Cast a Commercial for How People Play Together, Not Separate Headshots
Cast a Commercial for How People Play Together, Not Separate Headshots
A casting page usually arrives looking like a row of individual portraits. Late twenties. Warm. Open face. Easy laugh. Reads approachable. Then, at the bottom, the sentence that carries the entire scene: Needs chemistry with the other lead.
That last line asks for something the rest of the page cannot describe. Chemistry isn't a quality anyone has by themselves. It is a thing that happens between two people during a particular action, and if you want a collaborator to go find it, you have to hand over the action.
Fictional exchange, invented for this piece: two housemates, Ada and Ben, are making lunch in a shared kitchen. Ada reaches for a lid. Ben, meaning to be helpful, lifts one off the stack and hands it to her. She takes it, sets it on the container, and it's the wrong size — a lid from something else in the cupboard. She glances up. He's already seen it. She hands it back, he gives her the right one, and they both turn to the counter again. No lines. No apology. The task continues.
That's the whole commercial beat, or anyway the part casting has to solve. What is being requested from two performers is not a shared trait but a passage of behavior: she registers the wrongness, he receives that registration, the correction happens, and neither person ends the moment smaller than they started. Nobody in that exchange has charm in the abstract. What they have — or don't — is a way of handling each other.
Everything below is a way of writing that down.
Name the relationship the scene needs to reveal
Before you can describe the pair, describe the exchange. The useful versions tend to fall into a few shapes.
Anticipation. One person reaches before the other asks, and the scene's meaning lives in whether the reaching was welcome, presumptuous, or simply correct. A routine that breaks. Something both people have done a hundred times stops working, and the question becomes who adjusts. An interruption that changes who leads. A third person, a dropped object, a phone; someone was running the moment and now someone else is.
Pick one, then write what each person contributes to it. In the lunch beat, Ada contributes the noticing and the decision not to make anything of it. Ben contributes the read — the speed with which he sees that his helpful move landed wrong — and the lack of ceremony in the fix. Neither contribution makes sense alone. That's the test for your own sentence: if you could say it about one performer standing in an empty room, it isn't a relationship yet.
And it's worth saying plainly that "chemistry" is the name of the question here, not the answer. A treatment page that says these two must spark has told a collaborator the desired temperature without saying what burns. The instruction to investigate is real and legitimate — you're asking someone to go find out whether two people can do this, and you can't find that out from a photograph. But the treatment's job is to make the question askable, not to announce the result.
One more caution that saves a lot of time later: a real friendship, a marriage, a pair of siblings, a set of actual housemates — none of that establishes the screen behavior you need. It's a reasonable thing to be curious about, and it is not evidence. Two people who genuinely finish each other's sentences may be unable to show a stranger anything on camera, and two performers who met in the waiting room may find the timing in ten minutes. The relationship is on the page, not in anyone's private life, and casting can only assess what shows up in shared action.
Turn individual descriptions into complementary actions
Here's a typical list, the kind that survives into a treatment because each line sounds like a decision:
Ada — warm, quick, a little scattered. Ben — steady, capable, dry sense of humor. Both effortlessly charming.
Now the same two people described so that the traits only exist in relation to each other:
Ada moves fast and changes course mid-gesture; she's the one reaching for a lid that isn't there yet. Ben watches her hands, not her face, and is usually a half-step ahead of her — handing over what she's about to want. When he's wrong, he corrects inside the same movement, without a pause that would make it an event.
Notice what happened. "Scattered" became speed, but speed measured against somebody else's read. "Steady" became the half-step, which only means anything next to someone who keeps moving. "Dry" disappeared, because dry is a way of talking and this scene may not have dialogue at all.
The relational rewrite doesn't just read better. It tells a casting collaborator where to look. Speed matters only in relation to another person's pace or need. Confidence can be a gift or a tax, depending on what it does to the partner's contribution: a performer whose assurance fills every silence may be exactly right for a scene about being taken care of, and ruinous for one where the second person has to arrive at something themselves. If your page describes Ben as confident, you have described him; you haven't said whether that confidence gives Ada room or takes it.
