Writing a Clip That Has a Shape

By Tess Varga · Clip seller · Published 21 August 2026

Two clips can contain exactly the same material and sell at completely different rates. Most of the difference is structural. One of them was planned as a piece with a beginning, a middle that changes something, and an end. The other was planned as a list of things to do on camera. That is a perfectly reasonable way to spend an afternoon and a poor way to hold anybody’s attention.

Shape is also the cheapest quality upgrade available to a solo producer, because it costs nothing but the twenty minutes before the camera goes on.

A scene and a demonstration are different products

A demonstration shows something. It runs from the start of the activity to the end of it, and its logic is completeness: this happened, then this, then this. It is easy to shoot, easy to describe, and it is what most first clips are.

A scene has a situation. Someone wants something, something is at stake, and by the end the position has changed. The material can be identical. What differs is that a scene gives the viewer a place to stand and a reason to keep watching after the first minute, because he is waiting to see where it goes rather than watching a sequence complete itself.

Demonstrations have a real market and there is nothing wrong with making them. But they are interchangeable with every other competent version of the same activity, and a scene is not, which is why sellers who work in scenes tend to build an audience that follows them rather than an audience that searches for a subject. It is also, quietly, what makes a piece survive in the back catalogue: a demonstration ages against everyone else’s, and a scene is only ever yours.

The first fifteen seconds

Whatever a buyer sees before he decides is your advert, and it is almost always the opening of the clip. That has one hard consequence: you cannot spend the beginning building up to something, because the build is the part he has not purchased yet.

In those first seconds he needs to work out three things, none of which are the plot. Who is talking. What the situation is. What kind of piece this is going to be, in tone and in pacing.

So the opening has to arrive already in motion. Start with the situation stated, not approached. If your first line explains itself, cut it and start with the second. Trim anything before the moment the piece properly begins, because a clip that starts in the right place is worth more than a clip with a warm-up attached.

Two habits kill more openings than anything else. The first is the greeting: a sequence of pleasantries before the piece begins, which reads as filler and dates the clip. The second is the apology, meaning any variation of settling in, adjusting things, or telling the viewer what is about to happen. Neither belongs in the part he is being asked to buy on.

The beginning

The opening’s job is orientation. Establish the situation, the relationship, and the register, and do it fast enough that the viewer stops asking what this is and starts wondering what happens.

You want it concrete. A specific situation, stated plainly, does more than an atmospheric one, because a specific situation gives the viewer something to hold. Vagueness at the top of a clip reads as hesitation rather than mystery, and it leaves the viewer waiting to work out whether this is for him.

Aim to be finished with the setup inside the first tenth of the runtime. Longer than that and you have written an introduction rather than an opening.

The turn

The turn is the part most clips are missing. Somewhere in the middle, something changes: the situation shifts, a decision gets made, a rule is introduced, the balance moves. It does not need to be dramatic. It needs to make the second half a different thing from the first half.

Without one, a piece can only escalate, which gives you a slope rather than a shape. A viewer reads the whole slope inside the first thirty seconds, and that is the moment his attention starts to drift. With a turn, the second half becomes unpredictable in a small way, and small is enough.

The practical test is easy. Describe the situation at the start of your clip in one sentence, then describe it at the end in one sentence. If the two sentences are the same except in degree, there is no turn, and the piece will feel long even if it is short.

The ending

An ending is a decision, and most clips do not make one. They stop instead. A piece that stops leaves the buyer with nothing to carry away and no reason to remember whose clip it was.

The ending wants to resolve the situation the opening set up, in the same terms. If the piece started with something at stake, say what happened to it. Then finish on a deliberate last line, and know it before you shoot, because a solo producer with no planned ending will trail off and hope the edit rescues it. The edit will not rescue it. It will just cut the trailing off and leave you with a clip that stops abruptly.

A clean ending also has a commercial effect that is hard to see in the numbers. It is the last thing he experiences, and it is what he is judging when he decides whether to come back for another, which is most of how a repeat buyer is made.

Length follows shape, not the other way round

Deciding to shoot a fifteen minute clip and then finding fifteen minutes of material is how pieces get padded. Decide the shape, then let it tell you the length: a single situation with one turn is usually shorter than sellers expect, and stretching it is visible.

If a piece is genuinely running long, it is normally because it has two turns in it, and two turns is two clips. Splitting is nearly always better than padding, and it gives you two shapes instead of one long one.

Write it on one card

The whole plan fits on an index card, and the card is the reason the shoot goes quickly.

• One line for the situation.
• Three or four beats, in order.
• The turn, marked.
• The last line, written out in full.

That is enough structure to shoot against and loose enough to perform inside. Tape it just under the lens and your eyeline stays where it should be, which matters more than usual when there is nobody else in the room, as filming clips alone goes into.

The card also shortens the edit more than any editing technique will, because a piece shot to a known shape needs assembling rather than rescuing.

Shape survives the description too

One useful side effect: a clip with a shape is easy to write a title and a description for, since there is something to describe beyond the activity. A demonstration can only be labelled with its contents. A scene has a situation, and a situation is what a buyer recognises as the thing he was looking for.

If you cannot write a one-line description of your clip that includes something happening, the clip probably does not have a turn yet, and that is worth finding out on the card rather than in the edit.

In practice

A clip with a situation, a turn and a real ending holds attention differently from one that simply runs. The opening carries most of the commercial weight, because whatever a buyer sees before he pays is usually the first few seconds, so the piece has to start already in motion rather than building up to itself. The middle needs something to change, or the second half is readable from the first thirty seconds. The ending needs to be a decision, not a place where the recording stopped.

For beginners

Plan it on an index card before you switch anything on: one line for the situation, three or four beats, the turn marked, and the last line written out in full. Cut any greeting or settling-in from the top. If you can't describe your clip in one sentence with something happening in it, it doesn't have a turn yet, and that is much cheaper to fix on the card than in the edit.

For experienced sellers

Sellers working in scenes rather than demonstrations tend to build an audience that follows them instead of one that searches for a subject, and their work survives longer in the catalogue because it isn't interchangeable. They let the shape set the runtime rather than the other way round, split a piece with two turns into two clips instead of padding one, and write the closing line before the shoot so the edit assembles a finished piece rather than trimming a trail-off.