how to repurpose long-form video into short clips, from a podcast or YouTube show

In my experience, an hour-long podcast episode has somewhere between five and ten good short videos inside it, and almost none of them are where you'd expect. They're rarely the intro, the sponsor read or the polite opening question. They tend to come later, once the guest stops being careful. Learning how to repurpose long-form video into short clips is mostly learning to find those moments, then rebuilding each one so a stranger understands it in the first few seconds.

I edit long-form for the brands I work with and cut the clips from it, so this is my working method, start to finish. It covers how to find the moments, the four kinds of moment that survive as a clip, how to restructure and reframe them, captions and titles, the platform limits that catch people out, and how to send viewers back to the full episode. I'm the vendor here, so weight my opinion accordingly. Platform claims link to the platform's own help pages; everything else is my experience, labelled as such.

why repurpose long-form video into short clips at all

If you already record a podcast, a YouTube show, webinars or interviews, you've done the expensive part. The filming, the guest, the lighting and the hour of talking are paid for. Clips are how most people will find that work, because a stranger will give you thirty seconds long before they give you an hour.

The long-form audience is real, too. In February 2025 YouTube announced more than 1 billion monthly active viewers of podcast content on the platform, measured as of January 2025, and called itself the most frequently used service for listening to podcasts in the US. Clips are the bridge between that full episode and the people who haven't met you yet.

step 1: find the moments in the transcript, not the timeline

Scrubbing through an hour of video to find clips is slow and you'll miss things. I start with a transcript (most editors and recording tools will generate one), read it like an article, and highlight any passage where I'd stop and reread. Then I go to the timeline only for the passages I've marked.

Every candidate has to pass one test: would this make sense to someone who has never heard of you, landing on it cold in a feed? If the moment needs the previous ten minutes to make sense, or refers to "what we said earlier", it fails. Most of what I highlight fails this test, and that's fine. A first pass on an hour-long episode usually gives me fifteen or twenty candidates, and I end up cutting somewhere around five to ten. That's my experience, and it swings a lot with how dense the conversation is.

the four kinds of moment that work as a clip

When I look back at the clips that worked for the people I edit for, nearly all of them fit one of four shapes. Knowing the shapes makes the transcript much faster to read.

• The straight answer. Someone asks a question your customers ask, and the answer is complete in under a minute. "How long does a kitchen refit actually take?" followed by a real number and the reason.
• The story. A short anecdote with a beginning, a turn and an end. The job that went wrong, the first customer, the day the supplier let them down.
• The strong opinion. The moment the guest disagrees with something everyone in their industry says. It is the easiest clip to find, because the energy in the voice changes.
• The single step. One practical instruction pulled out of a longer how-to: the one setting, the one mistake, the one thing to check.

What rarely works on its own: introductions, thank-yous, inside jokes, and long build-ups where the payoff arrives four minutes later.

step 2: rebuild the clip so the payoff comes first

This is the step that gets skipped when clips are rushed out, and in my experience it's the one that matters most. In a conversation, people build up to the point. In a clip, the point has to arrive first, or the viewer is gone before it does.

So I restructure. If the guest spends forty seconds on context and then says the line that makes the moment, I open on that line, then cut back to the context, then let the answer play out. Sometimes that means opening on the last sentence of the passage. I also cut the hesitations, the "so, yeah", and any sentence that repeats the one before it. A clip that runs 2:10 in the episode often ends up at 0:50 with nothing important lost.

Instagram, at least, says it weighs this. Its own page, Instagram Ranking Explained, lists how likely you are to watch a reel all the way through among the most important predictions it makes. A clip with its payoff at the front gets watched to the end more often than one that makes people wait, in my experience. The opening line is a hook like any other, and I wrote separately about how to write a hook for a reel, which applies to clips line for line.

