I’ve written more than 300 video scripts over the years, from 30-second commercials and product videos to corporate films, interviews, training videos and animation.
By
Larry Mutenda
Quick answer: To write a video script, start by nailing down who it is for and the one action it should drive. Outline the hook, the main points, and the call to action, then write the spoken lines the way people talk. Add the matching visuals in a second column, read the whole thing out loud against a stopwatch, and cut it to length before anyone shoots or animates. For most branded video, use a two-column A/V script.
One thing I’ve learned is that a good video usually starts long before anyone picks up a camera. It starts with the script. The script is where you decide who you’re talking to, what you actually need to say, and what the viewer should see while you’re saying it. It’s also the cheapest place to solve problems. Changing a sentence takes a few seconds. Changing that same idea after a shoot, or after several scenes have been animated, can take hours or days.
In this guide I’ll show you how to write a video script that connects with your audience, and the process I use to go from a brief to a production-ready script. How I structure the message, how many words I write for different lengths, how I write for live action versus animation, and the script format we actually use in production.
A video script is the written plan for everything a viewer will hear and see, in the order it happens. Get it right and the shoot runs faster, the edit is cleaner, and the video stays on message. Get it wrong and you pay for it in wasted shoot days or reshot animation. This guide covers how to write a video script, from a blank page to a locked draft. It works for live-action and for animation, for a fifteen second ad or a longer brand film. You get the formats the industry actually uses, the structure, real word-count targets, a free template, and full example scripts you can copy.
A video script is the document that sets out what people say, what appears on screen, and the sequence it all runs in. It is the master reference for a video. Everyone who touches the project works from it, from the client approving the idea, to the crew on set, to the editor in post.
Most scripts do two jobs at once. They carry the words, whether that is narration, dialogue, or a voiceover. And they carry the on-screen action, meaning the shots, the visuals, the text, and the graphics that play against those words. In a simple script both live in one column. In a professional one they sit side by side so you can see how sound and picture line up.
A few things come up on almost every project. These are the ones I would tell anyone writing their first script.
Start with what the audience needs to hear, not everything the company wants to say. Companies naturally want to fit several messages, features and internal priorities into a short runtime. The writer’s job is to find the one message the viewer actually needs.
The first draft can be too long, and that is fine. Get the thinking onto the page, then cut. Do not force every idea into the final runtime while you are still working out what the video is about. Write with the picture in mind.
Read the script out loud. A sentence can look fine on the page and sound awkward when a person has to say it. Spoken language has a different rhythm than written copy, and the only way to catch that is to hear it. If the voiceover is explaining everything, the visuals are not doing enough work. A shot, a product screen, an expression or a graphic can often carry part of the message faster than a line of narration.
A script is not the same as an outline, a storyboard, or a shot list, and the four get confused often. The outline comes first and is just the order of ideas. The script turns that into exact words and action. The storyboard draws each shot. The shot list is the on-set checklist. The table below shows where each one fits.
| Document | What it is | When you use it |
|---|---|---|
| Outline | A bullet list of ideas in rough order | First, to plan the video before writing |
| Script | The exact words plus the on-screen action | Before the shoot or build, as the master reference |
| Storyboard | Frame-by-frame drawings of each shot | After the script, to plan the visuals |
| Shot list | A checklist of every shot to capture | On set, so nothing is missed |
You do not always need all four. A thirty second social video might go straight from a script to the shoot. A complex animation needs the script, a storyboard, and often an animatic on top. The script is the one document almost every project has in common.
Writing the script first is the cheapest way to protect the most expensive parts of a project. A shoot day costs money by the hour. Every second of animation costs hours of work to build. A locked script means nobody is inventing content on the clock.
A script also keeps the video on one message. Without a plan it is easy to wander, and wandering is what makes viewers leave. When the words are written down you can cut anything that does not earn its place before it costs you a frame. You can also hand the same document to a client and get sign-off before the crew is booked, which is far cheaper than changing direction after the shoot.
It aligns the whole team too. The camera operator, the sound recordist, the animator, and the editor all read from one page. On a two-column script they can see exactly what the viewer hears against what they see at each moment. That shared reference is what keeps a shoot day calm and a project on schedule.
The video below is an educational video we made for LeanBC, explaining the DMAIC method. On a process explainer like this, the script is where the method gets checked. Every step has to be right on the page before anyone animates a scene, because fixing a wrong step after the build means redoing the scenes around it.
Most video scripts share the same four parts: a hook, a setup, a body, and a call to action. The lengths shift with the format, but the order rarely changes.
