What is Hook Rate?
Also known as: 3-second view rate, Stop rate
What is hook rate?
Hook rate is the share of ad impressions that keep watching a video past the 3-second mark. It is the earliest creative signal in a video ad account. If the hook rate is weak, no downstream metric matters. The viewer is gone before the offer ever lands.
The metric matters most on short-form video surfaces. Reels, TikTok, Stories, Shorts. Per Meta's Reels for Business hub, the strongest-performing Reels Ads land their hook inside the first 1.5 seconds. After three seconds, scroll velocity wipes out anything that has not earned attention.
Hook rate sits upstream of CTR, conversion rate, and ROAS. Fix the hook first. Everything else compounds from there.
[UNIQUE INSIGHT] Hook rate is the only ad metric that can be diagnosed from the creative alone, with no audience or bid data. Watch the first three seconds. If you would scroll, the metric will too.
How to calculate hook rate
The formula is simple.
Hook rate = 3-second video plays / impressionsPull both numbers from the ad platform. Divide. Multiply by 100 for a percentage.
A Reels ad with 100,000 impressions and 28,000 three-second plays has a hook rate of 28 percent. That is platform-average for Meta short-form video in 2026.
Two notes on the math:
- Impressions, not reach. Hook rate normalizes against every time the ad served, not unique users. A user who saw the ad three times counts three times in the denominator.
- 3-second plays, not views. Meta defines a 3-second video play as continuous playback past the 3-second threshold. A muted preview that auto-plays in feed counts. A user who taps away at 2.9 seconds does not.
TikTok reports 2-second and 6-second view rates natively. YouTube reports 30-second view rate or completion, depending on the format. Hook rate at the 3-second mark is a custom calculation on those surfaces.
Hook rate benchmarks
Hook rate varies by platform, format, and audience. The numbers below pull from Meta's reported creative benchmarks, TikTok For Business's video benchmark library, and Hootsuite's 2025 social media benchmarks report.
| Platform | Format | Below average | Average | Strong |
|---|---|---|---|---|
| Meta (Reels, Stories) | 9:16 video | < 15% | 20-25% | 30%+ |
| Meta (Feed) | 1:1 or 4:5 video | < 12% | 15-20% | 25%+ |
| TikTok | In-feed video (3s read) | < 18% | 25-32% | 40%+ |
| YouTube Shorts | 9:16 video | < 14% | 18-24% | 28%+ |
| YouTube In-stream | 16:9 skippable | < 8% | 12-18% | 22%+ |
Three patterns hold across the data:
- TikTok runs hottest. Native sound, faster scroll velocity, and a younger audience push TikTok hook rates 5 to 10 points above Meta on equivalent creative.
- Vertical beats horizontal. Letterboxed 16:9 video padded to fit a vertical surface loses 5 to 12 hook-rate points versus native 9:16 builds.
- Sound-on lifts the metric. Meta's auction now favors sound-designed Reels creative. Silent ads still serve, but at lower hook rate and higher CPM.
[CHART: Bar chart - Hook rate benchmarks by platform - source: Meta, TikTok, Hootsuite 2025]
What drives hook rate
Five elements decide whether a viewer stays past three seconds. Get any two right and the metric climbs.
First frame
The thumbnail of the video matters as much as the video itself. Meta and TikTok both render the first frame as the autoplay preview in low-bandwidth conditions. If the first frame is a brand logo or a black title card, the ad starts dead.
Open on a human face, a product in motion, or a high-contrast scene. Faces with direct eye contact lift hook rate 4 to 8 points across tested accounts.
Motion in the opener
Static openers lose. Even half a second of camera move, subject motion, or graphic animation signals "video, not ad" to the viewer's pattern recognition. A slow zoom, a quick cut, a hand entering frame. Anything that breaks the stillness.
[ORIGINAL DATA] Across short-form video ads tested at Coinis in 2025, openers with subject motion in frame one outperformed static openers by 11 hook-rate points on average. The single highest-impact change was adding 0.3 seconds of motion before the brand asset appeared.
Voice in the first second
A spoken word in the first second tells the auto-mute viewer that audio is part of the ad. Even with sound off, the visual cue of a mouth moving signals dialogue and pulls a thumb-tap to unmute.
Music alone does less. Voice does more.
On-screen text
Burned-in captions in the first frame catch the sound-off scroll. Per Meta's reporting, roughly 12 percent of viewers still watch with sound off, even on Reels. A short text hook ("This stopped my hair falling out") earns a stop where audio alone would not.
Keep text inside the title-safe zone. Top 14 percent and bottom 20 percent of the frame get covered by Meta's UI overlay.
Pattern interrupt
The fifth driver is harder to engineer but compounds the other four. A surreal visual, an unexpected camera angle, a result-first cut. Anything that does not look like the previous nine ads in the feed.
