Blog
How do you test curiosity-gap hooks without a new edit ticket each time?
How do you test curiosity-gap hooks without a new edit ticket each time?
Curiosity hooks hide the payoff for a beat. To test them, lock the body and swap only the opening. You do not need a new editor ticket per line. Assemble those openings in eonik on Mac, approve, and read 3-second plays against your baseline; Meta formulas only; No invented lift.
How do you test curiosity-gap hooks without a new edit ticket each time?
Curiosity hooks hide the payoff for a beat. To test them, lock the body and swap only the opening. You do not need a new editor ticket per line. Assemble those openings in eonik on Mac, approve, and read 3-second plays against your baseline; Meta formulas only; No invented lift.
What a curiosity gap is doing
The opening withholds the answer the viewer now wants: a result, a name, a number you will show in the body. If the body does not pay it off, the gap is clickbait and hold collapses.
Write three openings that ask different questions. They must all earn the same proof block. If they cannot, you do not have a hook test. You have three ads.
Isolate and read
Same body, same kit, same placement. Swap hook only. Hook rate is 3-second video plays ÷ impressions (operator ratio). 3-second plays: ≥3s or nearly full length if shorter; per impression; replay excluded; Source: Meta Ads Help.
Hold: ThruPlays ÷ 3-second plays. ThruPlay: completion or ≥15s (≥97% if shorter: Meta documents that, not a benchmark); Compare to yourself, same window; Do not cite a fake 25% benchmark.
Assemble without the ticket
Record or generate three hook takes and drop them on the locked body in eonik. Captions from kit. Approve, then upload. Operator read: hook rate.
If a line needs a different B-roll, it is no longer a pure hook test. Either recut the body as a new experiment or rewrite the line so the same body still pays it off. Writing the gap: curiosity-gap hooks.
Curiosity hook cell. What to lock
| Layer | Lock | Swap |
|---|---|---|
| Opening | Must earn the same body | Question, withhold, poster |
| Body | Proof, offer, CTA | Do not rewrite per hook |
| Kit / music | Type, colour, logo, captions | Only if type is the test |
| Readout | Same account, window, placement | Hook rate, then hold |
Hook rate = 3-second video plays ÷ impressions. Hold = ThruPlays ÷ 3-second plays. Meta documents the plays, not a “good” percentage.
What usually breaks
- Curiosity openings that the body never answers.
- Changing music and hook together.
- Reading hook rate across placements as if they were one test.
Consider three questions the same proof can answer. Make those hook cuts in eonik, approve them, and read them yourself. how to generate AI ads.
“Almost any question can be answered, cheaply, quickly, and finally, by a test campaign. And that’s the way to answer them, not by arguments around a table.”
That principle still holds. The next on-brand cut is yours to approve.
Sources
Questions
Do I need an editor for curiosity tests?
Not for swapping openings on a locked body with kit already on. Record or generate three hook takes, drop them on the same proof, approve, upload. You still need craft if the body itself is a film, or if a line demands different B-roll and stops being a pure hook test.
Will curiosity hooks convert better?
Nobody can promise that. If the body does not pay off the gap, hold collapses and you shipped clickbait. Write openings that all earn the same proof. Read 3-second plays against your own baseline; eonik does not predict winners; You still decide in Ads Manager.
Where are the Meta formulas?
3-second video plays and ThruPlay are defined in Meta Ads Help. Linked in sources on this page. Hook rate and hold rate are operator ratios, not official benchmarks. There is no universal good percent; Compare the same account, window, and placement, not a folklore table.
Free, no card · macOS 15+ · Apple silicon