Beat Sheet for a Vertical Micro-Drama

A film's three-act structure assumes a runtime long enough to earn a slow build. A vertical micro-drama doesn't have that runtime, and borrowing the film structure anyway is the most common reason a first attempt at the format falls flat: setup that would be beat one of act one in a feature has to already be resolved and moving by the time a 60-second episode ends. Deloitte's own description of the format is precise about the shape — "60- to 90-second serialized episodes loaded with plot twists and cliffhangers," built around a steady schedule of drops timed to build an appointment-viewing habit rather than one long piece released once. That's not a short film. It's a different structural unit entirely, and it needs its own beat sheet, not a compressed version of a longer one.

Four beats, not three acts

A workable shape for a single episode is four beats, and the discipline is treating each one as a genuinely separate unit rather than four points on a smooth continuous arc:

Beat 1 — Resolve. Open mid-consequence from the previous episode's cliffhanger, not with a recap. If episode three ended on a hand reaching for a door, episode four opens with the door already opening — the audience gets the payoff in the first few seconds, fast, without a "previously on" preamble eating runtime that doesn't exist. This beat exists to close last episode's open loop quickly enough that the episode still has room left for a new one.

Beat 2 — Establish. The new information or complication that makes this episode its own story, not just the fallout from the last one. This is where a viewer who's never seen the series before gets just enough to follow along, and where a returning viewer gets the thing they actually came back for.

Beat 3 — Escalate. The stakes from beat 2 get worse, sharper, or more personal. This is usually the longest of the four beats and the one that actually carries the episode's content — the other three exist to open and close around it.

Beat 4 — Cliff. The new interrupted beat that becomes next episode's beat 1. Cut before the consequence lands, not after — an episode that resolves its own cliffhanger inside its own runtime has given the viewer no reason to come back.

Four beats in 60-90 seconds means roughly 15-20 seconds each, which is tight enough that nothing survives that doesn't need to be there. A line of dialogue that doesn't move the beat forward, a shot that's beautiful but doesn't carry information, an establishing pause that a feature film could afford — all of it gets cut in a format built this dense, because the runtime doesn't have room to spend on anything that isn't actively resolving or opening a loop.

Why every beat has to do both jobs

The structural rule underneath all four beats is the same one, repeated: every episode has to resolve something and open something, and ideally every individual beat is doing a version of that same double duty rather than just advancing a single throughline. Beat 1 resolves the prior cliffhanger. Beat 4 opens the next one. But beats 2 and 3 aren't just filler in between — the strongest micro-drama writing treats beat 2 as closing a small question the audience didn't know they had (why does this character react that way to the door opening?) while beat 3 opens the bigger one that beat 4 will interrupt.

That's the actual craft difference between a micro-drama episode and a scene from a longer piece cut down to size. A scene excerpted from something longer assumes context before and after it. A micro-drama episode has to function as a closed unit on its own terms — someone's first episode has to work as an episode, not just as fragment fourteen of something they'd need to have watched from the start to follow.

Locking the cast so beat 1 actually reads as continuous

None of the beat structure works if the character doesn't look like the same person from beat 4 of episode three to beat 1 of episode four. Character consistency across a serialized format is a harder version of the ordinary consistency problem, because the gap between generations is measured in days of release schedule, not minutes of an editing session — there's no continuous working session where the wardrobe and face are fresh in view. A fixed character reference, supplied at the start of the series and reused for every episode rather than redescribed per episode, is what keeps beat 1's "the door is already opening" actually reading as the same door, the same hand, the same person the audience left mid-motion.

Versely's catalog carries several models in the reference-to-video category built specifically to hold a fixed identity across separate generations rather than reinventing it each time: Minimax H3 Reference to Video, Kling O3 Pro Reference to Video, Seedance 2.0 Reference to Video, and Wan 2.7 Reference to Video. The practical habit for a series: generate or select one canonical reference photo per character before episode one, and reuse that exact reference for every episode going forward rather than picking a new "good" shot from a prior episode each time — a reference that drifts slightly episode over episode compounds into a character who's noticeably different by episode ten, the same way a photocopy of a photocopy degrades.

Pacing the drop, not just the episode

The beat sheet inside an episode is only half of what Deloitte's description of the format points at — the other half is the release cadence itself. A format built around "a steady schedule of drops timed to build an appointment-viewing habit" is making a claim about pacing across episodes, not just within one: consistency of release timing is doing real work here, not just consistency of quality. An episode landing at an unpredictable time each day undercuts the appointment-viewing habit the format depends on just as much as a weak cliffhanger does — the two failures look different but cost the same thing, a viewer who doesn't come back for the next one.

Walkthrough: drafting one episode's beat sheet

1. Write beat 4 first, not beat 1 — decide what this episode's cliffhanger interrupts before writing anything that leads up to it. Working backward from the cliff keeps beats 2 and 3 honestly building toward something specific rather than meandering until a cliffhanger gets bolted on at the end.
2. Write beat 1 as the direct payoff of the previous episode's beat 4 — no recap, straight into the consequence, aiming for the first handful of seconds.
3. Write beat 2 as the smallest new complication that justifies the rest of the episode existing — one new piece of information, not several.
4. Write beat 3 as that complication escalating, and check that it's the beat carrying most of the episode's actual runtime.
5. Turn the four beats into a script and run it through Versely's story-to-video tool, which segments prose into scenes automatically and keeps named characters consistent in look, voice and wardrobe across the whole piece — the mechanism that makes reusing one canonical reference practical at series scale rather than a manual check on every episode.
6. Save the result as a reusable Versely workflow rather than a one-off generation, so the same cast and style are one click away for the next episode instead of a fresh setup each time.

The four-beat shape isn't a rule that produces good writing on its own — it's a container tight enough that weak beats get noticed immediately, because there's no slack runtime for them to hide in. That's the actual value of building to a beat sheet this specific: it's less forgiving than a looser structure, which is exactly what a format with no time to spare needs.