Long-Form to Short-Form: How to Repurpose One Video Into 20 Pieces of Content
Written byNereida Enriquez • Communications Associate
Video - September 2025
Most brands sitting on a library of long-form content are sitting on months of short-form they have never made. The podcast episode, the webinar, the founder interview, the product deep dive. Each of these contains ten to twenty usable pieces, and the vast majority of them are never cut.
The reason is almost never a lack of interest. It is that repurposing is treated as an afterthought instead of a designed process, so it happens sporadically, badly, or not at all. The brands that do it well have made one structural decision that the others have not, which is to treat the long-form asset as raw material rather than as the finished product.
The single biggest improvement you can make costs nothing. Plan the short-form while you are planning the long-form.
If you know you want ten clips out of an hour-long conversation, you structure the conversation to produce them. In practice that means three specific things.
Ask questions whose answers are self-contained. A question like "and what happened after that" produces an answer that is meaningless without the previous ten minutes. A question like "what is the most common mistake you see people make with this" produces an answer that works cold, for a stranger, with no context. The second kind of question is worth structuring an entire interview around.
Prompt the guest to restate the question inside their answer. This feels unnatural in the room and it is the difference between a clip that works and one that does not. If someone says "yes, exactly," you have nothing. If they say "the biggest mistake I see is people posting daily without a point of view," you have a clip. Coach for this explicitly before you roll.
Deliberately create moments that stand alone. Tell the guest in advance that you will occasionally ask them to say something in a self-contained way, and then do it. Nobody minds, and it saves the edit.
Shoot for the crop while you are at it. Frame with vertical in mind, keep the subject centered enough to survive a 9:16 cut, and avoid graphics or lower-thirds that live at the edges of the frame, because those edges are the first thing to disappear.
Once you have the footage, work through it deliberately rather than hunting for highlights by feel, because feel is unreliable and slow.
We run a pass looking for five specific things.
Strong claims. Any moment where someone says something definite that a reasonable person could disagree with. Disagreement is the engine of distribution, and a claim that nobody could contest is a claim nobody will share.
Numbers and specifics. These travel because they are concrete and because they are memorable. "We cut our acquisition cost by forty percent" outperforms "we improved our efficiency" by an enormous margin, and it is the same fact.
Stories, even thirty-second ones. Narrative holds attention better than assertion, reliably, and a brief anecdote will outperform a well-argued point almost every time.
Tension. Moments of disagreement, correction, or discomfort. These are the most shareable content in almost any conversation and they are the ones most likely to be edited out by a nervous marketing team.
Practical instructions. Anything a viewer could act on. These get saved, and saves are among the strongest signals available.
Anything that does not fall into one of those five categories is probably not a clip, regardless of how much you enjoyed it in the room. The affection you feel for a moment because you were there is not a signal about how it plays to someone who was not.
Here is the practical breakdown we use, and the reasoning behind each piece.
Six to eight vertical clips, thirty to sixty seconds. Each built around one of the moments identified above, each with its own hook and its own text overlay. These are the core assets and they do the discovery work.
Two or three longer cuts, two to three minutes. For platforms and audiences that tolerate more, which is where the more nuanced material goes. Something that requires a caveat or a qualification cannot survive a forty-second edit, and cutting it to fit destroys the point.
Three to five static graphics. Pulling a quote, a statistic, or a framework from the conversation. These cost almost nothing and they consistently outperform their effort, particularly on platforms where video is not the native format.
One written piece. A newsletter, a blog post, or a long-form caption built from the transcript. This one matters more than it seems, because search engines cannot index your video and they can index this. A well-structured written version of a conversation is a permanent asset that accrues search traffic for years, while the clips have a shelf life measured in days.
Audiograms or captioned excerpts. For platforms that favor them, and for the substantial audience that consumes with sound off.
At least one response piece. A follow-up addressing what people actually said in the comments of the first clips. This closes the loop, generates a second wave from an audience that is already warm, and signals to the algorithm that the account produces conversation rather than broadcast.
Publishing all twenty pieces in a week is the most common failure and the most avoidable one. The audience overlaps, the content cannibalizes itself, and you have burned a month of material in six days.
Spread it across four to six weeks. Lead with your two strongest clips, then alternate formats so the feed does not feel repetitive. Hold one strong clip in reserve for a moment when you need it, because you will have a week where nothing is working and having something proven in the bank is worth a great deal.
And leave a gap before you repost the best performer. A clip that worked well can be reposted six to eight weeks later to a substantially different audience, and most brands never do this because it feels like cheating. It is not cheating. The overlap between who saw it the first time and who is in the feed now is smaller than you think.
Clips that assume the viewer saw the source. Every clip has to work cold, for someone with no context, or it works for nobody. If your clip references "what we talked about earlier," it is broken.
Reusing the same hook across variants. If five clips from the same conversation all open the same way, the algorithm and the audience will treat them as one piece of content, and four of them will be wasted.
Treating this as a mechanical task. The judgment about what constitutes a moment worth extracting is the entire skill. It is not a junior job, and handing it to whoever has time is how a genuinely good conversation produces twenty forgettable clips.
Skipping the written piece. It is the least glamorous item on the list and it is the only one that is still working for you in three years.
Nereida is a Communications Associate at Proach Media, covering platform strategy, short-form video, and the creative tooling shifting how content gets made.