North star is ROAS. Colours compare each value to that cell's own median
(good bad
~median). CPS is secondary.
Which cell wins? Board C (ROAS) is the decider — Board A (cheapest fan)
and Board B (highest AEPS) often disagree. Only cells with
30+ subs qualify.
Goal-seek: what must each lever hit for the target ROAS? “Required” is shown
against what this cell has already proven — so you see what's realistic.
Turn one-off wins into a playbook. Before/after numbers are pulled
automatically from the data — you just log the change and the verdict.
ROAS is the primary delta.
Log a test
Hypothesis generator: which cells are ROAS stars or losers vs the
cross-cell median, and why — cost axis or value axis. The mismatch finder
catches cheap-but-losing traps and pricey-but-winning hidden gems. Findings are
correlations — verify them in the Test Log.
Your OnlyFans models and their Kürzel (short code used in every
naming convention, e.g. Skylar Brooks → SB). Sync pulls them from OnlyFans;
edit a Kürzel inline.