When Does a Commercial Treatment Need Storyboards Instead of More Prose?
When Does a Commercial Treatment Need Storyboards Instead of More Prose?
Usually it doesn't. Not because boards are overrated, but because "storyboards" is often the wrong name for the thing you need. What most treatments are missing is one small picture that answers one specific question about one beat. That picture might be a plan, two frames, or six seconds of rough timing. Occasionally it's a real storyboard. Budget and goodwill go wrong when a team commissions the biggest version to answer a question the smallest one would have settled.
"Do we need boards?" is hard to answer because the word covers at least four different objects, and they cost wildly different amounts:
- a plan — a positional or overhead diagram showing where things and people are in relation to each other;
- selected frames — a handful of drawings showing states of a changing action;
- a timed study — a rough sequence with durations attached, sometimes called an animatic once it's been cut to time;
- boards — the full drawn sequence of the commercial's shots.
These are not rungs on a ladder of professionalism. Each answers a different kind of question, and one of them will answer yours while the other three sit there looking expensive.
Locate the incomprehensible beat
Start with the prose you already have. Hand the passage to someone who hasn't read the rest of the treatment and ask them to say it back. Not "what do you think?" — ask them four things:
- Where is everything? Who is where, what's between them, what can each person see, where does the camera stand?
- What changes? Something becomes visible, an object moves, one state becomes another.
- What happens at the same time? Two things overlap and the meaning depends on which overlaps which.
- How long does it take? A pause, a delay, a beat that needs room to land.
Whichever of those your reader gets wrong is your gap. "It feels thin" is not a diagnosis. "I couldn't tell whether they switched the bags or just stood there" is.
The reason to sort the gap this way is that prose is genuinely bad at the first question and genuinely good at the fourth, while pictures are the reverse. Prose carries position one relation at a time and asks the reader to assemble the rest, which means two readers assemble two different rooms. A drawing fixes the room and says nothing about how long anyone stands in it. A still frame has no duration at all. It cannot contain simultaneity, because simultaneity is a relationship between two things happening and a still is one moment.
Here's a beat to work with. It's invented to make the choice concrete, and nothing about it has been drawn, timed, tested or shot.
On the facing platform, a courier in a hi-vis vest stands with a black duffel. A woman in an olive jacket stands opposite him with a matching duffel. An express runs through on the near track. When it clears, both are still standing there, each holding a black duffel, and nothing about either of them has changed.
Read that cold and you have to guess four things: where the camera is and what it can see; whether the two exchanged bags or simply stood there; when the exchange happens relative to the train; and how long the train hides them. Four guesses, four different kinds of question. Now watch what happens to them one at a time.
Use a plan for a relationship
A plan is the cheapest useful drawing in a treatment, and it is the one most often skipped because it doesn't look like anything. It's usually a top-down or a side-on view: the camera, the tracks, the two people, the direction of travel, the sightlines. Ten minutes with a marker.
What a plan settles is who can see what and where two paths cross. In the station beat, it would show the camera set back roughly five metres from the near platform edge on one side, the two subjects about nine metres away on the facing platform, standing in profile to each other with about a metre of platform between them, and an express crossing the near track from left to right at something like thirty-five metres per second. The two sightlines from the camera to the two subjects cross the train's path about 0.9 metres apart along the direction of travel, which at that line speed works out to roughly a fortieth of a second — shorter than a single frame at 24 frames per second.
That is the plan earning its keep, and note what it does: it kills an idea rather than presenting one. Somebody in the room will suggest that the train could hide one of them a beat before the other, so the swap is half visible. The plan shows that difference is a fraction of a frame. The idea is dead, cheaply, before anyone draws it eighty times.
What the plan cannot establish is just as important. It won't tell you what the viewer actually notices, how long anything takes, or whether the beat is funny. It also won't tell you whether the camera can physically hold both the subjects and the platform edge in one frame at that distance — that's a lens question, and a plan drawn without one is a diagram, not a decision.
If nobody in the room disagrees about the geometry, skip the plan. A sentence can carry it: "Facing platform, static camera, express on the near track." A plan is worth drawing when the geometry is disputed, not because the geometry is complicated.
Use selected frames for a change
Frames earn their place when something visible changes state and the intermediate states matter. Two to four is usually enough. Choose the states that carry the missing information — the hand before and after, the door open and closed, the figure at the near mark and the far one — not one drawing per sentence.
Do this and you buy a specific thing: confirmation of shape. Two frames will tell you whether the courier's silhouette after the swap is distinguishable from her silhouette before it, at the size the camera will actually see her. If the answer is "yes, obviously, the bag now hangs a hand lower," you've settled a design decision in an afternoon.
Now apply it to the station beat, where it goes wrong in an instructive way. The before state and the after state in this commercial are deliberately the same picture: two people standing, each holding a black duffel in the same hand at the same height. Draw them and you get two nearly identical sheets. Draw the moment in between — the handoff at hip height, between two silhouettes, nine metres from the camera — and you get a gesture about the size of a thumbnail in frame. On paper, examined at desk distance, it reads clearly. On screen it's glanced at. The drawing would be telling you the shape of the gesture and nothing about whether anyone registers it, which is the actual live question.
That isn't a defect in the drawing. It's the boundary of what a still can hold. Examined is not the same as seen, and a treatment that treats them as the same will keep buying pictures that look convincing and answer nothing.
So: frames for a visible change you need to compare. Not for a change you have deliberately made invisible — and not for a question about timing, which no amount of shading will fix.
Use a timed study when time is the question
If the gap is simultaneity or duration, the representation has to have a clock in it. There is no still frame that answers "does the audience see this happen, or only the space where it happened?"
