The most common mistake I see in short-form video is treating it as a small version of a big thing. Take the interesting bit, cut it down, publish it.
It rarely works, and the reason is that the length is not a constraint on the same form. It is a different form with its own structure, closer to a joke or a magic trick than to a shortened documentary.
The structural difference
A long piece can spend time establishing. It can introduce a situation, develop it, complicate it and resolve it, and the audience will wait because they have committed.
A fifteen second piece has no committed audience. Every second is a decision to continue. Which means there is no establishing phase available — the situation has to be legible immediately, from the first frame, without explanation.
The things that work well in the format share this property. A visible situation, understood instantly, that develops in a way you did not predict.
What does not work is anything requiring context. A clip from a longer piece almost always requires context, which is why extracted clips generally underperform things made for the format.
The loop is a structural element
Something that has no equivalent in longer work. Short videos replay automatically, and the transition from end to beginning is a moment the viewer experiences.
Pieces that account for this — where the ending sets up the beginning, or where the loop point is invisible — hold people for several repetitions. Pieces that end flat get one viewing.
This is not a trick. It is closer to how a piece of music with a repeat works, and thinking about the join is part of composing the thing.
The measurable version is that watch time can exceed the video's length, sometimes by a lot, and that appears to be a strong signal to whatever decides distribution.
Sound is more optional and more important
A contradiction that took me time to resolve.
Most viewing starts silent or is watched in circumstances where audio is inconvenient. So the piece has to work with no sound, which means captions or visual clarity.
And yet the pieces that do best are generally strongly audio-led, because someone who turns the sound on has engaged more deeply and stays longer.
The resolution is that it has to work silently and reward sound. Which is a genuine design problem and is why so much short form has burned-in text carrying the meaning while the audio carries the tone.
Where the format is genuinely good
I spent a while being dismissive and have changed my mind about several things.
Demonstration. Showing that something works, or how a thing is done physically, is well served by a tight loop with no commentary. Craft, cooking, repair, technique. The format strips out everything except the doing.
Single ideas. One observation, stated well, is a complete piece. Long form pressures you to pad a single idea into ten minutes and the padding is the worst part of most long videos.
Comedy timing. The compression suits it. A joke that takes ninety seconds to set up is usually a worse joke.
Where it is genuinely bad
Anything requiring evidence or qualification. There is no room to say that a claim holds under some conditions and not others, and the format's incentives push toward confident overstatement.
This is not a moral failing of the people making it, it is what the length does. A claim with caveats does not fit and does not perform.
Anything cumulative. Argument, narrative, character. These need duration and there is no compressed version.
And anything where the interest is in the process rather than the outcome, since the format shows outcomes.
The funnel idea and why it mostly does not work
The standard advice is to use short form to bring people to longer work. I tried this for about eight months and the conversion was poor enough that I stopped.
What I think is happening is that the audiences overlap less than assumed. Someone who watches a lot of short video is in a browsing mode that is not adjacent to sitting down for twenty minutes. They are not a shorter-attention version of the same person, they are in a different situation.
The people I know who have made it work did so by making short pieces that were genuinely complete, building an audience that valued them, and then offering longer work as a separate proposition rather than as the real thing the shorts were advertising.
Making it deliberately
What changed my results was writing for the format from the start.
Deciding the single idea. Deciding what is visible in frame one. Deciding what the surprise is and where it lands. Deciding what happens at the loop.
Four decisions, made before shooting, for a fifteen second piece. That sounds disproportionate and it is the difference between something that works and something that is a clip.