The video below is a marketing video for Blockdaemon, a blockchain infrastructure platform, and it runs about a minute. You can hear the four parts in order. It opens on one line that says what the company is. The body carries the proof: the security layers, the monitoring, the insurance. Then it ends on a single ask, sign up. One message, one action.
The first five to ten seconds decide whether anyone keeps watching, so the hook has to work fast. A few types earn their keep: a sharp question the viewer wants answered, a promise of what they will get, a pattern break that stops the scroll, or a problem stated out loud. Do not open on a logo or a slow build. Give the viewer a reason to stay before you ask for anything.
Once you have attention, the setup tells the viewer they are in the right place. Say who the video is for and what it will cover. Keep it short. Ten to thirty seconds is usually enough. This is where you earn a little trust and set the expectation for the rest of the video.
The body carries the main content, and its job is to hold attention while it delivers. Break it into clear sections and change something every sixty to ninety seconds. A new location, a new speaker, a shift in music, or a fresh visual resets the viewer before their attention drifts. Make one point at a time and move on. If a sentence does not serve the goal, it does not belong.
End with one clear ask. Visit a page, book a call, subscribe, or try the product. One ask, not five. A crowded ending gives the viewer too many choices, and too many choices usually means they do nothing. Make the single next step obvious and easy.
There are three script formats you will run into, and the job decides which one to use. The two-column A/V script is the professional standard for most branded and commercial video. A single-column script suits simple pieces. Screenplay format is for narrative work.
The two-column, or A/V, script splits the page into audio on one side and video on the other. The audio column holds everything the viewer hears, so music, sound effects, dialogue, and voiceover. The video column holds everything they see, so the shots, camera framing and movement, on-screen text, and graphics. Because the two run side by side, you can line up sound and picture before anything is filmed. That is why it is the standard for commercials, corporate video, product video, and most animation. It also lets a client see what the video will look and sound like while it is still just a document.
A single-column script is a straight run of narration or dialogue with visual notes in brackets. It is quicker to write and fine for talking-head pieces, simple social clips, and rough drafts. The trade-off is control. You get less precision over how sound and picture align, so it suits shorter or looser projects rather than a tightly timed ad.
Screenplay format is the film and television standard, built around scene headings, action lines, and centred dialogue. Use it for narrative work, short films, and scripted ads that play like a scene. For a thirty second product video it is more structure than you need, and an A/V script will serve you better.
| Format | What it looks like | Best for |
|---|---|---|
| Two-column (A/V) | Audio in one column, the matching video in the other | Commercials, corporate, product video, animation |
| Single column | Narration or dialogue with visual notes in brackets | Talking-head, simple and social videos, rough drafts |
| Screenplay | Scene headings, action lines, centred dialogue | Short films and scripted, scene-based ads |
When I’m estimating a script before we’ve hired the voice actor, I budget around 130 to 150 spoken words per minute. I treat that as a ceiling, not a target. Product shots, pauses, reactions, titles and transitions all need room too. Decide the run time first, then work back to a word count. A 30-second spot lands around 60 to 75 words. A 60-second video sits near 130 to 150.
| Video length | Word count (about 150 wpm) | Common use |
|---|---|---|
| 15 seconds | 30 to 38 words | Social bumper, short ad |
| 30 seconds | 60 to 75 words | Ad spot, product teaser |
| 60 seconds | 130 to 150 words | Brand video, social feature |
| 90 seconds | 195 to 225 words | Product video, landing page |
| 2 minutes | 260 to 300 words | Case study, how-to |
| 5 minutes | 650 to 750 words | Training, detailed walkthrough |
Treat these as ceilings, not targets to fill. Read pace changes everything. A slow, deliberate luxury or health read drops closer to two words a second, which cuts the count by a fifth. A fast retail read can push past three. Silence and visuals eat time too. A logo hold, a product shot, or a reaction beat all burn seconds without a single word. So here’s the rule I work by: I use word count to estimate, and I use the read-through to decide. Read the full script out loud at the intended pace, pauses in, against a stopwatch. Trust the clock over the count every time.
The video below is a marketing video for CryptoSea, and it’s a real check on the table above. It runs 60 seconds and the voiceover is 136 words, inside the 130 to 150 I budget for a minute. That’s the ceiling doing its job. The words fit, and there’s still room for the visuals to breathe.
A video script is written to be heard, not read, so it should sound like a person talking. The fastest way to lose a viewer is copy that reads like a brochure. Write short sentences. Use contractions. Favour plain words over clever ones.
Read every line out loud as you draft it. If it trips your tongue, it will trip the presenter and the voice actor too. Keep to one idea per sentence, use active voice, and speak to one person, not a crowd. ‘You’ beats ‘our customers’ every time. Cut any word the sentence does not need.
