I don't have a good answer for you since I don't do this, but here's some intel/long-term ideas I got from a reliable source
GSA is a core tool, learn it inside out
Formula? Never shared with me. Regardless, boils down to testing/optimizing anyway as Google algo changes over several months
Get familiar with the basics like tier linking, building web 2.0s, familiar with spending money on linking services, whatever
The rest is simply coming up with two methods with difference of only 1 element (i.e., the way u structure ur tier linking), having two test subjects (two sites), then treat it like an a/b test
Find the peak time for your niche/audience throwing money at you, then blast it before the peak period
Take notes; Learn how fast things go and blow up, then play the timing game
Your objective is not about this 1 site, but the process and hardcore experimentation of a portfolio of properties - that's the big picture, and get you into a black hat badass in no time
Hope this was useful