Ramp new variants 1% → 5% → 50%, not straight to half your revenue.

Ramp new variants 1% → 5% → 50%, not straight to half your revenue.
The goal of the early phase isn't power — it's catching a catastrophe cheaply.
— Day 0-1 at 1%: watch error rate, latency, JS exceptions, hard checkout failures. Not conversion.
— Day 1-2 at 5%: confirm guardrails are stable, no SRM.
— Then 50/50 for the actual measurement window.
— Abort criteria are operational (variant throws errors) in the ramp, statistical only at full allocation.
A broken variant at 50% costs ten times what it costs at 5%. Read the number, not the story. [ramp 1/5/50%]