Click-through rate
Also known as: CTR, click rate
The share of people who saw a video listed somewhere and chose to open it, decided almost entirely by the thumbnail and the title.
Updated
What it means
It measures a decision made before any footage is seen. A video appears in a feed or a results page, and this is the proportion of those appearances that turned into a view.
Two things decide it and neither is the edit. The Thumbnail and the title are what a person judges, which is why a strong video with a weak cover performs badly.
It pairs with Audience retention as the other half of the picture. One says how many people arrived; the other says how many stayed once they did.
How it works in practice
A Frame grab from the edit rarely makes the best cover. A deliberate still, chosen or shot for the purpose, reads better at the size a feed actually displays it at, which is smaller than anybody designs for.
Legibility at thumbnail size is the whole test. Something that reads at a few centimetres across works, and anything needing a careful look does not, however good the photograph looks at full size on a monitor.
Comparing two covers answers more than theory. Platforms that allow testing settle the question with a real audience rather than with somebody general advice about what works.
What to watch out for
A high rate on a weak video is worse than nothing. People who click and leave immediately teach the platform that the video disappoints, which is a harder position than being ignored.
Misleading covers are punished twice. The clicks arrive and then the retention collapses, and a Recommendation algorithm reads the pair as a signal to stop showing it to anybody.
The figure depends on where a video appeared. The same cover in a subscriber feed and in a recommendation column produces different numbers, so comparisons across placements mean very little.
How RawClip handles it
A finished highlight is the video rather than the packaging, and the cover image is a separate decision that belongs to whoever publishes it. A frame from the result makes a reasonable starting point and rarely the best one available. A still shot for the purpose beats a frame every time.