From product launch to usable cutdowns

A launch video gets made. Three weeks later someone asks for a thirty-second version for LinkedIn, a ninety-second version for the docs page, and a cut that leads with the customer rather than the product. None of that was planned for, so it becomes a reshoot, a bad crop, or nothing.

The fix is entirely upstream. The derivative set is a decision you make before the capture, not a favour you ask for after it — because the things that make a cutdown possible (frame safety, isolated takes, a version-stamped UI) cannot be added later.

One scope note, because it matters for planning: this is about what a single technical capture can legitimately produce. A customer story still needs the customer, and an interview still needs an interview. Knowing which assets need their own shoot is half of what this document is for.

Name the assets at the script stage and every downstream decision gets easier. Each one carries a different job, and the job dictates the framing, the length and whether it can come out of the same capture at all.

The last row is the one people get wrong. A customer story cannot be extracted from a product capture, because the thing that makes it work is a human being saying something you did not script. Plan it as a second capture, or plan to do without it.

A technical video ages badly for exactly one reason: the interface in it stops matching the interface the viewer has open. Every choice below exists to make a stale screen cheap to replace instead of fatal to the asset.

Technical video review collapses in a predictable way. An engineer flags a wrong flag, product marketing wants the positioning changed, design wants the UI reshot, and nobody in the thread knows who decides. The cut goes round four times and arrives late without getting better.

The fix is not more process. It is naming, before the first review, who owns which layer and what each of them is allowed to block on.

Then run the reviews in an order that matches what is cheap to change. A technical correction arriving at fine cut is the single most expensive note in this format, because the motion graphics have already been built on the wrong frame.

A cutdown is not a shortened video. It is a different argument built from the same material, and the reason most of them fail is that they keep the setup and lose the claim.

The vertical versions carry one more constraint, and it is the one that has to be settled before the capture: you cannot crop a 16:9 interface into 9:16 and still have an accurate UI. Either the frame is composed with a centre-safe region from the start, or the vertical version rebuilds its frame around the interface rather than cropping to it.

Whoever makes the next re-cut is often not whoever made this one. The handover is what decides whether that costs an afternoon or a reshoot.