Most content testing optimizes opens among the current audience. Direction testing asks a better question: which kind of content attracts the people I want more of?
Setup
- Define 2–3 directions honestly distinct enough to attract different people (e.g. tactical checklists vs. business teardowns vs. personal essays)
- Tag every send with exactly one direction
- Run each direction at least 3 sends before judging — single-send noise is brutal
Read the right metric
Per direction, compare: open rate (current-audience fit), purchases (commercial fit), and new engaged readers — people whose first-ever open landed on that direction. The third number is the steering wheel; the first is just the speedometer.
The decision rule
- Direction wins on new engaged readers AND those readers match your target avatar → shift the calendar toward it
- Direction wins opens but attracts the wrong profile → it's comfort food; ration it
- Direction loses everything → kill it without ceremony after send #4
Review monthly. Audiences drift, and last quarter's winning direction quietly stops pulling. The point isn't finding the eternal answer — it's having a steering wheel at all.
Run this on your own list.