Back to the station. Two numbers matter, and both are assumptions the study exists to test rather than measurements anyone has taken. The exchange, played at a natural speed, takes roughly a second and a bit. The express — a six-car set, about a hundred and forty metres, through the station at the thirty-five metres per second already on the table — hides the two subjects for about four seconds.
Those two numbers together decide the beat. The action doesn't fit inside the occlusion in halves. It either finishes before the nose of the train reaches the pair, or it disappears entirely. There is no version where the audience catches the beginning and then the end, because the hole in the middle is four seconds wide and the whole gesture is barely one. Which leaves a real choice: either stage the exchange so it's fully visible just before the train covers them, or hide it completely on purpose and pay it off later. Both are legitimate commercials. Only one of them requires the audience to understand the swap at this moment.
The study itself is small. About six seconds of rough timing — the second before the train covers the pair, the four seconds of train, the second after — as three or four panels with durations written underneath, numbered, and marked provisional, because train speed, train length and distance along the platform all shift that four seconds and none of them is settled. The form doesn't matter much; three sketches and a stopwatch will do. What matters is that the question gets answered at the only speed where it's a question.
What the study settles is a placement: roughly how much clear time sits on either side of the occlusion, and therefore where the action has to start to be read. What it leaves open is almost everything else — whether the gag plays, whether the timing survives the music, whether that train exists on that day, whether the shot is shootable in the time available. A rough timed sequence is an argument about duration. It is not a feasibility test, and treating it as one is how a ten-minute sketch turns into a promise.
Keep the proof proportionate to the decision
Sort the options by what they cost and who is being asked to make them. A positional plan is minutes. Three timed panels are an hour plus a conversation. A cut animatic is a day or more of somebody's week. Boards of the whole commercial are days, and if they're finished rather than rough, they're the largest request in the room.
Before any of it starts, write one line about scope: what is being made, for whom, at what finish, and what it is not for. "Two rough frames and a six-second timing, for internal review, not for presentation" is enough. It costs nothing and prevents the ordinary drift where a rough study goes into a client deck and returns as an expectation.
Two labelling habits help more than they sound like they should. Mark whether a drawing is settled or exploring. A frame that fixes the courier's position at a mark is a decision somebody has made and can be held to; a frame that illustrates the sentence you already wrote is decoration, and the team should not discover the difference at the shoot. And keep the status of the object visible: a study is not a shot, a pitch illustration is not a committed frame, and an animatic is not proof that something is feasible.
Then there's the request that isn't a clarification at all. "We want the pitch to look more finished" is a different ask from "we can't tell what happens." Finish is a cost, not an explanation, and clarity of the request doesn't authorize unpaid expansion — if what's actually wanted is finished creative work before anyone is hired, that's a conversation about work and money, not about representation. Asking for it in the language of comprehension is a way of not having that conversation.
Finally, keep the prose. A treatment's sentences carry what pictures can't specify: why the courier does it, what she's afraid of, how the beat should feel. The drawings don't replace that. They answer the part of it the reader can't picture.
The smallest thing that answers the question
For the station beat, the answer isn't boards. If the geometry is disputed, one plan settles it in ten minutes; if it isn't, one sentence does. The frames aren't worth drawing, because the before and after are the same picture on purpose, and a drawing that looks clear at desk distance will flatter a moment that may not read at speed. What's left is one rough timed study of about six seconds, with the durations marked as guesses, answering exactly one question: how much clear time is there on either side of the train, and is that enough for the change to register?
Everything else stays open — and should stay open at this stage, because it hasn't been decided yet.
The general move is the same every time. Say out loud which question your treatment can't currently answer: where things are, what changed, what overlapped, or how long it took. Then buy the least drawing that answers it, at the finish that doesn't waste anyone's week. When the named question is answered, stop. More finish past that point isn't clarity. It's just cost.
Frequently asked questions
Does a commercial treatment need storyboards?
Usually not the full drawn sequence. The word covers four different objects — a positional plan, selected frames, a rough timed study, and full boards of the commercial's shots — and each answers a different kind of question at a very different cost. The method is to name which question the prose cannot currently answer, buy the least drawing that answers it, and stop when it is answered. More finish past that point is cost, not clarity.
How would a plan change the station beat?
A plan is a top-down or side-on diagram, ten minutes with a marker, and it settles who can see what and where two paths cross. In the station beat it shows the two sightlines crossing the train's path about 0.9 metres apart, which at line speed is roughly a fortieth of a second — shorter than a single frame. That kills, cheaply, the suggested idea that the train could hide one figure a beat before the other. What it cannot establish is what a viewer notices, how long anything takes, or whether the beat is funny.
Why aren't selected frames the right answer for this beat?
Because the before and after states are deliberately the same picture: two people standing, each holding a black duffel in the same hand at the same height. Drawing them yields two nearly identical sheets, and the handoff itself is a gesture about the size of a thumbnail at nine metres. A still fixes shape and says nothing about whether anyone registers it at speed. Examined is not the same as seen.
What does a timed study answer that a still cannot?
Anything involving simultaneity or duration has to be represented with a clock in it. Two assumed numbers decide this beat: the exchange takes about a second and a bit at natural speed, and the express hides the pair for about four seconds. The gesture cannot occupy the occlusion in halves, so it either finishes before the train reaches them or disappears entirely. The study answers how much clear time sits on either side of the occlusion and where the action must start to be read. It is an argument about duration, not a feasibility test.
What if the real request is that the pitch look more finished?
That is a different ask from not being able to tell what happens. Finish is a cost, not an explanation, and a request for finished creative work before anyone is hired is a conversation about work and money rather than about representation. Clarity of the request does not authorize unpaid expansion.