Design the experiment and define the success criterion
Packaged know-how that tells an agent how to do a job well.
You might say…
“I keep 'testing' things by feel and never actually knowing what moved what — I need to design the experiment properly before I run it.”
What it does
Frames one focused test — isolates the single variable, writes the hypothesis, and sets the success criterion and run window — before anything launches.
Trigger: Use when a growth hypothesis is ready and needs to be turned into a time-bounded, single-variable test with a pre-defined success criterion.
I/O: Growth hypothesis + metric context → experiment design with variable, hypothesis, success criterion, and run window
Recognise the problem?
The primitives are the commodity part. The fastest next step is a conversation about composing them into something that works for you.
Start a conversation