Take the cover frame out of your own video

• A face with its eyes open, mid-expression. Not mid-blink and not mid-word: a mouth caught halfway through a syllable reads as a mistake even when nobody can say why.
• Something readable at thumbnail size. The grid square is smaller than a postage stamp on most screens. If you cannot tell what it is at that size, it is not doing any work.
• Not the first frame. The opening frame of a recording is very often black, a slate or somebody settling into their chair, and it is what the app picks for you by default.
• Centre-safe. The grid crops to a square from the middle. Anything important near the top or bottom of a tall frame is not in that square.

The grid is also the thing people forget is a page of its own. Nine covers that look like one body of work do more for a follow than any single clip does - views but no subscribers covers why that is where the decision actually happens.

Each platform asks separately and each one forgets what the others chose, which is why a clip posted to three places often has three different covers and one of them is a blur of motion. It is worth doing at upload time, per platform, from the same saved frame.

Posting the same clip everywhere covers the rest of that per-platform pass, and what a wide recording loses going vertical covers the framing the cover is taken from in the first place.

SnapDub's clips come out as plain video files, so a frame taken here works as the cover wherever you upload them.