Hook rate, also called thumbstop rate, is the share of impressions that produced a 3-second video view.
Hook rate = (3-second video views ÷ impressions) × 100
Reported 2026 benchmark ranges across creative-analytics sources put roughly 25 percent as a working baseline, 30 to 40 percent as good, and 40 percent or above as elite. Many performance teams use 35 percent as the point at which an ad earns more budget. Below 25 percent is generally treated as needing a rebuild rather than a tweak.
Those are third-party reported ranges, not Meta-published rules, and they vary by source, placement and vertical. The number that matters more than any of them: hook rate is a triage metric, not a success metric. It tells you whether the opening earned attention. It tells you nothing about whether the ad earns money.
Read the definitions before you read the benchmarks
Benchmark ranges for this metric disagree with each other more than almost any other Meta metric, and the reason is usually definitional rather than empirical.
| Term | Common definition | Denominator |
|---|---|---|
| Hook rate | 3-second views ÷ impressions | Impressions |
| Thumbstop rate | Same as hook rate in most sources | Impressions |
| Hold rate | Longer-duration views ÷ 3-second views, or ÷ impressions | Varies by source |
| Engaged view | Video view at 5 seconds as of March 2026 | Impressions |
Two sources quoting hold rate can be measuring different things, because one divides by 3-second views and the other divides by impressions. Those produce numbers that differ by a factor of three or more. This is why you will see a good hold rate quoted as 20 to 25 percent in one place and 40 to 50 percent in another, with neither source being wrong. Before comparing your number to any published benchmark, confirm the denominator.
Note also that the engaged-view threshold moved from 10 seconds to 5 seconds in March 2026, alongside the introduction of engage-through attribution. That changed the engaged-view column without any change in creative performance. Any time series crossing that date needs annotation before it is read as a trend.
Where hook rate is genuinely useful
Fast, cheap triage. Hook rate reaches a readable sample far sooner than conversion rate does, because its denominator is impressions rather than clicks. At a $15 CPM, tens of thousands of impressions are inexpensive. Tens of thousands of clicks are not. This is why attention metrics are the practical early signal on small budgets, and why proving a conversion-rate difference between two creatives is usually out of reach at the budgets most accounts test on. The arithmetic is set out in the creative testing budget calculator.
Isolating the opening. If two creatives share an offer, an audience and a landing page, and differ only in the first three seconds, hook rate measures that difference cleanly. That is a narrow claim, and it is the claim the metric actually supports.
Localising where an ad loses people. Hook rate and hold rate read together tell you which part of the ad failed, which neither does on its own.
| Hook rate | Hold rate | Likely diagnosis |
|---|---|---|
| Low | Low | The opening is not stopping anyone. Rebuild the first frame. |
| High | Low | The opening over-promises relative to what follows. Bait, not hook. |
| Low | High | The content works for those who reach it. The first frame is the constraint. |
| High | High | Attention is not your constraint. Look at the offer, the landing page or the audience. |
The high-hook, low-hold pattern is the most common self-inflicted problem in a creative library. A pattern interrupt with no relationship to the product buys a 3-second view and loses the person at second four, and the account then optimises toward a metric that is measuring the interrupt rather than the product.
Where hook rate misleads
It is not a revenue metric. A 45 percent hook rate on an ad that does not convert is an expensive way to be noticed. No band on the chart above is a claim about profitability.
It is placement-dependent. Full-screen surfaces produce structurally different hook rates from feed surfaces. Comparing a Reels-heavy ad set to a feed-heavy one on hook rate compares placements, not creatives. This became harder to control in 2026, because placement selection is being removed from ad sets — see what replaced placement controls. If you cannot fix the placement mix, you cannot fully isolate the creative, and the honest response is to say so in the readout rather than to report the difference as a creative result.
It is not comparable across verticals. Categories with strong visual novelty produce higher hook rates than considered-purchase categories, permanently, with no implication about profitability. Comparing your hook rate to a published average from a different vertical measures the vertical.
It rewards the wrong optimisation when used as a target. Any metric becomes gameable once it becomes a goal, and hook rate is unusually easy to game: the first three seconds can be made startling without being relevant.
The defensible use is as a gate rather than as a goal. An ad below the fix-it band probably does not deserve further spend. An ad above it has earned the right to be judged on downstream metrics — which is where the actual decision gets made.
What to do with a low hook rate
In rough order of typical effect size:
- 1.Change what is on screen at frame one, not what is said at second two. Motion, a face, or the product in use beats a logo or a title card.
- 2.Remove the ramp. Openings that establish context before showing anything lose the impression during the establishment.
- 3.Show the subject at the size it will be viewed. A product shot framed for a desktop preview is a small object on a phone.
- 4.Make the first frame legible without sound. Sound-off is the default state, so a hook carried entirely by a voiceover is not a hook.
- 5.Test five genuinely different openings, not five colour variants of one. Distinct openings produce interpretable differences; near-duplicates produce noise.
The structures are the fastest way to generate genuinely different openings rather than variations on one: 12 Meta video ad script templates gives twelve shot-by-shot frameworks with one named open variable each, and first three seconds hook formulas goes deeper on the opening block itself. For the frame those openings have to survive in, see the 9:16 production specs.
What a hook rate cannot tell you
It cannot tell you whether an ad is fatiguing. A declining hook rate on a creative that has been running for weeks is consistent with fatigue, and also consistent with a placement mix drift, an audience saturation effect, or a seasonal change in who is scrolling. Distinguishing those requires more than one metric — see creative fatigue versus creative failure.
It cannot tell you whether the ad should be scaled. Scaling decisions need cost per result and the margin underneath it, and a 40 percent hook rate on a product with no headroom is still a loss.
And it cannot substitute for a controlled comparison. Hook rate makes a creative difference visible quickly and cheaply. Whether that difference survives into revenue is a separate question with a separate cost, and treating the first as evidence for the second is the most common way an account talks itself into scaling an ad that does not work.
AdRiseLab generates image and video ad creative for Facebook and Instagram from a connected Meta ad account, and reports performance back against the ads that ran. Budget, bidding and the decision to scale stay with the advertiser.
Related Reading
The hook rate hub covers the metric definition and where it sits in Meta reporting. For the wider set of creative benchmarks, hooks, formats and creative benchmarks for 2026 puts hook rate alongside the other reported ranges, and the industry benchmark tables give the CPM, CTR and CPA context those numbers sit inside. For the production side, AI video ads covers how the openings get made.