A campaign underperforms for one of four reasons and they need completely different fixes. Guessing wrong costs a fortnight.
The order is not arbitrary
If mail is not arriving, nothing about the copy is measurable. If the list is wrong, the offer is being tested on the wrong people. If the offer is wrong, the copy is competing for attention it was never going to get. Each layer has to be cleared before the one below it means anything.
Teams almost always start at copy, because copy is the visible part and the part everyone has an opinion about. That is why the same campaign gets rewritten three times before somebody checks the bounce rate.
Reading the replies properly
Classification counts are a summary and summaries hide the finding. Twelve people saying no is a number. Twelve people saying no because they thought you were selling something you do not sell is a positioning problem, and it is only visible in the text.
So every reply is read in full during a deep dive, not sampled, not summarised.
It zooms in
This is deliberately one campaign at a time. The cross-campaign view, which angle wins across everything and which objections recur portfolio-wide, is a separate and much larger read.
Runs well with
Skills compound. These are the ones we usually install alongside it.