Short-form,
without the busywork.

The footage for a short is usually easy: a phone, a few minutes, done. The work is everything after: cropping to vertical without ruining the composition, cutting hard enough that nothing drags, captioning it so it works muted, and getting to the hook before the thumb moves.

That is a lot of fiddly execution for a fifteen-second video, and it is why so many good ideas never get posted. The effort is wildly out of proportion to the runtime.

The stakes of the format make this worse rather than better: short-form is the top format for return on investment for nearly half of marketers, so everyone is posting. The bar for a single post is low; the bar for posting consistently is where everyone falls off.

Cropping 16:9 to 9:16 throws away nearly half the picture, and something has to decide what goes.

Automatic reframing tracks faces, which is right for a talking head and wrong the moment the point of the shot is something else: the product in your hand, the screen you are demoing, the second person in the frame. A face-tracking crop will confidently remove the thing the video is about.

Two ways to fix it here. Say it ("keep the laptop screen in shot") or pause and circle the thing that must stay visible, and the crop composes around it. The second is faster and unambiguous, which is the whole argument for pointing at the frame.

Short-form captions are a visual element. Large, centre-weighted, high contrast, appearing in rhythm with the speech, often with the key word emphasised. They carry a real share of the video's personality, and a generic caption track looks immediately out of place next to native content.

Practical constraints worth knowing:

• Stay out of the bottom third. The platform's own UI (username, caption text, buttons) lives there and will sit on top of yours.
• Do not cover your face. Slightly above centre usually works.
• Short groups. One to three words at a time reads better at speed than full sentences.
• Check it muted, on a phone. That is how it will be watched. Judging it on a desktop monitor with sound is judging a different video.

Ask for the style directly ("big centred captions, one word at a time, highlight the keyword in pink"), and adjust from there. More on captions.

The efficient way to work in short-form is not one video at a time. It is recording once and cutting several posts from it, different angles, different lengths, different hooks for different platforms.

Conventionally that means duplicating a project several times and editing each. Here it is a sentence: "give me a 30-second version and a 15-second version," or "cut a version that opens on the punchline instead."

Testing hooks becomes cheap in the same way. Three cuts of the same content with three different openings costs three sentences, and the feed will tell you which one works.