(thanks to Dr. Charlotte Angus - https://charlotte-angus.com/)
Telluric absorption features around ~6300A and ~6500A are often present in reduced spectra. Sometimes classifiers will treat these features as real and try to fit them against templates, sometimes skewing the results.
Many classifiers will provide an approximate age of the transient based on the templates it fits. It is always good practice to look at the light curve of the transient before you report this i.e. if the fit says +15 days but the light curve is clearly at max, then perhaps this is not the best fitting template
NB. exceptions come for classes of SNe with only a few spectral templates - usually unusual transients like SLSNe, Ca-Rich SNe. Also for SN-II, the evolution of the spectral features does not correlate well with age of the SN (e.g. a type IIP SNe looks the same at +9 days as it does at +50 days), so template age is not always reliable.
Most classifiers determine the redshift of the transient. If you have a good fit (i.e. one that’s visually good and matches the general features of your spectrum), then the redshift is normally good to within z +/- 0.1 (this is because your transient features typically have a broad velocity)
It is good practice to report the host galaxy redshift if you can find it. The easiest way to do this is to see if the galaxy has a pre-reported redshift (check NED/SDSS/Simbad etc)
Otherwise you can look for host galaxy emission or absorption features. These will be narrow (widths of the order ~10A). The best ones to look for are:
Emission lines (rest λ):
H-alpha: 6562 A
H-beta: 4861 A
O II: 3726 and 3729 A
O III: 4959 and 5007 A
Absorption lines (rest λ):
Ca H & K: 3934 and 3969 A
Na-D: 5891 and 5897 A
z = (λ(observed) / λ(rest)) - 1
At late times (>~ +15 days) it becomes difficult to distinguish between the spectra of SN-Ia and SN-Ic at the same redshift (the majority of features are very broad and similar) which makes it difficult to robustly classify at this phase.
One sanity check is to look at the lightcurve and determine the peak luminosity based on the redshift. If it’s within the SN-Ia luminosity range (~-18.5 to -19.5) then it’s more likely to be a SN-Ia but this is not definite!