What is thumb-stop ratio?
Thumb-stop ratio is the percentage of people who stop scrolling when your video ad appears, measured as short video plays (such as Meta's 3-second video plays) divided by impressions. It is mostly a verdict on the picture: the first frame, the movement and the contrast that make a thumb pause before any word lands.
Updated
How it is measured, and how it differs from hold rate
Most teams compute thumb-stop ratio exactly the way they compute hook rate: 3-second video plays divided by impressions. Meta counts a 3-second play when the video has played for at least 3 seconds, or 97% of its length if it is shorter, and excludes replays. Where teams separate the two names, thumb-stop is about the visual stop and hook rate about whether the opening line holds the viewer.
The partner metric is hold rate: how many of those stopped viewers stay. A common version divides Meta's 15-second ThruPlays (plays to completion, or of at least 15 seconds) by 3-second plays. High thumb-stop with low hold means the picture works and the message does not.
What moves it in an AI product ad
- Motion in frame one. A character already mid-action beats a still that fades in.
- The product, big and early. If the viewer cannot tell what is being sold, they keep scrolling.
- A look that differs from the feed. A clay or anime character stands out among filmed phone videos, while a realistic look blends in; choose the look that suits the product, not the trend.
- Readable captions, placed where the app will not cover them, for the many people scrolling with sound off.
An illustrative example
Illustrative example: a made-up hot sauce brand, Ember Ridge, tests two openings of the same 15-second clay ad. One starts on a calm table shot of the bottle; the other starts on the clay character mid-bite, eyes wide, bottle in hand. The second wins on thumb-stop, but hold rate drops after second four because the next shot is slow. The fix is to animate shot two as well, not to change the opening.
Common mistakes
- Opening on a logo or a title card. Nobody stops for a logo they do not know yet.
- Shock frames that have nothing to do with the product. They stop thumbs and lose the click.
- Ignoring sound-off viewing: a great sung first line does nothing for a muted viewer unless the caption carries it.
- Reading thumb-stop without hold rate, and so rewarding openings that bait and disappoint.
Related terms
See hook rate for the calculation and testing method, safe zone for where captions survive the interface, and 9:16 vertical video for the frame itself. To see how a clay character with your product in frame looks, visit clay animation ads or the example ads.
What it can't do
- A high thumb-stop ratio only shows the opening caught the eye; it says nothing about whether the ad sold anything.
What it costs
From $15 for one ad, or $29 a month for about 4 thirty-second video ads. A jingle take costs 4 credits; a typical 30-second video ad about 26. There is no free tier and no free sample.
One-time packs
One ad
$15 once
Five ads
$59 once
Monthly plans
Starter
$29 a month
Growth
$79 a month
Agency
$199 a month
Questions
- Is thumb-stop ratio a metric in Ads Manager?
Not under that name. You build it from Meta's 3-second video plays and impressions, or from TikTok's 2-second video views on TikTok.
- Why is my thumb-stop high but sales low?
Usually the opening promises something the rest of the ad or the landing page does not deliver, or the offer is unclear. Check hold rate and the click-through rate next.
- Do animated ads get better thumb-stop than filmed ones?
Sometimes, because they look different in a feed of phone videos, but it depends on the product and audience. Test one against the other with the same offer.
- Does the first frame matter if the video autoplays?
Yes. Autoplay means the first frame is what people see while deciding whether to keep scrolling, so it does more work than any other frame.
Sources
- Meta Business Help Center, 3-second video plays, read 17 Oct 2026
- Meta Business Help Center, 15-second ThruPlays, read 17 Oct 2026