Process / Satisfying

What this format actually is

Something is done, properly, from start to finish, and you watch it happen. No narration, no face, no explanation — often no sound but the process itself or a piece of music. Rust comes off metal. A surface is resurfaced. Something disordered becomes ordered. The payoff is completion.

It is the highest-reach format available to somebody who will neither speak nor appear, and that combination is rare enough to be worth naming plainly. Most advice for people who do not want to be on camera points at faceless voiceover, which still requires a voice, a script and a point of view. This requires none of those. It requires that you make or do something that looks good happening.

Why it travels so far

There is nothing to translate. A satisfying process video works identically for a viewer in Jakarta and one in Lisbon. Every other format on this site is gated by language; this one is not, which multiplies the addressable audience by roughly the entire non-English-speaking internet.

It is decided in under a second. Short-form feeds are won or lost in the first frame, and a visual process gives you the strongest possible opening — the thing is already happening. No hook to write, no promise to make, no personality to establish.

It loops. If your first and last frames match, the video replays seamlessly and watch time multiplies for free. Almost no other format gets that.

The honest ceiling

Reach is not the same as value, and this format has the widest gap between the two of anything here.

The audience is enormous and shallow. People watch, feel the small satisfaction, and scroll. They do not learn your name, do not seek out your other work, and convert to subscribers at a fraction of the rate that talking-head formats manage. A video with two million views can produce a few hundred followers, which is a ratio that would be alarming in any other format and is normal in this one.

Monetisation is correspondingly weak. Short-form revenue is low per view, the audience is international so advertiser rates are mixed, and there is no personal brand to sell anything through. People do make a living from this, but usually by attaching the videos to something else — a shop, a service, a real business the process is part of.

You should also know that it builds no authority whatsoever. Nobody watching a resurfacing video forms a view about your expertise. If your goal is to be known for something, this format actively works against it.

What separates the good from the ignored

The process has to be genuinely clean. Most real work is messier than it looks in your head. If the middle is fiddly, ugly or takes six hours, the video either becomes a lie or becomes boring. Choose processes that are visually legible end to end.

One unbroken shot beats six cuts. The satisfaction comes partly from continuity — from believing you watched the whole thing. Cutting away undermines it, and cutting away at the difficult part undermines it badly.

Frame it top-down and hold still. Handheld footage of a process is a different, worse format. A fixed overhead frame with the subject filling it is close to a rule here.

Resist the urge to narrate. The instinct to explain is strong and it converts a process video into a bad tutorial. If it genuinely needs explaining, it is a tutorial and should be built as one.

Where to start

Look at what you already do and find the single most visually satisfying thirty seconds in it. Not the most impressive, not the most skilled — the most visually satisfying. That is usually a moment of transformation: colour changing, a surface being revealed, disorder resolving.

Film it top-down, in one take, with the frame full. Try to match the first and last frames so it loops. Add a trending audio track, because it costs nothing and does a meaningful share of the distribution work on TikTok and Shorts.

Then publish twenty of them before drawing conclusions. This format has more variance per video than almost anything else — the same process filmed twice can do a thousand views and a million — and a sample of three tells you nothing at all.