An email is not a customer
Cost per lead is something you measure. Cost per customer is something you guess. Keep the two apart.
Mostly arithmetic, occasionally uncomfortable. Everything here is worked through with the same functions that run your experiment, so if a claim is wrong the product is wrong too.
Each note takes one common shortcut, runs the numbers, and ends with the move the evidence supports.
Cost per lead is something you measure. Cost per customer is something you guess. Keep the two apart.
The most expensive mistake in cheap demand testing is treating an empty inbox as an answer.
A small budget buys a real answer to a smaller question. Pretending otherwise is how these tests get a bad name.
Four landing-page variants on a small budget produce four uninterpretable results instead of one useful one.