How to write a vertical drama (microdrama): a screenwriting guide

A vertical drama, or microdrama, is a scripted series told in episodes of one to three minutes, shot in portrait format to be watched on a phone. To write one, you plan the whole season as a single arc and write each episode as a unit of retention: a hook in the first seconds, a conflict that escalates, a twist, and a cliffhanger that sends the viewer straight into the next episode.

This guide covers episode length, episode structure, season size, the paywall, tropes, dialogue for a small screen and script format.

What is a vertical drama?

It's serialized drama built for the thumb. Episodes are short, the frame is vertical — the phone held upright — and the story runs across dozens of episodes. The format grew up on dedicated apps such as ReelShort, DramaBox and ShortMax.

The difference from a TV series isn't just running time. A TV episode has acts and room to breathe. A vertical drama episode has to earn its place in seconds while pushing an arc that won't resolve for dozens of episodes.

That's why a vertical drama is not a feature cut into pieces. It's a format with its own grammar.

How long is a vertical drama episode?

Usually one to three minutes. Film Threat describes episodes in that range, built from a short setup, a development section and a 10–15 second cliffhanger at the end.

Within that range, length is a call you make with whoever commissioned the show. What matters is deciding before you write and measuring as you go. An episode planned at 90 seconds that comes in at 140 isn't a richer episode. It's an episode that lost the timing of its cliffhanger.

How do you structure an episode?

Think in microbeats, not acts. A grammar that works for episodes of about 90 seconds has five parts: hook, setup, conflict, twist and cliffhanger. The hook takes the first three seconds; the rest is distributed as the scene demands.

The table below is a starting point for a 90-second episode. The time ranges are illustrative — adjust them to your scene.

Time	Function	What the scene has to do
0–3 s	Hook	An image, gesture or line that stops the thumb. No introductions: start in the middle of the tension.
3–15 s	Setup	Re-anchor who's on screen, where, and what's at stake — usually by picking up the previous episode's cliffhanger.
15–60 s	Conflict	Put two wants in collision, on screen, and raise the pressure with every exchange.
60–78 s	Twist	Flip what the viewer thought they knew: a reveal, a betrayal, an unexpected entrance.
78–90 s	Cliffhanger	Cut on the question, not the answer. The last image leaves something open that only the next episode closes.

Here's an invented example to watch the table at work. Nora, a hotel housekeeper, is accused of stealing a ring.

0–3 s: close on Nora's hand closing around an engagement ring that isn't hers.
3–15 s: the manager gives her ten minutes to produce the ring before she calls the police.
15–60 s: the guest who owns the ring accuses Nora in front of the staff. Nora can't defend herself without explaining why she was in the room.
60–78 s: the guest's fiancé walks in. He's the man who vanished on the day of his wedding to Nora, five years earlier.
78–90 s: he recognizes her, goes pale and says, "You shouldn't be here." Cut.

Notice that each episode's setup carries the question the previous one left open. That's what stitches the season together, one episode at a time.

How many episodes are in a season?

It depends on the platform and the project. Film Threat puts typical season plans at 50 to 100 episodes.

The number matters less than the shape: the season is the unit of the arc. The episode holds attention; the season tells the story. So outline the whole season first — where the central conflict sets in, where it turns, where the paywall sits, how it all ends — and only then write episodes.

Writers who start at episode 1 and keep going without a season outline tend to find, somewhere in the middle, that their episodes repeat the same conflict instead of escalating it.

Where does the paywall go, and why is it a writing decision?

The paywall is the episode after which the viewer has to pay to keep watching. Which episode that is depends on the market and the deal; there's no universal number.

What doesn't change: the paywall is a dramatic moment, not a platform setting. It's the point where someone decides whether this story is worth money. Three questions help you write it:

What has the viewer already gotten? The free episodes have to deliver the trope they promised and at least one clear payoff — not just more promise.
What's the strongest question in the season so far? The paywall episode ends on it.
What does the next episode pay off? Someone who just paid needs a reward soon, or the trust breaks.

