How to Write a Video Script, From Blank Page to Shot List
A video script is four decisions, not a document: what the first line promises, how long the middle is allowed to be, what you actually say, and what has to be on screen while you say it.
How to write a video script, in the order the decisions actually get made: choose what the first line promises, decide how long the middle is allowed to be, write the words, then pull the shot list out of what you wrote. Everything else — formatting, columns, software — is bookkeeping around those four.
What this covers, and what it does not
This is the writing and the timing: the hook, the beat structure, the runtime, the prompter version and the shot list. It is not a camera guide, and it says nothing about editing software.
It also assumes the video is under about fifteen minutes and has one person talking. That covers Shorts, Reels, TikToks and most YouTube uploads, and it is the shape where a script pays for itself fastest.
Decision one: what the first line promises
The opening line is not a warm-up for the video. It is a contract, and the rest of the video either keeps it or loses the audience it just earned.
There are only a handful of shapes an opening can take, and each one obliges you differently. A contradiction hook — "everything you've been told about this is backwards" — obliges you to say what everyone gets wrong and be right about it. A result hook obliges you to have the result and to show it. A question hook obliges you only to answer the question, which is why it is the safest one and the one most under-used.
Pick the shape before you write the line. A hook whose promise your video cannot keep is worse than a flat one, because it spends attention you then have to give back.
The generator fills six pattern classes with your topic and names the promise each class makes, so the choice in front of you is which shape fits what you are about to say. The lines themselves are scaffolding — rewrite the one you pick. More on the patterns and when each one earns its place is in video hook ideas.
One thing the hook must not do is overpromise in a way the platform has written a rule about. That is a narrower category than "clickbait" and it is worth knowing where the line is: clickbait titles covers which patterns are actually against the rules and which are just taste.
Decision two: how long the middle is allowed to be
Runtime is a budget, and a script is how you spend it. Words are the unit.
Speaking pace varies more than people expect: a Shorts voiceover runs near 180 words a minute and documentary narration nearer 115. That is a 55% difference in how much script fits in the same minute, which is why a single "words per minute" figure is useless and why the pace you actually deliver at is an input rather than a constant.
The total, though, is rarely the real problem. The real problem is that the setup takes ninety seconds and the payoff gets forty, and no total tells you that.
Paste the prose and it splits on the breaks you already wrote, prices each beat in seconds, and gives every one an in and out timecode. The column worth reading is the length one: a beat taking 40% of the runtime is almost always a beat with two ideas in it.
It also hands back the beats as a chapter list, with YouTube's three requirements checked — first chapter at 0:00, at least three of them, none under ten seconds (YouTube Help).
Decision three: the words themselves
A first draft is long. That is not a character flaw; it is what a first draft is for.
The share of it that is hedging, throat-clearing and long-winded construction is the share that pushes your ending off the end of the runtime — and a sixty-second Reel is sixty seconds whatever you meant to say. Cutting filler is usually sold as a style improvement. It is really a timing one.
It removes what carries nothing, shortens the constructions with an exact shorter equivalent, and prices the whole trim in seconds at your pace. What it will not do is cut a word that changes a claim, edit anything inside quotation marks, or rewrite a passive sentence — the reasoning for each refusal is in filler words.
The other half of the writing is which words are doing work at all, and where they fall. On any surface that truncates, the opening line carries everything; a draft whose first sentence is built entirely from function words has spent its one chance on nothing. Power words is about that, and about the class that matters more than any of them — numbers and named things.
Decision four: what is on screen while you say it
Two outputs come out of a finished script, and they are for two different moments.
The first is the version you read from. A prompter app will wrap your text for you, which is exactly the problem: it wraps on width, so a line can end on "the", and a reader who reaches the end of a line on "the" either pauses in the wrong place or races the scroll. Both are audible.
It reflows to your prompter width, breaking after a sentence or a clause instead, and marks where the breaths go. Nothing is added or removed. The width and the all-caps question are covered in teleprompter script format.
The second is the list you shoot from. Your script already contains every shot it needs — written down in brackets, in "cut to" lines, and in sentences that point at something the viewer has to see. Reading them out into one ordered list before you pick the camera up is the difference between one setup and three.
It separates what you wrote down explicitly from what it is guessing at, adds the talking-head coverage that makes the list complete, and exports a CSV you can reorder by location. Shot list template covers what belongs in one and what does not.
How to write a video script when you have nothing yet
In order, and stopping is allowed at any point:
Write the topic as a noun phrase. "Sourdough starters", not "How to Make a Sourdough
Starter (Full Guide)". This is the input to everything else.
Pick a hook pattern, then write the line. Shape first, wording second.
Write the middle as prose, badly and long. Do not format anything yet.
Cut the filler and read the seconds saved. This is where the runtime comes back.
Split it into beats and look at the lengths. Fix the beat that is 40% of the video.
Format for the prompter, and pull the shot list. Two outputs, two moments.
Steps four and five are the ones people skip, and they are the two that decide whether the video is the length you meant it to be.
Common questions this raises
Two things come up often enough to answer here rather than in a footnote.
Do short videos need a script? They need one more than long ones do. A sixty-second video has no room to recover from a wasted ten seconds, and the whole thing is short enough that writing it costs almost nothing.
Does scripting make you sound stiff? It does if you read a document. It does not if you cut the filler, break it into beats, and read it back aloud — which is three of the six steps above, and is most of what they are for.
How do you write a video script for beginners?
Write the topic as a noun phrase, pick a hook pattern, draft the middle as prose without formatting it, then cut the filler and split it into timed beats. The formatting is the last step, not the first.
How long should a video script be?
Multiply your speaking pace by the runtime you want. At a normal talking-head pace of around 150 words a minute, a five-minute video is roughly 750 spoken words.
What is the difference between a script and a shot list?
A script is what you say. A shot list is what has to be on screen while you say it, in the order it appears - and it is pulled out of the script rather than written separately.
Should the hook be written first or last?
Pick the hook pattern first, because it decides what the rest of the video has to deliver. Write the exact wording last, once you know what you actually said.
Start with the shape: run your topic through the video hook generator and pick the class before you write a word of the line.
Every opening line belongs to one of about six shapes, and each shape obliges the rest of the video differently. Pick the shape first; the wording is the easy half.
Between 105 and 180 words, and the range is not vagueness - it is the difference between a Reels voiceover and a live presentation. Pick the pace and the number is exact.
A shot list has six columns and one job: make sure nothing is discovered on shoot day. Here is the template, what each column is for, and the two columns people add that make it worse.