Correction, 2026-10-01: an earlier version gave thumbstop benchmarks (25 to 30% strong, 35 to 45% top decile, under 20% weak) and said full-screen placements run several points higher than in-feed. We could not source either. Its example read like a real brand's results and priced Seedance 2.0 openings at "a few cents to well under a dollar per clip": on 8frame a 5-second Seedance 2.0 clip is 48 credits at 480p and 108 at 720p (about $0.48 and $1.08 at pack rate), and Nano Banana Pro, linked there as a video model, makes images. The platform view thresholds are now sourced and the example is labelled as illustrative.
Thumbstop rate is the percentage of ad impressions that convert into a short qualifying view of a video, most commonly a 3-second view, measuring how often the opening frames stop a scrolling viewer long enough to register the ad.
It is the first gate in the video funnel. Before a viewer can be persuaded, held, or clicked, they have to stop the thumb. If your thumbstop rate is low, little downstream matters, because most of your audience scrolled past before the ad had a chance. This makes thumbstop rate the cleanest single read on whether your opening frame and first motion beat are working. When it is low, check the creative before you touch the targeting.
How thumbstop rate works
The formula divides qualifying short views by impressions:
thumbstop rate = (3-second video views / impressions) × 100
The catch is that platforms don't all define the qualifying view the same way. Meta reports 3-second video plays and, separately, 2-second continuous video plays. TikTok reports 2-second and 6-second video views. So a raw thumbstop number from one platform is not directly comparable to another, and you have to know which threshold your number uses before you benchmark it. Mixing a 3-second Meta figure with a 6-second TikTok figure produces a comparison that means nothing.
Placement can shift the baseline too: a full-screen Reels or Stories ad and an in-feed ad compete for attention differently. Compare thumbstop rate within a placement, not across placements.
When you use thumbstop rate
Thumbstop rate is one of three watch-time metrics that describe a single creative, and each answers a different question:
- Thumbstop rate (short views / impressions): did the opening stop the scroll?
- Hold rate (views to 25–50% / short views): did it keep the people it stopped?
- Hook-to-click rate (clicks / short views): did the attention convert to intent?
Read together, they localize a problem. A strong thumbstop rate with a weak hold rate means the opening overpromises and the body loses people. A weak thumbstop rate with a strong hold rate means the creative is good but the first frame is invisible, so you are starving good content of an audience. Diagnosing which of the three is broken tells you whether to fix the opening frame, the middle, or the CTA, instead of scrapping the whole ad.
Thumbstop rate and hook rate are often used for the same short-views-over-impressions calculation. Whichever term you use, state the view threshold so nobody benchmarks the wrong number.
Examples
Benchmark read. We have not found a first-party benchmark for thumbstop rate that we can cite, and published ranges move with vertical, placement, audience temperature and view threshold. The more useful comparison is against your own account's history: a clip that beats your trailing 30-day median for the same placement and audience type is a winner regardless of where it lands against a public number. Benchmark cold and retargeting audiences separately.
AI variant test to move the number (illustrative numbers, not a real account). Say a brand's cold Meta creative sits near 18% thumbstop rate. The team generates eight new openings on 8frame, varying only the first two seconds: a product-in-hand reveal, a close-up pour, a face-first talking-head open, and a bold text-over-motion open, among others. On Seedance 2.0, eight 5-second openings cost 8 x 108 = 864 credits at 720p (8 x 48 = 384 at 480p), about $8.64 or $3.84 at pack rate. For the text-over-motion open, the team adds the text in the editor so it stays legible and easy to change. Same body, same CTA, matched budgets. If the text-over-motion open lands at 31% while the rest sit between 17 and 22%, that is a thumbstop problem solved with creative volume, not a media problem. The winner becomes the new opening template, and the brand tests the next batch of hooks against it.
Hold-rate contrast. A different ad opens at a 34% thumbstop rate but its hold rate to 50% collapses. The opening is doing its job; the ad is losing people once they are watching. No amount of new first frames fixes that, because the thumbstop metric is already healthy. The fix lives in the middle of the video, which is why you never optimize on thumbstop rate alone.
FAQ
Is thumbstop rate a metric Meta or TikTok reports? Not by that name. You compute it from impressions and a short-view metric: 3-second video plays (or 2-second continuous plays) on Meta, 2-second or 6-second views on TikTok.
What does a batch of AI openings cost on 8frame? It depends on the model. Eight 5-second Seedance 2.0 openings are 384 credits at 480p or 864 at 720p; stills for a text-led open cost 11 credits each on Nano Banana 2 at 1K.
Related concepts
- What is creative volume explains why generating many opening variants is the practical way to move thumbstop rate.
- UGC hook formulas for AI ads breaks down opening structures that aim to stop the scroll.
Sources
- Meta Business Help Center, "3-second video plays": https://www.facebook.com/business/help/743427195703387, checked 2026-10-01
- Meta Business Help Center, "2-second continuous video play": https://www.facebook.com/business/help/453999754940285, checked 2026-10-01
- TikTok Ads Manager, video play metrics (2-second and 6-second video views): https://ads.tiktok.com/help/article/video-play?lang=en, checked 2026-10-01
- Seedance 2.0 and Nano Banana 2 prices (calculateSeedance2Cost); credit value at pack rate: 8frame tool registry and billing catalog, read 2026-10-01
Want to move your thumbstop rate without a studio? Open the canvas on 8frame and generate opening variants to test against real spend.
Try it without signing up: the 8frame canvas opens on a real board. The canvas is free and unlimited; generation is paid from $19/month.