The hook is the whole funnel
On a feed, the first moments decide whether anything else you made gets seen. Everything downstream — the demonstration, the offer, the call to action — is contingent on surviving that gate, which makes the hook disproportionately worth attention relative to its runtime.
Teams routinely invert this, spending most of the production effort on the body of a piece that most viewers never reach.
What cheap generation changes
Historically you wrote one hook because shooting five was expensive. That constraint is gone: the same body can be paired with a dozen openings for close to nothing.
That should change the process, and mostly it has not. The habit of writing one hook and committing survives the disappearance of the reason for it.
Hook types worth having in rotation
The problem stated plainly, before any product appears. The outcome shown first, working backwards to how. The specific objection named out loud — usually the one your reviews complain about. The visual anomaly that is simply hard to scroll past. The direct address to a named situation, which trades reach for immediate relevance.
The point of a taxonomy is not that one type wins; it is that they fail differently across audiences, so a rotation tells you something a single execution cannot.
Test the hook alone
Changing the hook and the body together produces a result you cannot attribute. Hold the body constant and vary only the opening, and the comparison means something.
This is the single discipline that turns creative volume from noise into information, and it costs nothing beyond deciding to do it.
The honest limit
A hook cannot rescue a proposition nobody wants. Testing openings on an offer that does not land produces a well-optimised path to the same disappointing outcome. When every variant underperforms similarly, the creative is not the problem, and continuing to test it is a way of avoiding a harder conversation.