On-camera dialogue and voiceover read differently, and it helps to know which you are writing. Voiceover can be tighter and more scripted because it is recorded in a booth. On-camera lines need room to sound natural, so leave space for the way real people pause and breathe. Write to the delivery, not just to the page.
The video below is a how-to video for JM Bullion on packing and shipping valuables. Listen to how it’s written. Short sentences. One instruction at a time. Gather your materials. Place your items. Seal the seams. It’s written the way you’d talk someone through it if they were standing next to you, even though it’s really a manual.
Before I write a line, I put together a short script brief. It is five answers, not a long discovery questionnaire.
The first two questions I ask on almost every project are: who are we talking to, and what do we want them to do after watching? Everything else follows from those.
The video below is a startup video for Copilot AI, and it shows what a clear brief looks like on screen. The first sentence names the person it’s for: someone responsible for lead generation or revenue growth. That’s question one answered out loud. There’s one message, finding your prospects on LinkedIn at scale, and one action at the end, book a time.
The order matters. Each step makes the next one cheaper. Here is the sequence I follow.
Before you write a word, put down who the video is for and the one thing it should get them to do. Every later choice flows from these two answers. A video for existing customers sounds nothing like one aimed at people who have never heard of you.
Pull together the facts, the key messages, and anything the video has to include, such as a product feature, a legal line, or a specific offer. Getting this in front of you now stops you stopping mid-draft to go hunting for it later.
Get the ideas out of your head in rough order before you write full sentences. Bullet the hook, the main points, and the call to action. An outline is fast to move around. A finished script is not, so it is far cheaper to fix the structure here.
Draft what people will say before you worry about visuals. Write the way people talk and read each line aloud as you go. Do not polish yet. A messy first draft you can cut is worth more than a perfect first paragraph and nothing after it.
Now pair each line with what the viewer sees. In a two-column script this is the second column. Be specific. ‘Wide shot of the workshop in morning light’ gives the crew something to shoot. ‘Nice footage here’ gives them nothing.
Read the whole script at a natural pace with a stopwatch running. This is where you find out it is too long, because the first draft always is. Cut anything that repeats, stalls, or drifts off the goal.
Share the draft with the people who need to sign off before anyone books a crew. Take the notes, make the changes, then lock the script. A locked script is what the shoot and the edit are built on, and changing it after that point is where budgets go to die.
The four parts stay the same. How you weight them changes with the job. A corporate video leans on real people and locations. A commercial strips everything back to one message. This is roughly how I shift the approach.
| Video type | How to approach the script |
|---|---|
| Brand film | Story, emotion, interviews and strong visual sequences. |
| Corporate video | Clear messaging balanced with real people, locations and supporting B-roll. |
| Product video | Problem, product value, demonstration and a clear next step. |
| Commercial | One main message, very little setup, tightly controlled runtime. |
| Testimonial / interview | Do not script the answer. Script the questions and know what story the edit needs. |
| Training video | Accuracy and clarity come before clever wording. |
| Animated video | Write the audio and visual action together so the picture carries part of the explanation. |
| Short social video | Get to the point quickly and build around one idea. |
The video below is a documentary we made for Couch to 50, the storytelling end of the table above. On a documentary the script is the shape of the story, the questions, and the moments the crew has to capture. The lines come from the people on camera, not from a page.
A live-action script has to respect the real world. You can only shoot what exists, or what you can build, in the time and light you have. Keep the video column grounded in shots a crew can actually get on the day. Note the locations, the on-camera talent, and any product that has to appear.
If a shot is not in the script, no one plans for it, and it will not be there on the day. Spell out the key visuals a crew or an editor needs, and note the B-roll that will give the edit options later.
Write for the budget as well as the story. Every extra location, actor, or setup adds cost and time. A script that calls for one location and two speakers is a very different quote from one that hops across a city. Match the ambition of the script to the resources the project actually has.
The video below is a live-action piece we made for International Paper. Everything in the video column had to be a shot a crew could actually get, with the real product in a real location. That’s the discipline of live action. The script can only promise what the shoot day can deliver.
Interviews are their own case, and the big rule is simple: do not script the answers. You script the questions, you know the story you need, and you plan how the answers get supported in the edit.
Write open questions that invite a full spoken answer rather than a yes or a no. Tell me about the day you started gets you a story. Did you enjoy it gets you one word. Then plan the B-roll that will cover the cuts, because you are shaping real answers in post, not reciting a page. For interviews, I do not script the answer. I script the questions and plan the edit.