There's a second reason to make this conversion: it forces you to separate what the story requires from what you merely prefer. Story requirements are things like the correction is noticed and completed without either person losing standing — miss that and the lunch scene becomes a small humiliation, or a scolding, or a joke at somebody's expense. Appearance preferences are everything else, and they should sit in their own small open list, clearly marked as open, so that nobody reads them as a lock. Ages, genders, resemblance or the deliberate absence of it, family likeness, whether the pair reads as lifelong or newly met — for a four-second kitchen exchange, most of this is available. Say so. A collaborator who receives a locked list will cast to the list; a collaborator who receives an open one may bring you a pairing you wouldn't have pictured, which is the point of hiring them.
One boundary worth writing into the brief itself: the wrong lid must not read as a product failure. It's a lid from the wrong container, and nothing on screen implies the container is defective. The ad's story is the adjustment, not the flaw. That sounds obvious until you watch the beat played with a lingering close-up on a lid that doesn't fit, at which point it is a different commercial.
Leave more than one convincing pairing available
The same story event — wrong lid, notice, swap, resume — can be played at least two ways. Write both into the treatment, and say what each one buys you.
The practiced route. Ben catches his own error as the lid leaves his hand; Ada's fingers are already closing on the right one from the stack. The wrong lid barely lands. Nothing is negotiated, nothing needs acknowledging, and they're back at the counter inside the same gesture. Tonally this reads as shorthand, an old ease, a household with a rhythm. It also has a risk: a closed loop can look like an inside joke the audience isn't in on, and a routine that efficient can leave no visible room for anything to change.
The tentative route. Ada holds the wrong lid, and instead of setting it down and helping herself, she waits — half a beat, eyes up — until Ben takes it back. He does. She takes the right one. Then the task resumes. Tonally this reads as courtesy, a small adjustment made on purpose by two people who haven't yet assumed they can skip the acknowledgment.
Both routes preserve the event. What changes is who the pair seems to be, and what the scene seems to be about. For a stipulated brief about learning to share a home, the tentative route is the one with room in it — the ad's story is a small adjustment being made deliberately, and the wait for acknowledgment is that story. For a long-running campaign about an already-established partnership, the practiced route is likely stronger, because the ad's story is that these two have this and you're watching it work. Neither is the good one in general. The brief decides, and if your treatment names the brief's requirement, the choice explains itself.
What stays essential across both: the mismatch is noticed, the correction completes, the task resumes, and neither person's standing moves. What's flexible: everything about who they are. And a warning against a shortcut that looks like insight — demographic similarity, demographic contrast, or a real-life bond does not tell you which route a pair will find, or whether they'll find either. Those are assumptions dressed as casting instinct. The only place the answer lives is in the doing.
Here is what all of this turns into on the page, as a proposed paragraph for discussion:
Pair, lunch scene. One kitchen, two people, one wrong lid. What the scene requires: the person handing over the lid can register a mistake in the other person's face and take it back without making a gift of it or an apology out of it; the person receiving it can hold or set down the wrong lid without shrinking and lift their eyes without accusing. Both can complete the swap and turn back to the counter inside the same small beat, then work side by side without ceremony. Open: ages, genders, resemblance or none, and whether the two read as long-time housemates or recent ones. Not required: matching looks, contrasting looks, or charm held individually.
No pair described here has been cast, auditioned, or assessed. This is a casting proposition — a statement of what the selection has to support — and it's the useful kind precisely because it doesn't pretend to be a result.
Hand over the question, not an invasive exercise
A treatment can name what must be visible between two performers. It cannot decide how anyone finds out whether it's there. That belongs to the casting collaborator and the production team, and handing it over is not a retreat from rigor; it's the difference between a scene question and an audition protocol.
