There are no steadfast rules.
You can get some great sounds by breaking the rules, but the fact that you’re reading this in the first place tells me that you want some guidelines to start you off.
What is a good starting point for compressing vocals?
I would say that general wisdom tells us to compress for -1db to -5db most of the time. The ratios that I’ve seen for vocals usually fall between 2:1 to 6:1, depending on if you wanted to give a constant thickening of the vocals or if you wanted to simply tame the loud parts (which would call for a higher ratio.)
How does the sound of vocals change based on the threshold and ratio changes?
Check it out: