A quick opinion, and I’m no expert by any means:
A compressor will make the louder bits quieter and the quieter bits louder (laymans terms). So think funky guitar that has a huge range of volume dynamics. Compressed this will even out the levels making it easier to mix.
So, it makes sense in this case to compress prior to automation.
The same could be argued re reverb - makes things sound further away…
My technique in a few lines:
Write the music.
Set up group tracks for say Drums/perc, strings, synth pads, bass, vocals etc.
Solo each group and get your levels right for each group - and eq’s etc
Set up a marker track for each section of your song (intro, verse, chorus, outro)
Start setting levels for all groups for each section (let each one loop)
Write a track sheet- notes for each track for things like the level and pan in each song section.
Apply automation to suit to your volumes etc.
This doesn’t take into account any work prior to mixing of choosing your synth sounds, exporting your midi/VST/instrument tracks to audio, adding effects such as compression, reverb etc.
Simply - don’t start automating a mix until the rest of the hard work is done.
Don’t spend too long on it due to ear fatigue. Do a bit, walk away. Have another listen, do a bit more.
And this IS the general forum - there is a forum for bugs, issues etc, so a general forum is where we can share things like this.
Finally - backup regularly and keep saving different versions of your track.
I’m sure that everyone else will have slightly different methods.
Hope that helps.