Turn a YouTube video into shorts: the method that survives month two

You have a long video sitting there. An interview, a podcast, a livestream, a twenty minute tutorial. You know there are ten short videos inside it. You also know that if you open an editor tonight, you will get one out, maybe two, and you will not touch it again for three weeks.

That is where most creators stop. Not for lack of ideas, for lack of cadence. This article covers how to cut a long video into shorts, what makes a clip hold attention, and above all how to make the whole operation fast enough that you can do it every day.

The real problem is not editing, it is repetition

A good short is not technically hard: a vertically framed excerpt, readable captions, a hook in the first two seconds. You can learn that in one evening.

The problem shows up at clip ten. Then at clip thirty. Manual editing costs somewhere between twenty minutes and an hour per clip depending on your skill and your software. At three posts a day, that is a full time job that pays nothing until you have an audience. Which is exactly why most accounts quit before finding out whether they would have worked.

So the useful question is not "how do I make a good short" but "how do I make thirty without spending thirty evenings".

What makes an excerpt work

Before talking about tools, you need to know what to look for in the source video. A good excerpt almost always has the same four traits.

It opens on tension, not on an introduction. A question, a claim that surprises, a number that clashes with intuition. If the first two seconds are used to say hello, the clip is dead.

It stands on its own. An excerpt that assumes you watched the previous five minutes does not work. The viewer arrives from nowhere and knows neither you nor the context.

It runs between twenty and sixty seconds. Shorter, and it has no time to say anything. Longer, and the content really has to carry it.

It ends on an ending, not on a cut. A conclusion, a punchline, a sentence that lands. A cut in the middle of a sentence reads as a mistake.

Those four criteria are what you scan for when you reread your transcript. They are also what an automatic clipping tool has to detect to be worth anything.

The three ways to do it

Method	Time per clip	What it costs	Who it suits
Full manual edit	20 to 60 minutes	Your time, nothing else	One heavily worked clip a week
Manual cut plus automatic captions	8 to 15 minutes	Your time, plus a subscription	A few clips a week
Automatic cutting and captioning	1 to 3 minutes	A subscription	A daily cadence

The last column is the one that matters. No method is better in the abstract: there is a method that matches the pace you want to hold. If you post once a week and editing is part of the fun, manual is perfect. If your goal is to post daily for six months to find out what lands, manual will break you first.

The method, step by step

1. Pick a dense source. A video where somebody talks a lot and says things. A podcast, an interview, a talk. A mostly visual video with little speech gives you very little to work with.

2. Start from the text, not the picture. Pull the transcript and read it. Good moments jump out of a page far faster than they jump out of a timeline. Mark anything that reads like a sharp claim, a story, or a disagreement.

3. Cut wide, then tighten. Take thirty seconds around the passage, then trim the start until you land on the first sentence that hooks. Almost every time, the real start of the clip is later than you thought.

4. Reframe to 9:16 on whoever is speaking. Not on the centre of the original frame, on the face. A centred crop cuts both heads in half in a two person interview.

5. Burn in captions. A large share of viewing happens with the sound off. Without captions you lose those people in the first second.

6. Post, then look at what worked. Not clip by clip, in batches of ten. One clip taking off tells you nothing. Ten clips of which three take off tells you which format to keep.

What to know about rights

A point most articles skip, and it matters: cutting up somebody else's video and posting it under your own name is not neutral. You are using someone's work.

Platforms have rules on reused content, and those rules move. Some creators explicitly allow their content to be clipped, sometimes encourage it, because it spreads their name. Others do not. The simple rule: either it is your video, or you have permission, or you are exposed.

On monetization, let us be precise

Platform payout programs exist and are gated behind public thresholds: a follower count, a view count over a given window, a minimum amount of original content. Those thresholds change regularly, and the amounts paid vary by country, season and topic.

What nobody can honestly promise you is a figure. If you read somewhere that a tool will earn you a specific amount per month, the person writing it does not know. What can be said is this: the volume of content you publish is the only lever you actually control. The rest depends on algorithms nobody steers.

What to watch in the first thirty days

When you start, the temptation is to watch the follower count. It is the least useful number of the lot, because it moves too slowly to teach you anything about your content.

Three numbers serve you better. Three second retention tells you whether your hook works: it is the only place where a bad opening shows up immediately. Average watch time divided by clip length tells you whether the middle holds, and therefore whether you are cutting too long. The view count of your tenth best video, not your best one, tells you where your real floor sits: one video blowing up can be an accident, ten decent videos are a method.

Write those three numbers down once a week, never daily. At a daily resolution you are reading noise, and you will end up changing format at the exact moment you should have doubled down. The useful window for drawing a conclusion is two weeks minimum, over a complete batch.

And one rule that saves a lot of wasted time: change one thing at a time. If you alter the hook, the length and the caption style within the same batch, you will never know which of the three produced the result.

The three line summary

Editing is not your problem, cadence is. A good excerpt opens on tension, stands alone and ends cleanly. And the only method that survives six months is the one that costs you minutes per clip, not hours.

If you want to see what that looks like on your own video, paste a YouTube link on the home page.