Case Studies

Every case study here shows the tracked metric before and after, the volatility band the change had to clear, and what was actually done.

That last part matters. An unchanged brandu2019s visibility swings a few points on its own from week to week, so any result smaller than that band is noise dressed up as a win. Nothing gets published here unless the change clears it. That rule is why these are worth reading.