If your video ads stall out before the three-second mark, no amount of offer tweaking will save them. So what is a good hook rate for video ads? As a starting point to test, aim for a 3-second hook rate above roughly 25-30% on TikTok and Reels-style placements, and treat anything under 15% as a signal that the first few frames need work. These are ranges to benchmark against your own account history, not fixed industry rules — your product, placement and audience will shift the number.
What hook rate actually measures
Hook rate is the percentage of people who saw your video ad start and kept watching past an early checkpoint, most commonly 3 seconds. Some platforms and reporting tools also surface a 5-second or 15-second view-through rate, which tells you something slightly different: not whether the opening grabbed attention, but whether the next beat kept people there.
In practice, hook rate is a proxy for one question: did the first frame, the first line of copy, or the first sound give someone a reason to not scroll past. It is calculated as 3-second plays (or video plays at 3 seconds, depending on the platform's naming) divided by impressions, or sometimes divided by total video plays. Check which denominator your reporting is using before you compare numbers across campaigns, because the same ad can show a very different hook rate depending on that math.
What is a good hook rate for video ads, by platform and placement
There is no single universal number, because hook rate is shaped by where the ad runs, how long the full video is, and what the audience was doing right before they saw it. Someone mid-scroll through an entertainment feed behaves differently than someone watching a YouTube Short between videos they chose. Use the ranges below as a rough starting point to test against your own account, then build your real benchmark from your last 10-20 campaigns.
| Placement | Typical 3-second hook rate to test against | Notes |
|---|---|---|
| TikTok / Reels / Shorts feed | 25-35% | Fast scroll environment; the first second does most of the work. |
| Facebook/Instagram feed (non-Reels) | 20-30% | Slightly slower scroll pace than short-form video feeds. |
| Stories/Reels placements with CTA stickers | 20-30% | A visible sticker or caption can lift or hurt hook rate depending on placement. |
| In-stream / pre-roll (skippable) | 30-45% | Viewers already chose content; a skip button raises the bar in the first seconds. |
| Spark Ads / boosted organic posts | 30%+ | Often starts from an organic post with proven early retention. |
If your numbers land well below these ranges across multiple ads, the issue is rarely the platform or the audience — it is almost always the opening seconds of the creative. For spark-ad specific guidance on pacing a full video around that early hook, see video ad length for TikTok Spark Ads.
Why hook rate moves even when nothing about the offer changes
A few variables quietly change hook rate without touching your product, price or landing page:
- Video length — a 30-second ad and a 10-second ad are not directly comparable; shorter ads often show a higher hook rate simply because the 3-second mark is a bigger share of total runtime.
- Placement mix — the same ad running across feed, Stories and in-stream will blend into one reported number that hides very different real performance per placement.
- Audience familiarity — a retargeting audience that already knows your brand tends to hook differently than a cold prospecting audience seeing you for the first time.
- Sound-on vs sound-off environments — if your hook relies on a voiceover line or a sound effect and a chunk of your audience watches on mute, the on-screen opening frame and caption carry more weight than the audio.
- Thumbnail/first-frame vs auto-play — on some placements the static first frame is what people see before the video plays, so a weak first frame can suppress hook rate before the video even starts moving.
This is why comparing your hook rate to a number you saw in a forum post is less useful than comparing this week's ad to last month's ad, same placement, same length, same audience type.
How to track hook rate without guessing
Hook rate lives in your ad platform's own reporting, not in a creative tool. Pull the 3-second video play metric (or the platform's equivalent) alongside impressions or total plays, and build a simple column for hook rate per ad. A few habits make this more useful:
- Break reporting out by placement where the platform allows it, rather than reading one blended number.
- Compare ads of the same length and same hook format, not a 15-second ad against a 30-second ad.
- Log hook rate alongside hold rate (the 25%, 50%, 75%, 100% watch-through marks) so you can see where exactly attention drops — a strong hook that collapses at 25% has a different problem than a weak hook that never takes off.
- Track hook rate next to click-through rate, since a high hook rate with a low click-through usually points to a weak middle or offer, not a weak opening. For more on reading click-through alongside hook rate, see average click-through rate for video ads.
A note on tools: creative generators and editors, including FrameNotion, do not publish ads or report on live performance — that reporting only exists inside the ad platform itself. Treat creative tools as where you build and iterate the hook, and treat the ad platform as the only source of truth for whether that hook actually worked.
