Pitch a Panel Show Through Its Conversation, Not Its Celebrity Guest List
Pitch a Panel Show Through Its Conversation, Not Its Celebrity Guest List
A panel-show pitch usually arrives with two things in it: a subject everyone nods at, and three names everyone recognizes. Cover the names with your hand and read what is left, because that is what somebody is being asked to say yes to.
Here is the version of that pitch I'll use as an invented example throughout. It is a working proposal, not a real show, and the three performers named in it are invented too.
What Makes an Entrance Funny — a half-hour studio panel with Ana Feld, Marcus Iyer and Roz Kim, on comedy's first silent beat.
Cover the cast line, and what remains is a subject, a runtime and a hope. The proposal doesn't say what the panel will be looking at, why these three rather than any other three, what the moderator does at minute nine, or what the room knows by the time it ends.
None of that is an argument against names. A recognizable face is a real thing to sell, and any producer is right to want several. But a name establishes who might be on screen. It cannot establish what happens once they are. Those are two different documents, and the second one is the one that survives a scheduling problem. Build the conversation first, and names can drop in and out of the cast line without the rest of the proposal moving. Build the names first, and the format gets reverse-engineered from whoever says yes, which is how four funny people end up agreeing with each other for twenty-two minutes.
Decide what the conversation is for, then decide what is in front of it
A panel is a container, not a genre. It can explain, argue, remember, rank, witness, tease. Each of those wants different raw material, a different moderator, and a different ending. None of them is the default.
So the first line of work is choosing what the conversation should produce. Comic exploration? Deliberation? A guided tour of something the audience doesn't know yet? Write it as a sentence someone could disagree with. Have a fun conversation about comedy is not a purpose; it's a wish. For the invented example: a silent entrance is funny only when the audience has been made to predict a completion that never arrives.
The purpose then chooses the prompt, which is where most pitches go soft. A topic asks people to hold opinions — Is physical comedy dead? — and opinions are cheap and endlessly repeatable. What a panel needs is something to examine. For the invented example: What could make an entrance funny before anyone speaks? Now there's an object instead of a theme, and an audience has something to picture.
In this proposal, the object is two described entrances. Nothing has been filmed or recorded; they exist as written descriptions in the document, which is all a proposal needs at this stage.
Entrance A. A person in a coat comes through a side door, walks to the middle of the space, stops for a beat, then turns and leaves the way they came. No line.
Entrance B. A person carries a folding chair in, opens it, and sits down facing away from the room.
Neither is funny on its own, and that's deliberate. Both hand the audience a purpose — a line is coming, a conversation is starting — and then withdraw it. Whether the withdrawal lands is precisely what this panel has been gathered to pick apart.
Notice one useful weakness. Entrance A's stop lasts "a beat." That's a word, not a length. The first thing an editor would want to do with that entrance is turn it into seconds. That's a good sign: the material is specific enough that a specialist already has somewhere to start.
Cast the perspectives before you cast the names
Fill the seats with jobs first. Put a name in each one afterwards, if you have one.
A seat is a job description that says what this person can do to the others' claims. Test it by covering the names again. Does the stage direction still make sense?
For the invented segment, three seats:
A performer who has made entrances and can stand up and make one without warning.
An editor who has cut comedy and thinks about timing after the fact, in units that can be changed without disturbing anything else in the shot.
A theatergoer who has sat in a great many rooms and remembers what those rooms did.
Three kinds of authority: making, shaping, receiving. The theatergoer isn't there as a decorative civilian. That seat has one job — to say what a person who doesn't know the craft does with the material before anyone explains it to them. If the seat can't do that, it isn't a seat. It's a guest.
Then run the interchangeability test. Could you swap the editor for a second performer and get the same conversation? If yes, one of those two seats is furniture. It isn't enough for each panelist to be interesting; each has to be able to question, extend or complicate something another one said.
Celebrity is a measure of how many people recognize someone, not of what that person can do to a claim. A famous comic and an unknown one might bring the identical contribution, and a pitch that can't tell the difference is selling recognition rather than conversation.
Keep availability honest while you're here. Attached, in negotiation and aspirational are three different states, and the proposal should say which one it's describing. Nothing in a pitch establishes that anyone has agreed to appear. Write the cast as a proposal and keep it in its own paragraph, where it can be revised without touching the format.
