Clipping for YouTubers: turning long-form uploads into Shorts

YouTubers are in an unusual position compared to every other kind of creator: you already have data about which parts of your video worked. The audience retention graph is a moment-by-moment record of where people leaned in and where they left. Almost nobody uses it for clipping, and it's the single best clip-finding tool you own.

Use the retention graph as a shortlist

Open any upload with a few thousand views, go to Analytics → Engagement → Audience retention, and look for the spikes: the small upward bumps where the line goes up instead of down. YouTube even labels the top one. Those bumps are re-watches: people scrubbed back to hear something again.

A re-watch spike is about the strongest signal a moment is clippable that exists anywhere, because it's real behaviour from your actual audience rather than a model's guess. Note the timestamps of the top three or four.

Then run the video through automatic clipping and compare. Where the tool's picks and your retention spikes agree, post that clip first. You have two independent signals pointing at the same 40 seconds. Where they disagree, your retention graph wins for older videos and the tool wins for videos too new to have data.

For a brand new upload with no retention data yet, the automatic pass is all you have, which is exactly when it's most useful.

Chapters are a shortcut

If you already write chapters, you've done half the segmentation work. Chapters mark topic boundaries, and topic boundaries are where clips naturally start and end. A clip that begins at a chapter marker almost never opens mid-thought.

This isn't automatic: the tool works from the transcript, not your chapter list, but when you're triaging the returned clips, the ones that align with a chapter boundary tend to be the cleanest.

Shorts are discovery, not a second channel

The mistake is treating Shorts as separate content that has to stand alone forever. Shorts are the top of a funnel that ends at your long-form uploads, which is where watch time and revenue actually live.

That changes what makes a good clip. You are not trying to deliver a complete, self-sufficient video in 45 seconds. You're trying to make someone think who is this and what else have they made. A clip that resolves a bit too neatly gives the viewer no reason to click through. A clip that lands a sharp point and stops leaves them curious.

Practically:

• Post the clip, then pin a comment pointing at the full video. Do this every time.
• Don't clip your best moment from a video that's still new. Let the long-form get its initial push first, then clip it a week or two later.
• Space them out. One upload should feed three to five Shorts across a fortnight, not five in one afternoon. The algorithm treats them as separate bets, and spacing gives each one its own shot.

Vertical reframing when you shot 16:9

Almost all long-form YouTube is horizontal, and Shorts are vertical. Cropping the centre of a 16:9 frame works fine for a talking head sitting still. It works badly for anything else: two people in a wide shot, a screen share, a whiteboard, or a presenter who moves around.

Clips are rendered vertical with the speaker kept in frame, and captions are burned in. But be realistic about what survives the crop: if your video's value is in what's on screen rather than what's being said, a vertical clip of it will lose the point. Screen-recording tutorials in particular tend not to reframe well, and you're usually better off clipping the parts where you're talking to camera about why something matters rather than the parts where you're demonstrating it.

Where the boundaries matter most

Shorts loop. That's the format's defining behaviour, and it makes a bad end point far more punishing than it is elsewhere: a clip that stops mid-sentence loops straight back into itself and produces an obviously broken artefact that viewers notice immediately.

Getting clips to end on a real sentence boundary is a solvable engineering problem rather than a matter of taste, and we wrote up what actually causes mid-sentence cuts, including the measurement that showed 7 out of 10 clips were affected before we fixed it.

Cost, if you upload weekly

A weekly YouTuber posting 20–40 minute videos is 80–160 minutes of source material a month, and that's before you re-run anything. Metered tools price by input minutes, so your bill scales with how long your videos are, which is a strange thing to be penalised for.

GetClipMachine is €15/month flat, whatever the length or count. If you want the specific numbers against a metered competitor, see GetClipMachine vs Opus Clip.

Free accounts get 3 videos a month. That's roughly three weeks of a weekly upload schedule, enough to see whether the clips are worth posting before paying for anything.

What it won't do

• It won't reframe a screen share sensibly. Cropping a 1080p screen recording to vertical makes the text unreadable.
• It won't write your titles. You get the clip and a reason it was picked. The Short's title and hook text are still yours.
• It won't find a moment that isn't there. A 40-minute video with no strong beat produces low scores, and the scores should be low.