How to Write a TV Pilot

A pilot has a harder job than any other script. It has to tell a great story, and it has to prove there are a hundred more where that one came from. Most pilots I read manage one or the other. The ones that get people excited manage both.

A pilot isn't a short movie

The note I give most on pilots is some version of "this feels like a movie."

A movie is about change. The character starts one way, ends another, and the story's over. A pilot that spends all its pages on change burns its one chance to show us what a normal week of the show looks like. So a pilot has two jobs. It shows the transition, the event that knocks the lead's life into a new shape. And it previews the engine, the kind of story the show will tell every single week.

The event in a pilot also has to be bigger than a movie's inciting incident. Think of it as an emotional apocalypse. It permanently changes the character's world, and the world after it is the show.

Succession does both jobs cleanly. Logan Roy is supposed to hand his media empire to his son Kendall around his 80th birthday, then changes his mind, and by the end of the pilot he's had a brain hemorrhage. That's the transition. But the pilot also shows us every future episode: four grown kids circling their father's throne, each one sure they deserve it, none of them ready.

Start with the Star

Every pilot needs a Star, and the Star isn't necessarily the most likable character or the one with the most lines. The Star is the character who brings the most conflict.

Pilots, and most of the first season, are about the Star. Ensembles grow into ensembles. They rarely start as one. If your pilot gives five characters equal weight, a reader won't know who to hold onto, and neither will a network.

Give the Star a want the pilot won't resolve. That unresolvable problem is the series engine. The longer it takes to solve, the longer the show can run. I go deeper on building that kind of character in the TV character paradox post.

Cast by conflict

Every regular should exist because of the conflict they bring to the Star, not because the show "needs a best friend" or "needs a boss."

The antagonist is your #2 character, and they should be as complicated as the Star. Make it a person, not an abstraction. "The students" or "the system" can't argue back in a scene.

New Girl is the model. Jess is a relentlessly earnest, slightly dorky teacher who's just walked in on her boyfriend cheating, and the pilot drops her into a loft with three single guys who have no idea what to do with her. Each of them pushes on her in a different way, and Nick, a grumpy bartender stuck on his own breakup, is the clear #2: as complicated as she is, and the cynic to her optimist. That's a show that can generate stories forever, because the conflict is built into who's in the room.

Keep the pilot cast tight, four characters you really dig into at most. And give yourself two stories, both about the Star, weighted something like three to one between the A story and the B story. A C story can wait for the second draft, if it needs to exist at all.

The arena should test the dysfunction

Your arena is the world of the show, and your home base is the place it keeps coming back to: the bar, the office, the kitchen. Design that arena to test your lead's specific dysfunction. If you started with the arena instead, find the character it would test hardest.

A good gut check: is what's broken in the place the same thing that's broken in the people? When it is, every story you set there is automatically about your characters.

The 100-episode test

When you finish the pilot, ask yourself honestly: is it obvious there are thirteen episodes here? A hundred?

Try writing ten one-sentence episode ideas in ten minutes. If you stall at four, the engine isn't there yet. A few common engine killers:

• The mystery solves the show. If the pilot's central question gets answered, what's the rest of the series?
• Two characters and no one else. A two-hander can be a great movie. It's very hard to sustain as a series.
• The premise is a one-time event. A wedding, a heist or a trip ends. A job, a family or a town doesn't.

Don't save it for episode 17

Be honest about what a pilot is for. Most pilots, even great ones, never get made. What a great pilot does is get you read, get you repped and get you staffed. It's a writing sample.

So don't save your best material for later in the season. Nobody hires anybody off episode seventeen. Put everything you've got into these pages, and end on something that points to trouble ahead, a final beat that makes us want the next episode.

Page counts and structure

• Half-hour pilot: around 30 to 35 pages. Usually a two- or three-page teaser, then three acts of roughly ten pages each, with the last act the longest.
• Hour-long pilot: around 50 to 60 pages. Usually a teaser plus four or five acts, even for streaming shows with no commercial breaks.

Start each act at the top of a new page. Readers read toward the act breaks, so end every act on an escalation of the A story or a dilemma your Star can't wriggle out of. For half-hour comedy specifically, here's how the acts work.

And keep it lean. Nobody wants to read long from a writer they don't know yet. If a half-hour comes in at 40 pages, that's the first thing a reader will notice.

The teaser matters more than you think

The first two or three pages are where a reader decides whether they trust you. The best pilot teasers show the whole show in miniature: its tone, its central relationship and its core problem, all in a single scene. More on that in the cold open post.

Common pilot problems

These are the ones I flag most when I read pilots:

• It feels like a movie. All transition, no engine.
• Too many storylines. Focus on two and hit them hard.
• Exposition machines. Voiceover, talking heads and characters explaining backstory to each other are usually covering for a story that isn't dramatizing itself.
• It's the episode before the show. Nothing really happens until the last page, when the premise finally kicks in.
• The storylines never collide. The A and B stories should crash into each other near the end.
• The act-outs don't threaten anything. Each act break should put the Star's goal in real danger.

Once the pilot works, you'll probably need to pitch it, and maybe write a bible too.

Get a read on your pilot

Auteur covers TV pilots as well as features, with the same 20-category ratings grid and page-specific notes on structure, character and whether the engine is there. Send us your pilot and you'll have coverage back in minutes.