How to edit a talking-head video without a separate editor

The edit most pieces to camera need

1. Trim the top and tail. The walk to the camera and the reach for the stop button are always in there.
2. Cut the obvious fluffs. One clean cut per mistake, no transition. Nobody notices a hard cut on a talking head.
3. Add captions, because most people will watch without sound.
4. Adjust the picture once, globally, rather than shot by shot. There is only one shot.
5. Export once, at the settings the destination wants.

That is usually the whole job. Resist adding more: music under a piece to camera almost always fights the voice, and transitions between two angles of the same person look like a mistake.

Cutting without it showing

A straight cut in a static shot produces a small jump, because your head will not be in exactly the same place either side. Three ways to hide it, in order of effort: cut on a moment you were still, cut where your hand passes the frame, or punch in slightly on one side of the cut so the two framings differ enough to read as deliberate.

If a mistake is in the middle of a sentence, cut the whole sentence rather than the error. Half a sentence joined to another half almost never sounds right.

Captions, filters and stabilisation

Captions generated from your script are accurate by construction, because the words were known before you spoke them. That is a meaningful advantage over transcribing the audio, particularly for names and technical terms.

Filters are worth a light touch: correct the white balance, lift the shadows a little, and stop. Stabilisation is for handheld footage only, and on a tripod it can introduce a slow drift that is worse than the problem.

More on captions specifically in burning captions into a video.