Data Study
We measured predicted attention in the first 3 seconds of 76 proven TikTok ads. Here's the uncomfortable finding: the ads with the strongest relative openings mostly ranked near the bottom of the benchmark, and the #1 ad overall opened at just 25/100. The data says a hook is necessary — and wildly insufficient.
| Ad | Hook | Overall rank | Opening |
|---|---|---|---|
| Denis Beck Fitness | 96/100 | #63 of 76 | “my friend got broken up with yesterday so I had to give him the grip strengthener” |
| Alex Stien | 93/100 | #54 of 76 | “Okay, next time you do this again, I'm keeping it.” |
| AI companion app | 86/100 | #72 of 76 | “Maybe it's time for something real…” |
| Evony / DoorDash creative | 84/100 | #47 of 76 | Spanish-language story open |
| Rivue Apartments | 81/100 | #76 of 76 | (music only — opens on its best shot) |
Five monster hooks; not one cracked the top half of the leaderboard. Meanwhile StickyLint (#1 overall) opened at 25/100 and MarkOgier (#4) at 2/100.
Attention curves are normalized to each ad's own range, so "hook = 96" really means "the opening was nearly this ad's best moment." That's the problem. The bottom three ads in the entire benchmark all peak at second zero and decay from there — the shape of a great thumbnail attached to nothing. Across all 76 ads, relative hook strength correlated negatively with overall rank (r = −0.52): the top 20 ads averaged a hook of just 8/100 relative to their own later peaks, because their later peaks were huge. The winners' median peak lands 57% of the way through the ad.
None of this says hooks don't matter — on a paid feed the swipe decision still happens in the first second. It says the hook's job is narrower than the industry treats it: buy the next 3 seconds, then hand off to something that escalates. The pattern to copy from the top of the leaderboard is open with a competent question or action, then build — and the pattern to avoid is spending your best material before second three. If you're drafting openings, pair this with how to write a TikTok hook and ad hook examples.
A PreTestAds report gives you both numbers this study is built on: Hook Strength for your first 3 seconds, and the full attention curve showing whether the ad builds or decays after the open. If the curve peaks at second one, you've made a thumbnail — recut before you pay for impressions. You can test a hook variant without reshooting anything; see how to test a hook without filming.
Score your creative and see whether attention builds after the open. First analysis free.
Score My HookA hook's job is to stop the swipe — but brain data from 76 top TikTok ads shows a hook can't carry an ad. Ads whose first 3 seconds were their attention high-point tended to rank near the bottom of the set overall, because everything after the open decayed. The winners opened competently and built to a bigger moment later. You need a hook AND an escalation.
In a 76-ad benchmark, relative hook strength correlated negatively (r = −0.52) with overall engagement rank. That's not because good openings hurt — it's because attention curves are relative to each ad's own range. An ad whose opening is its peak has nowhere to go but down; an ad that builds makes its opening look small by comparison. Peaking early is the real failure pattern.
Upload the ad to PreTestAds and check two things in the report: Hook Strength (average predicted attention in the first 3 seconds) and where the curve goes afterward. A strong hook followed by a crash means you have a thumbnail, not an ad. It predicts attention, not conversions.