Map the paths instead of writing the lines
This is the part most pitches skip: what could actually happen between these people, given this material.
Not a transcript. A map of routes, at least two of them, written in the language of possibility rather than prediction. For the invented segment, a route where the panel converges might run like this. The performer and the theatergoer land on the same mechanism — both entrances withhold a completion the audience has already predicted — and then sharpen it against each other. The theatergoer's version is about which prediction the room actually made. The performer's is about what a body has to do to keep a prediction alive. The editor develops it rather than echoing it: on screen, the withheld completion can be placed exactly, because the cut decides the moment when the audience has finished guessing.
A route where the panel splits might run like this. The editor's claim is that duration is the operative variable. Two seconds of stillness reads as intention; six reads as malfunction. In a cutting room those are two different numbers on the same shot, and the difference between them is the whole effect. The theatergoer's claim is that duration only means something once the room knows who is walking. In a scene where we already know the character, a long stop reads as a decision. Opening cold with a stranger, the same length reads as a mishap. On that account, what the audience was holding when the silence began is the variable, and seconds are downstream of it.
Now the panel has a disagreement worth thirty seconds of air, and it arrived without anyone writing a line for anyone. That's the standard a pitch should hold itself to. Invented dialogue is the one part of a proposal that cannot survive contact with the person you've invented it for.
Two cautions. First, don't require the split. Some segments are strongest when the panel agrees and the interest lies in how precisely they can agree — how far a shared observation can be pushed before it cracks. Second, don't fake the split either. A format that needs conflict in order to function will manufacture it whether or not the material contains any.
Give the moderator moves, not charisma
The moderator is usually the vaguest line in a panel pitch: warm, quick, a safe pair of hands. Temperament is not a plan. The proposal should name what the moderator does.
Three situations worth preparing, using the invented segment.
When one voice dominates. The moderator doesn't ask the loudest panelist to stop; that's a correction, and it makes the room polite and dull. Instead, turn to someone who has been quiet and ask for an instance rather than an opinion: Tell me about an entrance you've seen fail. An instance can be worked on by the other two. An opinion can only be agreed with or contradicted.
When a specialist assumes knowledge. The performer says something like you have to earn the pause. Fine sentence, no handle on it. The moderator asks for one entrance where the pause was earned and one where it wasn't. A term has become two pieces of material, and everyone at the table can look at them.
When the exchange stalls. The moderator goes back to the original material and changes one condition. What happens to Entrance B if the chair is already in the room, so that the sitting down is the entire entrance? A changed condition produces new material without asking anyone for a fresh opinion.
One more move belongs to this prompt specifically. When the editor and the theatergoer begin restating their positions in the abstract — length, context, length, context — the moderator can ask the performer to stand up and run Entrance A twice, once with a short stop and once with a long one. Set that up honestly, though, because the demonstration proves less than it appears to. By the second run, everyone in the studio has already seen the first, so there is no clean first reading left to compare against. The demonstration makes duration visible and, in doing so, shows that what a room already knows changes what it sees. Both claims survive.
Which is why a useful moderator names the limit out loud rather than calling for a verdict the room can't produce. A live demonstration tests a live room. The editor's evidence lives in a cutting room, with a different audience and a different command of time. Saying so doesn't weaken the segment; it shows the panel knows the difference between what it has demonstrated and what it has merely asserted.
What the moderator is not: a referee, the person who knows the answer, or a guarantee against silence. A panel that goes quiet for a moment is not a format failure. A pitch that promises no silence is promising something it cannot deliver.
Let the segment stop where the question becomes specific
Panels rarely fail because they end badly. They fail because they don't end. They run out of time mid-thought, or they close with everybody agreeing that comedy is subjective, which is the conversational equivalent of switching the lights off.
The purpose you chose at the start decides what completion looks like. For a segment built on comic exploration, the conversation is complete when the panel has separated the variables and a viewer could say which one a given entrance depends on: how long the silence lasted, and what the audience was holding when it began. That's a clarified difference, not a verdict. Nobody has to be right for the segment to have delivered.