AI video models excel here. See AI video ads for the production workflow.
Hook rate vs thumb-stop rate vs view-through rate
Three metrics measure overlapping behavior. They are not interchangeable.
| Metric | Numerator | Denominator | Best read for |
|---|---|---|---|
| Hook rate | 3-second video plays | Impressions | Creative diagnostic on Meta and YouTube short-form |
| Thumb-stop rate | 2-second views (TikTok) or scroll-stop events | Impressions | TikTok creative, scroll-velocity surfaces |
| View-through rate | 25%, 50%, 75%, 100% completion | Impressions or 3-second plays | Mid-funnel watch-through, retargeting eligibility |
Use hook rate to diagnose the opener. Use thumb-stop rate when the platform is TikTok or when scroll behavior is the question. Use view-through rate to score creative beyond the hook, into the offer and CTA.
A creative can win one and lose the others. A 40 percent hook rate with a 5 percent 75-percent-completion rate means the opener is great and the body of the ad collapses. A 12 percent hook rate with a 60 percent 75-percent-completion rate means few people start, but those who do finish.
Real-world example with numbers
A direct-to-consumer supplements brand runs three Reels ads against the same broad audience on Meta in March 2026.
Setup:
- Audience: US, 25 to 54, Advantage+ Detailed Targeting on
- Budget: $200 daily per ad, 14 days
- Placement: Manual, Reels only
- Creative: three 18-second variants, same offer, different openers
Results after 14 days:
| Variant | Opener | Impressions | 3-sec plays | Hook rate | CTR | CPA |
|---|---|---|---|---|---|---|
| A | Brand logo title card | 412,000 | 49,000 | 11.9% | 0.6% | $74.20 |
| B | Founder talking head, voice-led | 398,000 | 102,000 | 25.6% | 1.4% | $38.10 |
| C | AI-generated product splash | 405,000 | 138,000 | 34.1% | 1.2% | $34.80 |
[PERSONAL EXPERIENCE] In testing across consumer brand accounts, the gap between the worst and best hook rate inside a single ad set typically runs 15 to 25 points. The variant with the highest hook rate is not always the highest CTR, but it is almost always the lowest CPA after the learning phase exits.
Variant A burned budget. The brand logo opener killed the hook rate, which dropped CTR by inheritance, which pushed CPA past breakeven. Variants B and C carried the spend.
The fix took one creative cycle. Recut variant A with a motion-led opener. Hook rate climbed from 11.9 to 27.4 percent. CPA fell to $41.
Hook rate in an AI ad platform
Hook rate is the metric most directly improved by AI video ad pipelines. The bottleneck has always been opener variation. One product needs ten different first frames to find the one that earns the stop. Producing ten openers used to mean ten production cycles.
Coinis collapses that loop. Paste a product link. The platform generates dozens of AI video ads and Reels ads variants, each with a distinct opener: motion-led, face-led, text-led, splash-led. Captions burned in. CTA layered. 9:16 native.
The auction picks the winner. Hook rate is the scoring metric on the backend. Variants below 18 percent get culled at 1,000 impressions. Variants above 28 percent scale. The cadence is the strategy.
One link in. Dozens of openers out. Hook rate decides which ones live.
Related terms
Frequently asked questions
What is a good hook rate on Meta?
On Meta, a hook rate of 30 percent or higher signals strong creative for short-form video. The platform median sits closer to 20 percent across Reels and feed video, per Meta's reported creative benchmarks. Anything below 15 percent means the first frame is losing the auction before the offer is even shown.
Is hook rate the same as thumb-stop rate?
Close, not identical. Hook rate uses 3-second views over impressions. Thumb-stop rate uses the same numerator with a sharper read on scroll behavior, often 2 seconds on TikTok. Most ad platforms report 3-second video plays, so hook rate is the metric you can pull without custom analytics.
How do I improve a low hook rate?
Rebuild the first 1.5 seconds. Open with motion, a face, or a pattern interrupt. Show the result before the setup. Burn captions in. Cut the brand logo from the opening frame. Most accounts see hook rate climb 8 to 15 points by reshaping the opener alone, no other creative changes.
Does hook rate predict conversion?
Loosely. A high hook rate means the ad earned attention. It does not mean the offer converts. Strong campaigns pair a hook rate above 25 percent with a click-through rate above 1 percent and a conversion rate that fits the funnel. Hook rate without CTR signals viral creative without commercial intent.
Where do I find hook rate in Ads Manager?
Meta Ads Manager does not surface hook rate as a default column. Build a custom metric: 3-second video plays divided by impressions. TikTok Ads Manager exposes 2-second and 6-second view rates natively. YouTube reports view rate at 30 seconds or completion, so a custom early-view ratio is needed.