If the paywall moves, the season moves around it. Pushing the paywall five episodes later isn't a configuration change. It means rewriting everything between the old spot and the new one.

Which tropes work?

In vertical drama, a trope is a contract with the audience: the premise promises an experience, and the viewer comes back to collect it. Secret heiress, fake marriage that turns into real love, mafia bodyguard — viewers recognize the shape in seconds, and that's an advantage in a format with no time for introductions.

Tropes work when you:

Declare the trope early. Episode 1 makes it clear what kind of story this is.
Hit the beats it promises. In the Telos library, Fake Marriage, Real Love runs through five beats: contract signed, first genuine moment, jealousy that exposes the feeling, the contract threatening to end, and a real marriage.
Vary the execution, not the promise. The surprise lives in how a beat happens, not in swapping it for a different one.
Build the conflict into the premise. If you have to invent a new conflict every episode, the season wears out. A hidden heiress working for the people who humiliate her generates conflict on her own.

How do you write dialogue and action for a phone screen?

The screen is small and vertical, so the face fills the frame. The blog of Filmustage — a company that sells production software, so a vendor source — recommends writing for the close-up and treating dialogue as action. In practice:

Write the reaction. In a close-up, what a character does with their eyes and hands is the scene. Put it on the page: whether she looks away, almost speaks, grips the object.
Lines are moves. Every line attacks, deflects, denies or gives ground. A character explaining how they feel is standing still.
Keep lines short. A monologue eats time the episode doesn't have.
Keep the frame small. Two people in a standoff play better than a crowded room the vertical frame can't hold.
Keep geography minimal. Describe the space only where it matters to the action. A wide shot of a city skyline burns seconds and barely reads on a phone.

How do you format the script?

Use standard screenplay format — scene headings, action, character, dialogue — and mark clearly where each episode begins. Final Draft added "Micro Drama" and "ReelShort Vertical Drama" templates to the program.

EPISODE 12

INT. HOTEL - ROOM 1408 - NIGHT

Close on NORA's (28) hand as it closes around an engagement ring.

She looks at the door. Footsteps in the hallway.

MANAGER (O.S.)
Nora. Ten minutes.

To measure running time, pick one ruler and stick with it: 1 page ≈ 60 seconds of screen time. That's the ruler Telos uses. By that count, a 90-second episode runs about a page and a half, and a two-minute episode runs two pages. If the PDF shows three pages, the episode won't fit in 90 seconds, no matter how fast the cast talks.

Measure every episode, not just the season. Running 20 seconds long per episode looks harmless until every single one becomes a problem in the edit.

Common mistakes

Slicing up a feature. Cutting a traditional screenplay into 90-second chunks gives you episodes with no hook and no cliffhanger.
Opening with introductions. The first three seconds aren't for orientation; they're for grabbing the viewer.
Resolving the cliffhanger in the same episode. If the question gets answered before the cut, it isn't a cliffhanger.
Eyeballing length. Without a ruler, the season runs long one episode at a time.
Treating the paywall as a setting. It decides whether the story is worth money; it deserves the care you'd give a climax.
Writing for a horizontal screen. Wide shots, group scenes and choreographed action get lost in a vertical frame.
Starting without a season outline. Without the season's arc mapped, the middle episodes repeat conflict instead of escalating it.

How Telos helps

Telos was born from a vertical drama, and the format has its own path in the system:

Length measured on the page. Every episode has a running-time target you set — 90 seconds by default, adjustable episode by episode — measured on the page at 1 page ≈ 60 seconds. The AI writes to fit the target.
The season as the unit of the arc. The initial season targets 35 episodes by default and goes up to 90.
The episode's grammar. Hook (0–3 s) → setup → conflict → twist → cliffhanger.
The paywall as a story decision. The paywall episode is suggested from the market's range, and the writing queue stops there, at every arc turn and at the end of each six-episode block.
Tropes with their beats. A library of 64 industry tropes, each with its beats.
The AI writes from your project. World, characters, structure and your voice go into every generation.
Exports to PDF, Final Draft (.fdx) and Fountain.

Plans and credits at /en/pricing.

Written with AI assistance, from the sources cited.