The video below is an interview with Lloyd Tevedzai. I didn’t script a single answer. I scripted the questions, knew the story I needed, and planned the B-roll to cover the cuts. Everything you hear him say is his.
An animation script carries more weight than a live-action one, because nothing exists until someone builds it. There is no b-roll to fall back on and no happy accident on set. If an action is not described in the script, it will not appear on screen.
The video below is a 2D animation we made for Fujitsu on surgical capacity planning for hospitals. It runs over three minutes and the voiceover is about 400 words, read at a slower pace because the ideas are dense. Every scene, the scheduling engine, the what-if scenarios, the digital twin of an operating room, had to be described on the page clearly enough for an animator to build, because none of it existed until it was drawn.
That means the video column has to describe the action clearly enough for an animator to board and build it, whatever animation style the project uses. Be specific about what moves, in what order, and roughly how long each beat takes. Expect the writing itself to run longer. A page of live-action script is about a minute on screen, but a page of animation script is closer to forty seconds, so the same run time takes more written detail.
The order of work is different too. On most animation projects the voiceover is recorded from the locked script first, before any drawing starts. That scratch or final track then drives the storyboard and the animatic, which is the timed storyboard set to the audio. Because timing is baked in this early, a vague line like ‘character reacts’ can cost a day of work, so it pays to be precise on the page.
The script sits near the front of the process, right after the brief and before anything visual gets made. On a normal project the order runs brief, script, storyboard, production, then edit. Lock each stage before you start the next.
The brief sets the goal, the audience, and the message. The script turns that into words and action. From there a live-action project moves into the shot planning and pre-production I cover in how to make a corporate video, while an animation project moves to a storyboard and an animatic. The edit assembles the footage or the finished frames against the script. When the script changes late, every one of those downstream stages has to change with it, which is the whole reason to get sign-off before the script is locked.
The first draft is never the one that ships. The edit is where the video actually gets good. The first version usually tells me everything the company wants to say. The edit tells me what the audience actually needs to hear.
Here is a typical before and after on a single product-video opening.
Before: In today’s fast-moving world, businesses of all sizes are constantly looking for smarter, more efficient ways to manage their expenses and stay on top of their finances, which is why we built our platform.
After: Money is easy to spend and hard to see. We built something that fixes that.
The second version is shorter, easier to say, and it leaves room for the picture. I cut the setup, swapped written language for spoken language, dropped the duplicate ideas, and trusted the visuals to do part of the work. That is the pass I make every time: remove setup, replace brochure phrasing with speech, cut repeats, and give the picture something to do. If everything important is being explained by the voiceover, the picture is not doing enough work.
The clearest way to see all of this is on a real project. Here is a corporate video we produced for Providius, from the brief to the finished script. After it, three short format templates you can copy and fill in.
Watch the finished Providius video first, then read how the script behind it was built.
Providius makes real-time observability software for mission-critical networks, the kind that carry live media where a dropped packet or a timing error shows up on air. We produced a corporate video for them that runs about a minute and fifty, live action with motion-graphic overlays. The audience is the engineers and operations teams who feel that pain, so the script had to respect that they know their networks better than we do.
The challenge was focus. The platform maps network topologies, identifies devices, measures performance metrics, detects anomalies, monitors PTP and NTP timing, and profiles traffic for security risks. That is a long list, and a script that gives each feature equal weight says nothing. So we built the whole video on one spine: you cannot fix what you cannot see. Every capability in the script hangs off that single idea of real-time observability instead of competing with it.
The outline is simple. Open on the stakes, that media networks are mission critical and their failures stay invisible until they disrupt delivery. Turn to what Providius does, real-time observability and proactive anomaly detection. Show the capabilities as proof. Pay it off with fewer costly outages and faster resolution. Close on one ask, visit the site. The audio below is the delivered voiceover.
| Video | Audio |
|---|---|
| Data-center b-roll. Server racks with status lights. An engineer studies a wall of monitors. | NARRATOR (V.O.): Any organization working with media must ensure real time, low latency, and high quality delivery across the network. |
| Animated line icons for network, signal and an eye light up over the racks. Keywords surface: packet loss, latency, routing errors. | Challenges such as packet loss, latency, and multicast routing errors can disrupt seamless content delivery. |
| The Providius mark resolves. A live network topology map draws itself across the screen. | Providius ensures the uninterrupted delivery of time sensitive data across performance critical networks, providing real time observability and proactive anomaly detection. |
| Motion-graphic UI. A topology map with labelled devices and a live metrics readout. | Map network topologies, identify media devices, and measure key performance metrics to pinpoint vulnerabilities in critical paths across distributed networks. |
| A calm control room. An alert clears on the dashboard. | Providius delivers measurable ROI by minimizing costly outages, accelerating problem resolution, and saving valuable time. |
| End card. Providius logo and web address. | Discover how Providius can elevate your media workflows. Visit providius.com today. |
Here is a live-action brand video we produced for MTG Traders. It is a clean example of a tight script turning into a finished piece.
