When a Short should come from long-form footage
The Shorts versus long-form comparison decides what each format is for on a business channel: Shorts earn ReachUnique people who saw an ad. among people who have never seen the channel and test ideas cheaply, and long-form earns watch time, trust and the clicks that carry a viewer to a website. The channel strategy hub notes that a single long-form video can produce several Shorts. This page assumes both decisions are made and deals with the cut itself.
Cutting from long-form makes sense when the footage already contains a complete idea that a stranger would stop for. Demonstrations, before-and-after results, a common mistake shown and corrected, and a clear answer to a question the audience asks are the usual candidates. It makes less sense when the long-form video is a conversation that needs its build-up, when the best moment depends on a screen too detailed to read on a phone, or when the footage was shot so tightly for a horizontal frame that the subject cannot be kept in a vertical one. In those cases a Short filmed for the feed is the better tool, and the cut is not worth forcing.
The approach is close to the clip rung of the format ladder in the guide to repurposing content, which covers repurposing across every channel. Here the destination is the YouTube Shorts feed, and the source and the derivative live on the same channel, so the link back to the long-form video is part of the job rather than an afterthought.
Before you start
- A long-form library with retention data. The candidate moments come from the audience retention report, which the retention curves guide explains, including how to read spikes and the separate Shorts view.
- The original project files or the highest-quality export. Reframing for a vertical screen crops the image, so starting from a compressed download loses detail the Short needs.
- An editor that can reframe, trim and caption. YouTube documents that Shorts can be made in the app or uploaded, and that a Short can be made from an existing video; the steps below work with either route.
- The channel's Shorts strategy in writing. Which pillars Shorts serve and where they send viewers, as the comparison sets out. The video brief for the source video, where one exists, says what the video promised and which moments were meant to carry it.
- A simple log. A spreadsheet with a row per Short: the source video, the timestamp of the moment, the kind of moment, the new opening, the publish date and the results read at a fixed age.
The steps
List candidate moments from the retention curve and the edit
Open the audience retention report for each long-form video in the batch and note every spike, where viewers rewatched a moment, and every plateau that held through a demonstration. The retention curves guide reads a spike as a moment to build the next video around, and it is just as often a moment to cut. Then watch the edit and add the moments the curve cannot show, such as a strong line in a video too new to have data. Record each with its timestamp and a one-line description of what happens in it.
Keep only moments that stand alone
For each candidate, ask whether a viewer who has never seen the channel and never heard the setup would understand it and get something from it. A result that refers to "the method from earlier" fails; a demonstration with its own before and after passes. Strike out anything that needs the previous minutes to make sense. This is the step that separates a Short from a fragment, and it usually removes more candidates than it keeps.
Write a new opening for the feed
In the long-form video the moment was reached after an introduction; in the Shorts feed nobody has seen that introduction, and the first second decides whether the viewer stays or swipes. Write a new opening as general practice suggests: start on the most interesting frame, or state the problem the moment solves in a few words of on-screen text or a short spoken line recorded for the purpose. Do not open on a greeting, a logo or a reference to the longer video. The hook should be honest about what follows, for the same reason the retention guide gives for long-form: a hook that wins the opening and loses the rest shows up in the curve.
Reframe the moment for vertical viewing
Shorts are vertical, as YouTube's Shorts documentation states. Reframe the footage so the subject stays in the frame throughout: follow the hands in a demonstration, the face in a spoken tip, the product in a result. Where a wide shot cannot be cropped without losing what matters, split it into two framings or drop the moment. Check every frame on a phone, because text, screens and small details that read well on a monitor often do not survive the crop.
Cut to a single payoff
A Short carries one idea and ends when it lands. Trim everything that does not lead to that payoff, including asides, repeated explanations and the sign-off from the original edit. YouTube documents a maximum length for Shorts, but the useful length is set by the moment: as long as the idea takes and no longer. The retention guide notes that Shorts loop and that Shorts which end close to where they began tend to hold better than ones with a sign-off, so consider whether the final frame can lead naturally back into the first.
Add captions and on-screen text
A Short whose message is spoken says nothing to someone who cannot hear it, or to a viewer somewhere they cannot play sound. Add captions and review them against the audio; the guide to accessible social media posts sets out what captions and text in video need, and the guide to AI for alt text, captions and transcripts covers reviewing an automatic caption file. Keep on-screen text short and clear of the parts of the screen the player covers, and make sure any information shown only visually is also said or written.
Attach the related long-form video and name it at the end
YouTube lets a Short carry a link to a related video, as the comparison and the channel strategy hub describe. Attach the long-form video the moment came from, and use the final seconds of the Short to say what the viewer gets by following it: the full method, the rest of the comparison, the answer to the next question. That closing line is the Short's CTAThe prompt that tells a reader the one action to take next, such as a button or link.. Choose it when the Short is planned, not after it is published, as the diagnosis for views without subscribers advises, so the edit builds towards it.
Publish on a rhythm beside new long-form
Release the cuts on a steady rhythm rather than in a single batch: an early cut close to the long-form video's publish date, so the related link points somewhere fresh, and the rest spread across the weeks that follow, between new long-form uploads. Moments from older Evergreen ContentContent that stays relevant and useful long after it is published, rather than being tied to a date or a news moment. can be cut at any time, which helps keep the Shorts rhythm going when the long-form pipeline is slow. Record each publish date in the log.
