Find the second people leave — before you publish
Retention graphs tell you that people left. They do not tell you why, and by then the video is already out. The hook test plays your first 60 seconds to reviewers in your niche and records the moment each one would have stopped watching, plus the reason.
Sign up gives you 5 credits. One review you give = one credit.
What the reviewer does
The clip plays. The moment they would have left, they say so — one tap. If they watch to the end, that counts too. Then they pick reasons from a fixed list: slow intro, unclear promise, bad audio, low energy, too long a setup, visual quality, did not match the thumbnail, or kept watching. One line of written feedback is optional and usually the most useful part.
What you get back
A drop-off timeline across the 60 seconds, the median leave second, the share who watched to the end, and the reason tags ranked. Reading them together usually points at a single fixable thing: fifteen seconds of logo, a promise the thumbnail did not make, or audio that is quieter than everything else in the feed.
Why 60 seconds
The first minute is where a video is won or lost, and it is short enough that a reviewer can give you real attention for it. Longer clips get skimmed, and a skimmed answer is worse than no answer. Clips are compressed on the device to 720p, and anything longer is cut at 60 seconds before upload.
Reviews that are too fast do not count
The server measures how long a review actually took. Anything under 20 seconds is thrown away, the reviewer loses reputation and no credit is paid. Reviewers who always leave at the same second or always pick the same position get flagged automatically.
Questions
Do reviewers watch the whole video?
No — only the first 60 seconds, which is what the hook test is about. Your full video stays unpublished and off the platform.
Is my unpublished clip safe?
It is shown only to the reviewers assigned to your test, inside the app, and it is never public. Deleting your account removes your clips.
How is this different from YouTube retention analytics?
Retention analytics arrive after publishing and show a curve without reasons. This runs before publishing and gives you the second, the reason tag and a sentence from a person in your niche.
What if everyone watches to the end?
That is a result too, and a good one: your hook is not what is holding the video back. Then the thumbnail and title test is where to look.