Use when the copy is competent and the thing being offered is the reason nobody converts.
Replace each placeholder with your own detail. The more specific you are, the less the model invents.
When the copy is competent and still nothing converts. At that point rewriting headlines is rearranging the description of something people do not want. This works on the offer itself: what is promised, how quickly, how much effort and risk the buyer carries, and why now. It explicitly refuses to rewrite your ads.
The current offer copied word for word off the page rather than as you intend it, since the gap between the two is frequently the whole problem. Then {{MARGIN_ROOM}}, meaning what you can genuinely give away, the reasons prospects actually state for not buying, and what competitors offer at your price point.
A teardown of the current offer against six components marked present, vague or absent, three rebuilds each changing a different component, a stress test table, and the one to run first with its deciding metric. The column asking what breaks if twenty per cent take it up is the one to show finance before any guarantee ships.
Accepting three rebuilds that are all price changes. A discount moves one component and leaves the other five untouched, so check that at least two rebuilds change something other than price. Whether discounting damages a brand over time is argued both ways, but as a test of your offer it tells you very little.