Our engine
Four steps, one week per turn. People decide what to test. Automation builds it. Every round starts from what the last one proved.
Every round starts with a persona: who they are, what they have already tried, what will make them stop. Mapped from your best seller and your past results.
First creative lands in 3 to 5 days. After that a fresh drop ships every week, named and tracked down to the variant.
Spend comes back as signal. Every variant is scored against targets we set together, so a win means the same thing to both of us.
Losers get a note on record. Winners get remixed, and we tell you which to scale. Round five opens knowing everything round four proved.
Variants across every persona worth winning. The baseline is set.
Losers come down with a reason on record. Survivors are remixed.
We double down on what is working and cut what is not, with the reasons written up.
Every new variant starts from three weeks of proof. That is the engine compounding.
No upselling. No fluff. Proof before we expand.