Notice how it opens on a clear hook and builds toward one message, rather than trying to say everything at once.
Here is an animated video we produced for Contruent. An animation script has to spell out the on-screen action in detail, because every shot is built rather than filmed.
Watch how the animation carries the explanation, with the voiceover and the visuals working together beat for beat.
A twenty second vertical video for social, written in a single column with visual notes in brackets.
HOOK (0 to 3s): [Creator to camera, fast pace] Stop scrolling. You are wasting money on this every month.
PROBLEM (3 to 8s): [Cut to a screen recording of a subscriptions list] Look at all these. Half of them you forgot about.
SOLUTION (8 to 15s): [Back to creator, holding phone] This app finds them and cancels the ones you do not use. Took me two minutes.
CTA (15 to 20s): [Text on screen: link in bio] The link is in my bio. Go check yours right now.
Why it works: the hook lands in the first three seconds, the whole thing is one idea, and the call to action is a single tap.
I use the same two-column skeleton for most scripts. Before the grid, the template header carries the details I set before writing: project name, audience, goal, core message, target runtime, and the call to action. Then the grid runs Time, Video, Audio, and On-screen text or production notes, one row per beat. Fill the audio column with what the viewer hears and the video column with what they see.
| Video | Audio |
|---|---|
| [Hook shot. What is on screen in the first five seconds.] | [Hook line. The reason to keep watching.] |
| [Setup visual.] | [Who this is for and what it covers.] |
| [Body visual 1.] | [Main point 1.] |
| [Body visual 2.] | [Main point 2.] |
| [Closing shot. Logo, product, or presenter.] | [One clear call to action.] |
For a quick single-column draft, drop the columns and write the narration in order, with each visual note in square brackets on its own line. It is faster to write and easy to expand into a two-column script later.
Start with two answers. Who is it for, and what should they do after watching. Outline the hook, the main points, and the call to action. Write the spoken lines the way you talk, read them out loud, then add the matching visuals. Time the draft and cut anything that runs long.
For most branded video the two-column A/V script is the standard, with audio in one column and the matching video in the other. Use a single column for simple talking-head or social videos, and screenplay format for narrative or scripted scenes.
An AV, or two-column, script puts audio and video side by side and is built for commercials, corporate video, and animation where sound and picture have to line up. A screenplay runs linearly with scene headings and dialogue and is built for narrative film and television. Most branded video uses an AV script.
About 130 to 150 words at a natural pace of roughly 2.5 words per second. On-screen text, pauses, and music can carry time without words, so read your draft against a stopwatch to confirm it fits.
Around 60 to 75 words at a natural speaking pace. Fewer if the video leans on visuals or on-screen text, or if the read is slow and deliberate.
No. The script holds the words and the on-screen action. The storyboard turns that into shot by shot drawings. Most projects use both, and the script always comes first.
Yes, and the script matters even more than in live-action. Nothing exists until it is built, so the script has to describe the action clearly enough to board and animate. On most projects the voiceover is recorded from the script before any drawing starts.
Lead with a strong hook in the first few seconds, keep one core message, and end with a single call to action. Match the length to where it will run. A promotional video runs shorter than a broadcast spot, and every second of silence or visual counts against your word budget.
AI can speed up an outline or a first draft. It still needs a human pass for voice, accuracy, and timing. Read anything it produces out loud and cut hard before you shoot.
A video script is a plan for what viewers hear and see, in order, and it is the cheapest place to fix a video. Nail down the audience and the goal first. Pick the format that fits the job, write for the ear, and add the visuals beside the words. Then read it out loud, time it, and lock it before anyone shoots or animates a frame. After 300-plus scripts, that is still the process I trust.
Larry is the Founder and Producer at Mammothic Films, founded in 2009. He leads a team that creates story-driven videos for businesses and brands. A marketer, designer, and animator, Larry has overseen the production of more than 1,500 videos for Fujitsu, Logitech, Warner Brothers, NFL, Blockdaemon, Astronomer, Canonical, and 370+ other companies in Canada, USA, and around the world.
Larry is a devout christian who believes in placing God first, and that everything else is a blessing that comes from serving Him. When he’s not working, he’s hanging out with his wife and three kids, playing the piano, or sight-seeing.
