Hook rate on TikTok is the share of impressions that turn into a qualifying early view, and the median sits at about 33 percent, higher than the 28 percent typical of Meta feed and Reels. Anything under about 20 percent on TikTok in-feed means the first second is broken and nothing downstream deserves attention yet. The trap is treating it as a target: a hook rate above 40 percent is frequently a sign you have bought attention with a pattern interrupt that has nothing to do with what you sell.
What is a good hook rate on TikTok?
The figures below come from the platform medians in our 2026 fatigue benchmark, a synthesis of published platform and agency sources rather than our own measurement, plus the ranges we see day to day across accounts.
| Band | TikTok in-feed | What it means | Action |
|---|---|---|---|
| Under 15 percent | Broken | First frame doing nothing, or mis-targeted | Rebuild the opening, do not test around it |
| 15 to 20 percent | Weak | Recognisable as an ad in frame one | Five to eight new openings, same body |
| 20 to 30 percent | Normal | Working range for most in-feed accounts | Iterate, hook testing pays here |
| 30 to 33 percent | At or near median | Competitive | Move effort to completion and the offer |
| Above 40 percent | Suspicious | Often a pattern interrupt | Compare cost per acquisition, not hook rate |
For context: TikTok's 33 percent median sits above Meta Reels and Meta feed at 28 percent and well above YouTube in-stream at 22 percent. TikTok is not harder to hook on, it is easier, because delivery is full-screen with sound on and nothing competes in the viewport. The harder part is downstream, which is why TikTok's median video CTR is about 0.84 percent against Meta feed's 1.62 percent.
That changes where you spend testing effort. On Meta, the first frame carries much of the load. On TikTok, clearing 25 percent is achievable with basic discipline, and profitability usually lives in seconds three to eight.
The formula, and the threshold that invalidates comparisons
Hook rate = early views / impressions
The definitional problem is "early views", and it is why most published hook rate benchmarks are not comparable with each other. Meta's convention is 3-second video views. TikTok reports several thresholds, including 2-second and 6-second focused views alongside video views and quartile completions, and a focused view is stricter than a plain view because it demands the video actually be in focus on screen.
So the same creative produces very different hook rates depending on the numerator, in a predictable direction: a 2-second focused view rate reads lower than a 3-second view rate on the same ad. The 33 percent median is a reference point, not a measurement you can reproduce without stating your threshold.
Three rules that keep the number honest:
- Pick one threshold and never change it mid-quarter. Historical comparisons break the day you switch.
- Do not compare your TikTok hook rate to your Meta hook rate unless the thresholds match. They usually do not.
- Compare variations within one platform and one campaign. Hook rate is a relative instrument: excellent for ranking eight openings against each other, poor as an absolute grade.
The hook rate glossary entry has the definitional detail, and the hook rate calculator works it out from your impressions and view counts.
The four levers that move hook rate, ranked
Ranked by how much movement we see per unit of effort, highest first.
1. What is in the first frame
Not what happens in the first second: what is on screen at zero. The viewer's decision starts before any motion resolves, so a static first frame that reads as an advert has already lost people who never saw your cut.
The reliable version is a human face at close range, or a hands-and-problem shot. The reliable failures are a product on a clean background, a logo, a title card and a wide establishing shot. Swapping the first frame alone is the cheapest meaningful test in paid social.
2. The first spoken line
Sound is on, so speech is a first-second lever on TikTok in a way it is not on Meta feed. A claim delivered flat beats a question almost every time, because a question is a filter and filters remove buyers as well as non-buyers. "Struggling with dry skin?" invites a no. "This costs eleven pounds and replaced my four hundred pound one" invites a why.
Cut the wind-up. Any breath, greeting or "so basically" in front of the claim is dead airtime at the moment you cannot afford it.
3. Motion in the first half second
Movement holds. A reframe, a hand entering the frame, a cut on the first syllable, a camera clearly held by a person. This is why fast-cut edits work, and the reason is worth stating precisely: it is not that speed is virtuous, it is that a locked-off static shot with narration is a production signal, and production signals get read as advertising.
4. On-screen text in the first frame
A short line, four to seven words, in the middle third of the frame where TikTok's interface does not cover it. The weakest of the four and still worth running, because it insures against sound-off viewing and lets you state the claim twice without saying it twice.
| Lever | Effort | Typical movement | Test as |
|---|---|---|---|
| First frame subject | Low | Largest of the four | 5 to 8 openings, one body |
| First spoken line | Low | Large | Claim vs question vs result |
| Motion in first 0.5s | Medium | Moderate | Static vs handheld vs cut-on-speech |
| First-frame text | Low | Small, insures sound-off | With and without |
The shock hook trade, in numbers you should be watching
A hook with nothing to do with your product will lift hook rate and wreck everything after it. A loud noise, a near-miss, a face doing something strange, a claim bordering on a lie: all raise early views and pull in an audience with no reason to buy.
The pattern in reporting looks like this. Hook rate goes from 26 percent to 44 percent. Completion rate falls, because the video cannot pay off what the hook promised. Click-through may hold, because curiosity converts to clicks. Conversion rate falls hardest, because the clicks came from people who did not know what they were about to be sold. Net effect: CPA up, dashboard looking better.
So never approve a hook on hook rate alone. Read hook rate, completion rate and CPA as a set, and treat any variant whose hook rate rose while completion fell as a suspect rather than a winner. Our post on what creative can and cannot move in TikTok conversion rate has the funnel split behind this.
The honest exception: for prospecting on an unknown brand, a mild pattern interrupt that then delivers a relevant payoff is fine and often necessary. The test is whether the payoff relates to the interrupt. A dropped glass in front of a phone case, yes. In front of accounting software, no.
Testing hooks at the volume it takes
Hook rate has high variance, which is good for testing and bad for small samples. Five to eight openings against one body is the standard unit we run, and two or three percentage points on a few thousand impressions is noise rather than a result.
Genyad is our product and hook variation is the job it does most cheaply. You upload footage once, it transcribes and tags every clip, and each variation is a fresh script, shot selection, voiceover and caption set drawn from that library rather than a re-cut of one timeline, so eight openings can genuinely differ in first frame and first line rather than being the same clip with new text on it. The hook variation generator page has the workflow. One credit per variation, no subscription, and the free plan gives five with no card.
The limits: no AI avatars, no static formats, no predicted performance score that would rank your eight hooks before they run, and no direct publishing to TikTok. You find out which hook won by spending on it.
Frequently asked questions
What is a good hook rate for TikTok ads?
The median is about 33 percent, and 20 to 30 percent is the normal working range for in-feed campaigns. Under 20 percent means the opening needs rebuilding rather than tuning. Above 40 percent, check cost per acquisition first, because unusually high hook rates often come from pattern interrupts that dilute intent.
How is hook rate calculated on TikTok?
Divide qualifying early views by impressions. The complication is the threshold: TikTok reports 2-second and 6-second focused views while Meta's convention is 3-second views, and a focused view is stricter, so the same creative scores differently depending on the numerator. Pick one and hold it steady.
Can a hook rate be too high?
Yes. A hook that promises something the ad does not deliver raises early views and then loses completion and conversion rate, showing up as a rising cost per acquisition on an ad that looks better in the creative report. Read hook rate next to completion rate and CPA, never alone.
How many hooks should I test at once?
Five to eight against the same body, so the only variable is the opening. Smaller sets rarely separate cleanly given how much variance hook rate carries, and two or three percentage points on a few thousand impressions is noise.