So: what should a collaborator be assessing in shared action? For the lunch beat, roughly this. Can this pair complete the swap and return to the counter, or does one of them stay in the moment after it's over? Does the pause read as consideration or as embarrassment? Because those are different commercials. If Ada plays the wait as I have done something wrong, the scene turns into a small humiliation, and a brief about sharing a home cannot survive that read. If Ben plays the correction as a favor granted, the same problem arrives from the other direction. What you're looking for is an adjustment that costs nothing.
Evidence that would inform the choice comes from paired work on that fragment — the doing, not a description of it. Whether the two have worked together, what they each make of the beat, how it changes the second time they run it: all of that is fair game for a collaborator to weigh. The method is theirs. Your contribution is the question, stated so that more than one combination of performers could answer it.
Two things a treatment is not. It is not a guarantee of compatibility. Writing these two will have chemistry guarantees nothing; it just commits you to a claim you can't support. And it is not permission to manufacture intimacy by asking two people to disclose private history to each other in a room. That approach is intrusive, it's unreliable — personal disclosure tells you about a person's willingness to be open, not about their timing with a partner — and it puts your collaborator in a position they didn't agree to. Keep the scene's intimacy in the scene.
For the record: the piece that would strengthen this most is a casting director describing, in their own words, what they watch for when they're evaluating a pair rather than a headshot. I don't have that account in hand, so the argument above rests on how the exchange has to work rather than on reporting from inside a casting session. Worth getting before this page is more than a proposal.
What you're actually asking for
Strip it back and the request is three sentences long.
The required exchange: one person offers the wrong thing, the other notices, the correction happens between them, and the task continues with neither person diminished by it. The flexible attributes: who these people are — their ages, their looks, whether they resemble each other, whether they've lived together for a decade or a week. The open question: who can make that wrong lid cost nothing, and can they do it in the same beat, without a pause that turns the moment into an event?
Note that the question is about a screen relationship. It is not a request for evidence that two performers' private lives match the roles, and it shouldn't be read as one — by your collaborator, by the production team, or by the actors who get the page. You are casting an exchange. The people are how you get there, and the sooner the treatment says so, the sooner somebody can go find it.
Frequently asked questions
How can a casting treatment describe chemistry rather than just asking for it?
State the exchange the pair must carry. In the lunch example, one person offers the wrong lid, the other notices, the correction happens between them, and the task continues with neither person diminished. Chemistry is the name of the question, not the answer; the treatment should make that question askable rather than announce the result.
Why are individual trait lists like 'warm, quick, scattered' not enough?
Those traits describe people standing alone. A relational rewrite makes them exist only in relation to each other: speed measured against someone else's read, steadiness as a half-step ahead of a partner who keeps moving, or confidence that either gives room or takes it. If a sentence could describe one performer in an empty room, it is not a relationship yet.
How should open appearance preferences be handled?
Separate story requirements from preferences. A requirement is something like the correction being noticed and completed without either person losing standing. Ages, genders, resemblance or its absence, and whether the pair reads as long-time or recent housemates should sit in a small open list clearly marked as open, because a locked list leads a collaborator to cast to the list rather than bring an unexpected pairing.
Why write more than one convincing pairing route?
The same wrong-lid event can be played as a practiced route, where the error barely lands and the pair moves on inside the same gesture, or a tentative route, where one waits half a beat for the other to take the wrong lid back. The practiced route reads as shorthand but risks a closed loop; the tentative route reads as courtesy and deliberate adjustment. The brief decides which fits, and neither is good in general.
What should a treatment not ask of performers?
It should not manufacture intimacy by asking two people to disclose private history to each other. That is intrusive, unreliable because disclosure shows willingness to be open rather than timing with a partner, and puts the collaborator in a position they did not agree to. Keep the scene's intimacy in the scene. Also, no pair described in the piece has been cast, auditioned or assessed, and no casting director's first-person account is in hand; the argument rests on how the exchange has to work.