Read viewed against swiped away and the crossover
Read each Short at a fixed age with the Content tab filtered to Shorts, as the comparison recommends. The first measure is viewed against swiped away, which the retention guide describes as the Shorts equivalent of click-through. The second is the crossover: whether the long-form video attached to the Short gains views from viewers who arrived through Shorts, which the traffic sources guide and the views without subscribers diagnosis show how to read. Compare each Short with the channel's own earlier Shorts at the same age, not with figures from elsewhere, and look at Average View DurationTotal watch time of a video divided by its number of views. on the long-form videos that receive the crossover to see whether those viewers stay.
Feed what you learn into the long-form plan
Group the results by kind of moment in the log: demonstrations, results, tips, mistakes, answers. After several rounds, patterns appear, and they feed two decisions. The first is which kinds of moment to keep cutting. The second is the long-form plan itself: a Short that held viewers is evidence that its idea deserves a fuller video, and the next brief can mark the moments meant to become Shorts before filming. A kind of moment that keeps holding viewers can even become a format of its own, which the tutorial on planning a YouTube video series turns into a named run of long-form episodes. Treat each round as a turn of the experimentation loop: a stated expectation, a cut, a reading, a change.
Which moments make good Shorts
The log becomes more useful once each moment is tagged by kind. The table describes the kinds that business channels most often cut, as general practice; which of them work for a given audience is something only the channel's own results can show.
| Kind of moment | Why it can stand alone | What the new opening does | Where the closing line sends the viewer |
|---|---|---|---|
| Demonstration | The action is visible and complete; no explanation is needed to follow it | Shows the most striking frame first, or names the task | The full tutorial, for the steps and the reasons behind them |
| Before and after | The change is the payoff and is understood at a glance | Opens on the before, or on the after with a question | The long-form video that explains how the change was made |
| Common mistake corrected | The mistake is recognisable and the fix is short | Names the mistake in a few words | The video that covers the rest of the method |
| Direct answer to a question | The question is one the audience types or asks | Puts the question on screen | The long-form video that covers the follow-up questions |
| Single tip | Useful on its own if it is specific | States the tip's result, not its topic | The video the tip came from, or a playlist on the subject |
| Opinion or verdict | A clear position is easy to react to | States the position | The full comparison or review with the evidence |
Moments that rarely work are setup without payoff, summaries of what a video will cover, sponsor reads, and any segment that depends on a chart or a screen viewers cannot read on a phone. A paid flight that needs vertical ad edits from the same footage is a different job with its own creative rules, covered in the YouTube setup for brand awareness.
A climbing centre cuts its technique videos (illustrative)
Ixtravolm Climbing finds which moments cross over
Ixtravolm Climbing, an invented indoor climbing centre, has 40 long-form technique videos of 12 to 20 minutes each and has never published a Short. Its instructor and a part-time editor go through the retention reports for the 10 most-watched videos and list 46 candidate moments from spikes and plateaus. Applying the stand-alone test strikes out 27, mostly tips that refer back to "the drill from before". The remaining 19 are tagged as demonstrations (9) or tips (10), each gets a new opening, a vertical reframe that follows the climber's feet or hands, reviewed captions and the source video attached as the related video, and a closing line naming what the full video adds. The centre publishes 3 a week for 6 weeks beside its usual long-form upload. Read at 14 days, demonstrations average 71% viewed against 29% swiped away and tips 58% against 42%. The source videos of the demonstration Shorts gain an average of 240 views each from viewers who arrived through Shorts; those of the tip Shorts gain 35. The next 5 long-form briefs each mark two demonstration moments to be filmed with a vertical framing in mind, and tips are cut only when they show the move rather than describe it.
The figures belong to the example. What transfers is the order: the stand-alone test before any editing, a new opening and a single payoff for every cut, the related video chosen in advance, and results grouped by kind of moment and read at a fixed age against the channel's own earlier Shorts.
Common mistakes
- Cutting at random. A slice from the middle of a tutorial with no setup and no payoff is a fragment; run the stand-alone test before opening the editor.
- Keeping the original opening. The long-form introduction was written for a viewer who chose the video. The feed viewer did not, and needs a new first second.
- Losing the subject in the crop. A horizontal shot cropped without following the action leaves hands, faces and products half out of frame. Reframe shot by shot and check on a phone.
- Ending on a sign-off. "Thanks for watching" at the end of a Short wastes the loop point and the closing line. End on the payoff and the reason to watch the long-form video.
- No related video. A Short without a link to the source video leaves the viewer in the feed, which is the missing bridge the views without subscribers diagnosis describes.
- Publishing every cut on one day. A burst spends the Shorts at once and leaves nothing for the weeks between long-form uploads.
- Reading Shorts on long-form measures. Viewed against swiped away and crossover are the Shorts questions; a channel average across both formats describes neither.
- Never feeding results back. If what the Shorts teach never reaches the next long-form brief, the channel pays for the same lessons every round.
