TikTok hooks that convert: what to do with the first three seconds
Almost every underperforming UGC ad fails in the opening frame, not the body. These are the hook structures we return to most, and how to test them without burning budget.
7 min read
Why the hook decides the result
On TikTok and Meta, the majority of viewers who will ever leave your ad leave in the first three seconds. Everything you spent on production is judged against that window, which is why two videos with identical bodies and different openings can produce wildly different costs per acquisition.
Treat the hook as a separate deliverable from the video. Same footage, several openings, tested against each other. That single change usually produces more improvement than a bigger production budget.
Nine hook structures that work
The problem callout: name the frustration out loud in the viewer's own words. Works when the pain is specific and immediately recognisable.
The negative hook: 'stop buying X until you've seen this'. Pattern-interrupts because it contradicts what an ad is expected to say.
The result-first hook: open on the outcome, then rewind to how it happened. Strong for beauty, home and fitness.
The curiosity gap: 'I didn't expect this to work' — enough tension to earn another two seconds without over-promising.
The comparison: your product against what the viewer currently uses, side by side, no commentary needed.
The listicle: 'three things I wish I knew before buying' — implies a payoff and sets up sequential watching.
The direct recommendation: creator straight to camera, talking to a friend. Simple, and consistently underrated.
The demonstration: start mid-action with the most visually satisfying moment — the pour, the peel, the mix, the fit check.
The objection hook: raise the doubt the buyer already has, then answer it. Effective in sceptical categories like supplements.
How to test hooks properly
Test three to five hooks per concept, keeping the body constant. If you change the hook and the edit at once, you learn nothing about either.
Give each variation enough impressions to be meaningful before judging it, and watch three-second and fifteen-second retention alongside cost per acquisition. Retention tells you whether the hook worked; CPA tells you whether the rest of the video followed through.
When a hook wins, don't just scale it — write four more variations of the same structure. Winning structures usually generalise across products in a way that specific scripts don't.
Common hook mistakes
Starting with branding. A logo or product beauty shot in the first second reads as an ad and gets skipped.
Front-loading too much information. One idea per hook. Anything else belongs in the body.
Over-produced delivery. If the opening line sounds performed, the credibility you're borrowing from the creator disappears immediately.
Ignoring subtitles. Most of the feed is watched with sound off at least some of the time, so the hook has to work as text on screen too.