Consensus is not compulsory, and neither is a winner. A moderator who ends by awarding the point has imported a competition into a format that never described one.
It's worth testing the stopping point against a second prompt, to find out whether you've built a method or a vehicle. For the invented segment: Could the same entrance be funny twice? Same seats, same material, new condition — recognition. If the second prompt only works with these exact three people in this exact order, the format is a personality vehicle and the proposal should say so. If it works with the roles, it's a format, and the roles can be recast.
A pitch should also be able to say what the segment gives up. A half-hour conversation will not settle what makes an entrance funny, and a proposal that implies otherwise is setting the panel up to end with a shrug.
The same pitch with the names taken out
Here is the invented proposal with the cast line removed. This is what a buyer would be reading.
Format and purpose. Half-hour studio conversation, three seats and a moderator. Purpose: to test the claim that a silent entrance is funny only when the audience has been made to predict a completion that never arrives.
Prompt. What could make an entrance funny before anyone speaks?
Material. Two described entrances. Entrance A: enter, stop, leave. Entrance B: carry in a chair, sit down facing away. Written, not filmed. The length of A's stop is given only as "a beat."
Seats. Performer — makes entrances, can demonstrate one. Editor — thinks about timing after the fact, in adjustable units. Theatergoer — reports what audiences who don't know the craft do with the material, and remembers rooms rather than measuring them.
Possible path, convergence. Performer and theatergoer identify the same withheld completion; the editor adds that a cut can place the withholding exactly.
Possible path, divergence. Editor: the length of the silence is the variable. Theatergoer: prior footing is the variable, and length is downstream of it.
Moderator moves. Ask for an instance rather than an opinion when one voice dominates. Convert a term into two contrasting examples when a specialist assumes knowledge. Change one condition and return to the material when the exchange stalls. Ask for a demonstration when the claims go abstract, and say out loud what the demonstration cannot settle.
Stop condition. The two variables are separated and a viewer could say which one an entrance depends on. No winner declared.
Gives up. Any verdict on what makes an entrance funny. The half hour separates the variables; it does not settle them.
Renewal prompt. Could the same entrance be funny twice?
Open variable. Whether there is a studio audience. With one, the room's reaction becomes material the panel can work on in real time. Without one, the theatergoer's memory of other rooms is the only audience evidence in the studio, and that is a different segment.
Everything above that page — the title, the one-line description, the names — is packaging. If someone can read the map and hear the half hour, the cast line is the last thing they need in order to be persuaded, and the first thing the proposal can afford to lose.
Frequently asked questions
If the cast names are covered in a panel-show pitch, what should still be doing the persuading?
A subject, a runtime, and a hope are not enough by themselves. The remaining pitch should show what the panel will examine, why these roles rather than any other roles, what the moderator does at specific points, and what the room knows by the end. A name can establish who might be on screen; it cannot establish what happens once they are.
How do you turn a panel topic into a prompt the panel can work on?
Write the purpose as a sentence someone could disagree with, then let that purpose choose a prompt. A topic asks people to hold opinions, which are cheap and repeatable. A prompt gives them something to examine, such as 'What could make an entrance funny before anyone speaks?' rather than 'Is physical comedy dead?'
What is the interchangeability test for panel seats?
Fill the seats with jobs before names. Each seat should be able to question, extend, or complicate another panelist's claim. Test it by covering the names: if swapping an editor for a second performer produces the same conversation, one seat is furniture. Celebrity measures recognition, not what someone can do to a claim.
What practical moves can a moderator use when a panel dominates, assumes knowledge, or stalls?
When one voice dominates, ask a quieter person for an instance rather than an opinion. When a specialist uses a term like 'earn the pause,' ask for one entrance where it was earned and one where it was not. When the exchange stalls, return to the material and change one condition. A demonstration can make a variable visible, but the pitch should also name what the demonstration cannot settle.
How should a panel pitch define its ending and its limits?
The purpose chosen at the start decides completion. For comic exploration, the segment is complete when the panel has separated the variables and a viewer could say which one an entrance depends on: how long the silence lasted, and what the audience was holding when it began. No winner has to be declared. The pitch can also say what the segment gives up, such as any verdict on what makes an entrance funny.