How long should your video be

Everybody who asks this is hoping for a figure. Eight minutes. Twelve. Under a minute. Whatever figure you settle on will be wrong, because it was never yours to begin with: it came from somebody else's material, somebody else's audience and somebody else's video.

Your video is as long as the material that earns its place. Everything that earns its place stays, everything that does not comes out, and what is left over is the length. That is not a way of avoiding the question. It is a working method, and it reliably produces a shorter video than a target does.

Why does everybody want a number?

Because a number can be checked and a judgement cannot. A target feels like progress. You can hold it up against the timeline and know whether you have hit it.

The trouble is what a target does to an edit, and it does damage in both directions. Come in under it and you start padding: an extra example nobody asked for, a longer introduction, a recap of the thing you said ninety seconds ago. Come in over it and you cut whatever is easiest to remove rather than whatever is weakest, which is almost never the same clip.

Let the material decide

A recipe with four stages is a four stage video, and stretching it to hit a runtime does not make it a better recipe. A lesson with one idea in it is short, and there is no shame in that. A demonstration where the hands have to be watched properly is long, because the hands take the time they take.

Runtime follows the purpose, material and delivery constraints. A target can help plan a piece; it should not justify padding or removing an essential explanation.

Has this beat already landed?

This is the test we use, and it applies to more or less any video.

A beat is one point a viewer takes away. Go through your edit section by section and ask whether the viewer already has this one. If they do, it goes, however well it was filmed and however long it took to film. Repetition is the single most common reason a video overstays, and it hides in three usual shapes.

The instruction given twice

An explanation followed by a demonstration may repeat useful information. Keep the repetition if it helps the learner follow the action; shorten it when both versions do the same job.

The section that restates the one before it

Two sections, different words, same point. Each looks reasonable in isolation. Together they are the reason the middle of your video sags.

The run-up

Most people warm up before they say the thing they mean. The warm-up is useful for the speaker and dead on screen. Find the sentence where you started meaning it and begin there.

Where do the extra minutes come from?

Almost never from one section being too long by itself. They come from the same point turning up in three different places, spread far enough apart that each appearance looks fair on its own. That is why a length problem is invisible when you scrub through the timeline and obvious when you sit and watch the whole thing.

Which gives you something to do this week, without any software. Write out the points your video makes, one line each, in the order they appear. Not what you filmed, what the video teaches. Duplicates are visible on a page in a way they never are on a timeline. Delete the weaker instance of each pair, whichever it is.

What if the video really is long?

Then it is long, and that is allowed. Some subjects take forty minutes because they contain forty minutes of material. That video will keep the people who came for the material, and it should be chaptered so they can find their way back to the part they need.

Long videos can work when the material and audience warrant them. Check structure, navigation and pacing as well as repetition, rather than using runtime alone to diagnose the problem.

What can you cut before the weekend?

In order, and stop when the list stops producing anything:

1. The run-up at the top, down to the first sentence that means something. 2. Every instruction you gave twice, keeping whichever version was clearer. 3. Any section that restates a section you have already made. 4. Any stretch you skip past when you watch your own video back. You know the ones. 5. Nothing else. If it survived those four, it stays.

Where this stops being your rule and becomes a house rule

We do not keep material simply because it was expensive to film, or pad an edit to reach a number. Each passage needs a job; a recap or repeated demonstration can have one.

Knowing the rule is easy. Applying it to an afternoon of your own filming is the work. Of everything you shot, which point would you defend if somebody made you lose a third of the video? Answer that before you open the timeline and the length settles itself.

Sources and evidence

What is established. YouTube's own guidance on video performance states that publish time is not known to affect a video's long-term performance and that creators should aim for the video length that suits the content rather than a fixed target, which is the platform declining to endorse the idea of one ideal duration. That is a statement about YouTube specifically, not proof that repetition is the only reason a video overstays its welcome, and it does not extend to other platforms with their own norms.

Our method. Review each passage for its purpose, remove repetition that adds nothing and retain the explanation or practice the learner needs. Runtime is decided with the brief and delivery constraints in view.

References.

• YouTube Help. Good to know about recommendations (no ideal length). support.google.com. YouTube's own statement that there is no universal ideal video length; fit content, audience and purpose. Does not say shorter or longer performs better as a rule.