Diagnosing a low hook rate: a checklist
If your hook rate sits below the range you are testing against, work through this before touching the offer:
- Does the first frame make sense with the sound off? If the opening relies entirely on a spoken line, add a caption or on-screen text that carries the same idea visually.
- Is there motion or a visual change in the first second? A completely static opening frame, even a good-looking one, often underperforms an opening with a small camera move, a product reveal, or a quick cut.
- Does the hook promise something specific? Vague openers ("Check this out") hook worse than specific ones ("My skin cleared up after I stopped doing this one thing").
- Is the product or brand visible too early? An opening that looks like an ad gets scrolled past faster than one that opens like content — a question, a relatable scene, or a bold claim before the logo appears.
- Is the pacing too slow? If the first beat takes more than a second or two to land, tighten the edit before the 3-second mark.
- Have you tested more than one hook on the same body? A single ad with one opening line is a sample size of one — run hook variants to isolate what the opening itself is doing.
Hook frameworks and before/after rewrites
Here are weak-to-strong rewrites across common ad types, the kind of change that usually moves hook rate without touching the offer:
| Weak opener | Stronger opener | Why it tends to hook better |
|---|---|---|
| "Introducing our new vitamin gummy." | "I was exhausted by 2pm every day until I started eating these." | Leads with a relatable problem instead of a product name. |
| "Shop our phone stand collection now." | "Your phone stand is probably ruining your videos. Here's why." | Opens with a specific, slightly provocative claim. |
| "50% off this weekend only." | "Why is this jacket suddenly half price? Here's the catch." | Turns a flat discount line into a curiosity hook. |
| "Our skincare routine in 3 steps." | "Dermatologists keep telling people to stop doing this one thing to their skin." | Creates tension before revealing the brand. |
Notice the pattern: the stronger versions delay the brand pitch by a beat and lead with a problem, a claim, or a question instead. If you want a faster way to generate and test multiple openers like these against the same body footage, an AI hook generator for video ads can produce several hook variants to run side by side.
Where FrameNotion fits into hook testing
FrameNotion builds a 30-second vertical ad from a product link or page, writing the hook, problem, benefit, proof, offer and call to action from scratch rather than from a template, with an AI voiceover, captions and music timed to the cut. Because every ad is generated fresh, you can request an A/B hook variant on Pro and Agency plans and run two openers against the same body to see which one actually moves hook rate in your ad account's own reporting. A finished ad takes about 10-20 minutes, and the first one is free, so testing a second hook on the same product costs very little time. From there, hook rate and hold rate live entirely in your ad platform — FrameNotion builds the creative, your ad account tells you which hook won. If you want a sense of hook styles across product types before building your own, the example ads page is a useful starting point.
Putting it together
A good hook rate is less a fixed number and more a moving target you set from your own account's history, placement by placement. Use the ranges in this guide as a first benchmark, pair hook rate with hold rate and click-through so you know where attention and intent actually break down, and treat every low number as a note to rewrite the first three seconds, not the whole ad. Test more than one opening per body, log the results consistently, and let the ad platform's own numbers — not a borrowed industry average — decide what counts as good for your product.
Frequently asked questions
Is hook rate the same as click-through rate?+
No. Hook rate measures how many people kept watching past the first few seconds of the video, while click-through rate measures how many people clicked the ad. A video can hook well and still have a low click-through rate if the middle or the offer doesn't follow through on the opening's promise.
Should I compare hook rate across different video lengths?+
Not directly. A 3-second checkpoint is a much larger share of a 10-second ad than a 30-second ad, so hook rate tends to read higher on shorter videos. Compare ads of similar length when you're benchmarking.
What counts as a bad hook rate?+
There's no universal cutoff, but as a starting point to test, a 3-second hook rate consistently below roughly 15% across similar placements usually means the opening isn't giving viewers a reason to stay, and the first 1-3 seconds need a rework before you change anything else.
Can I see hook rate inside a video ad creation tool?+
Generally no. Creative tools, including FrameNotion, build and edit the ad itself but don't publish to ad platforms or report on live performance. Hook rate is a platform-side metric you check in your ad account's own reporting after the ad runs.
Does a strong hook rate guarantee good ad performance?+
No. A strong hook rate means people stayed past the opening, but performance still depends on hold rate through the rest of the video, click-through, and what happens after the click. Use hook rate as one diagnostic signal, not the whole scorecard.