Learn
How do you write and test curiosity-gap hooks without turning them into clickbait?
How do you write and test curiosity-gap hooks without turning them into clickbait?
Write and test curiosity-gap hooks without turning them into clickbait by withholding a payoff the body must actually deliver. Draft several openings that ask different questions, all landing on the same proof. Assemble those swaps in eonik, approve, and read 3-second plays on your baseline. If the body does not pay it off, hold dies; Cite Meta for formulas only.
How do you write and test curiosity-gap hooks without turning them into clickbait?
Write and test curiosity-gap hooks without turning them into clickbait by withholding a payoff the body must actually deliver. Draft several openings that ask different questions, all landing on the same proof. Assemble those swaps in eonik, approve, and read 3-second plays on your baseline. If the body does not pay it off, hold dies; Cite Meta for formulas only.
Write the gap, then the payoff
Draft the body answer first: the demo, the number you can defend, the named workflow. Then write openings that create a question that body answers in the next breath.
If the opening teases a celebrity, a price, or a result the body never shows, it is clickbait. Rewrite or drop it.
Test like a hook cell
Lock body, kit, and music, then swap curiosity openings in eonik and upload. Hook rate is 3-second video plays divided by impressions; hold rate is ThruPlays divided by 3-second plays. Formulas from Meta Ads Help. Omit invented hold percentages.
If hook is high and hold is poor versus your baseline, the payoff failed. Fix the body or stop using that gap family. Assemble the swaps as hook variants; you still sandbox in Ads Manager.
Curiosity gap vs clickbait
| Opening | Honest gap | Clickbait |
|---|---|---|
| Promise | A question the body answers next. | A celebrity, price, or result you never show. |
| Body | Same proof for every opening. | A new demo per hook. |
| Hold | ThruPlay among people who started. | Dies after the tease. |
| Test | Swap openings in eonik. Lock body. | Not a hook cell. Three ads. |
Cite Meta for 3-second video plays and ThruPlay. Omit a “good curiosity” percentage. There is not one.
What usually breaks
- Teasing a payoff you cannot show.
- Curiosity plus a new demo per hook (not isolated).
- Inventing a benchmark for “good curiosity.”
Consider the next cut. Make it in eonik on Mac, approve every frame, then upload yourself. The loop is how to generate AI ads. eonik never spends.
“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
Is curiosity better than PAS?
It is a different opening. Test against your other families. Do not declare a 2026 winner. Write the payoff first, then curiosity-gap hooks that all share B-roll; Isolate them like any hook cell: lock body, kit, music; Assemble in eonik, approve, upload; Read hook rate and hold on your baseline; Legal claims still need substantiation; Curiosity is not a loophole.
How do I measure the gap?
Same as any hook: 3-second plays for stop, ThruPlay for hold, your baseline. Cite Meta definitions. Hook rate is 3-second video plays divided by impressions. Hold is ThruPlays divided by 3-second plays; If hook is high and hold is poor, the payoff failed; Fix the body or stop using that gap family; There is no official “good curiosity” number; Omit invented benchmarks.
Can eonik write curiosity lines?
An assistant can draft. You own payoff, claims, and approval. If the opening teases a celebrity, a price, or a result the body never shows, it is clickbait. Rewrite or drop it; Spoken and on-screen should agree; Captions must not spoil the gap in frame one unless that is the test; Make the cuts in eonik; You still decide in Ads Manager.
Free, no card · macOS 15+ · Apple silicon