Where p.a. fits: I edit long-form podcasts and YouTube shows and cut the month's clips from the same recording, so one filming session feeds the whole slate. You can see the finished work on the homepage.
Ask me for a quote →

step 3: reframe from 16:9 to 9:16 around the speaker

Most long-form is filmed wide and horizontal. Reels, Shorts and TikTok are vertical. Cropping a 16:9 frame to 9:16 keeps roughly a third of the width (my arithmetic: about 32%), so what decides whether a clip works is who's in that third.

• One speaker: crop to them, with their eyes about a third of the way down the frame, and leave room at the bottom for captions.
• Two speakers: either cut between them as each one talks, or stack them, one above the other. Stacking works when the reaction matters as much as the line.
• Anything on screen: slides, a product on the desk or a screen share often sits at the edge of a wide frame and gets cropped out. Cut to it full-frame for a beat, or leave that moment for the long-form.

If you know you'll be clipping, it changes how you record next time: a separate camera on each speaker, people seated closer together, and nothing important parked at the edge of the frame.

step 4: captions, a title, and a reason to stay

Every clip I deliver has burned-in captions. Plenty of people scroll with the sound off, and captions are also an accessibility basic. The W3C's guidance on captions for prerecorded video says captions should include the dialogue, identify who is speaking, and include meaningful non-speech sound. Identifying who is speaking matters a lot on a two-person podcast clip, so I set the two speakers' captions in different colours or put a name on screen at the first change of speaker.

On top of the captions, a short on-screen title in the first seconds tells the viewer what the clip is about before the speaker gets there: "the quote that lost us our biggest job" or "why we stopped offering free site visits". Keep it to one line, in the brand's own fonts and colours, so a month of clips looks like one show. Avoid a full-screen logo intro, because it costs the first second and shows the viewer nothing.

step 5: know the platform limits before you export

Clip length is one place where the rules really have changed, so check them against the platforms' own pages rather than an old guide.

• YouTube Shorts: YouTube's help page on three-minute Shorts says square or vertical videos up to three minutes, uploaded on or after 15 October 2024, are treated as Shorts.
• Instagram Reels: Instagram's help page on recording a reel says you can make reels up to 20 minutes long, but reels over 3 minutes won't be recommended to new audiences. For a clip meant to reach strangers, three minutes is the real ceiling.
• Copyright on longer Shorts: the same YouTube page says any Short over one minute with an active copyright claim will be blocked globally. If your podcast has intro music or a song playing in the background that could draw a copyright claim, either keep that clip under a minute or cut the music out.

Most of the clips I cut land between 30 and 75 seconds. That's my habit from what holds attention for the brands I edit for, not a platform rule, and a strong story can earn longer.

step 6: send viewers back to the full episode

A clip's second job is to bring people to the long-form. YouTube has two built-in ways I use.

The first is in the YouTube app. Its help page on creating Shorts from your videos describes a Remix, then Edit into a Short option that lets you select up to 60 seconds of a public video you've uploaded, and says Shorts made this way are linked back to the original video. It is quick, but it's a trim, so it gives you none of the restructuring, reframing or caption styling above.

The second works for clips you've edited properly. YouTube's page on adding a related video to a Short explains that in YouTube Studio you can attach one of your videos to a Short, and viewers then see a clickable link below your channel handle. It needs advanced feature access on your channel. On Instagram and TikTok, name the episode in the caption and point to the link in your bio.

what about automatic clipping tools?

There are plenty of tools that take an episode and hand back a batch of vertical clips with captions. They're useful for a first pass on a long back catalogue. In my experience they're good at spotting where the energy rises and poor at the clip test from step 1: they'll happily pick a lively moment that only makes sense if you heard the ten minutes before it. Their crops also tend to drift between speakers. If you use one, treat its picks as candidates, check each against the stranger test yourself, and fix the opening line by hand.

one recording, a month of clips

To repurpose long-form video into short clips well, treat each clip as a new short video made from old footage: find the moment in the transcript, check it makes sense to a stranger, put the payoff first, reframe around the speaker, caption it, and link it home. Do that, and one good episode can feed a month of clips.