Hold rate measures how well a video keeps the people its hook already won. It is the share of viewers who kept watching well into the video, out of everyone who made it past the opening — the stage after the hook, not instead of it.
A video can have a strong hook and a weak hold rate: the opening earns the click and the middle loses the audience anyway. Reading the two together shows whether a scenario needs a new opening or a new body.
A format that repeatedly shows a strong hook rate and a weak hold rate is telling an account something specific: the promise in the opening is not the video that follows it.
Because it is read against hook rate rather than alone, a single number never settles the question — it is the shape of the two together, across a handful of videos in the same format, that points at a fix worth trying.
A common mistake is reading hold rate on its own, as if a low number always meant a weak middle — a video with a very strong hook can post a lower hold rate than one with a weaker hook, simply because it earned a bigger, less committed audience to begin with.
It is most useful read across a run of videos in the same format rather than judged from one: a single low hold rate could be one weak edit, but the same pattern repeating across several videos in a format is the format itself telling an account where it loses people.
Where it comes from
Motion's own creative-analytics glossary defines it as through-plays divided by three-second plays — the audience the hook already earned, measured again further in.
(Through plays ÷ 3-Second Video Plays) × 100 = Hold rate
Read the source: Motion — Hold rate.
What Rankfield reads
The landing page answers this directly, in its own words:
Your account end to end, your competitors, and the wider market around you: posts, videos, captions, comments, rhythm, what you offer. It watches the videos, not just the numbers under them.