transitions has been detected toward 73 sources (85% of the sample), with 12 nondetections and one source not observed in the frequency range of CH3CN(5K-4K). The emission of CH3CN has been detected toward all evolutionary stages, with the mean abundances showing a clear increase of an order of magnitude from high-mass starless cores to later evolutionary stages. We found a conservative abundance upper limit for high-mass starless cores of XCH3CN < 4.0 × 10−11, and a range in abundance of 4.0 × 10−11 < XCH3CN < 7.0 × 10−11 for those sources that are likely high-mass starless cores or very early high-mass protostellar objects. In fact, in this range of abundance we have identified five sources previously not classified as being in a very early evolutionary stage. The abundance of CH3CN can thus be used to identify high-mass star-forming regions in early phases of star-formation.