How to plan a 15-second skit that actually lands
Fifteen seconds is not a short film. It is a setup, a turn and an exit - and most first attempts spend the first five seconds on nothing.
Fifteen seconds sounds generous until you count what fits: roughly three shots, four spoken lines, one idea. That is the whole budget. Every first-draft skit I have seen - including mine - spends a third of it on establishing something nobody asked about.
So here is the structure that survives contact with a feed, and the timings that go with it.
Three beats, and what each one is for
- Beat 1, about 4 seconds: the situation, stated by someone doing something. Not narration. Somebody in a room, mid-action.
- Beat 2, about 5 seconds: the turn. New information, a reaction, the thing that makes beat 1 mean something else.
- Beat 3, about 4 seconds: the exit line, and then out. No wind-down.
Two characters is the sweet spot. One character is a monologue and needs a very good line to survive; three characters means somebody is standing there silently while the other two talk, and in a vertical frame that somebody is taking up a third of the picture.
Open on the second-most-interesting moment
The standard advice is to open on the most interesting moment. That is nearly right and slightly wrong: open one step before it, so the interesting thing happens on screen rather than having already happened. A character sipping coffee and then noticing something is a better opening than a character already reacting - the reaction needs something to be a reaction to.
What you are cutting is everything before that: the walk into the room, the greeting, the "so anyway". Start the clock at the first thing that would make someone stop scrolling.
The line-level rules
- Six to twelve words a line. Longer and it reads as two rows of caption in a vertical frame.
- No greetings. "Hey!" is a wasted second.
- The last line should be the shortest one in the video.
- Do not explain the joke in the third beat. If it needs explaining, the second beat is wrong.
Where the timing goes wrong
Three failure modes account for almost everything that feels off on first watch, and all three are fixed by editing rather than rewriting:
- Dead air at the start of a shot. Generated clips habitually begin with half a second of a character standing still. Trim it - every shot, every time.
- The punchline arriving after the exit. If the last shot runs on after the line, cut it there. Silence after a joke is only good on stage.
- Two beats doing the same job. If shot two and shot three are both reactions, one of them goes.
Watch it muted, then watch it once more
Two checks before you call it done. First, with the sound off: does the video still make sense? Most short video is watched muted, and if the beats do not read visually, subtitles are carrying more than they should. Second, at full attention: is there a single second you would skip? Cut that second. It is always there, and it is usually at the start of shot one.
Fifteen seconds is a tight budget, which is exactly why it is a good format to practise in. There is nowhere for a weak beat to hide, so you find out quickly which of your three ideas was the real one.