A product launch video works when it is built around one idea. Open on the viewer's world, name the product once they care, show it doing three to five real things, end on what changes for them, and cut the hero film into shorter versions for every channel.

A product launch video has one job: make a product nobody has used yet understood, and wanted, before the viewer scrolls away. Everything in this blueprint serves that job.

It comes from three launch films we made at Brand Appart: Promethee (2026, 1 min 45), RockFi Intelligence (2026, 53 seconds) and the InCard rebrand (2024, 1 min 24). The timecodes below refer to those films, which are embedded in this article.

What a product launch video has to do

A launch video plays to three audiences on the same day. The first scrolls past it on X or LinkedIn, often with the sound off. The second lands on your homepage and gives it a few seconds before deciding to stay. The third is an investor, a partner or a future hire who gets the link forwarded and watches it to judge the company.

One film can serve all three only if it is built around a single idea: the one sentence you want each of them to repeat after watching. Write that sentence before anything else. If you need two sentences, you have two films.

The blueprint: five beats

The five beats below hold whatever the style of the film. Here they are with the Promethee timecodes.

Promethee launch video (2026), by Brand Appart. Hook at 0:00, name at 0:10, product in action from 0:30, payoff at 1:24.

1. The hook: the viewer's world, in under ten seconds

Open on something the viewer already knows, not on your logo. Promethee opens with a question to people who play games: "You've probably spent hours leveling up a character. What if this time that character was you?" InCard opens on the people it serves: e-commerce businesses that want to grow. In both films, the first seconds are about the viewer, and the product has not been named yet.

A useful test: if the first frame and the first line would work just as well in a competitor's film, the hook is too generic.

2. The name: once they care, and not much later

Name the product as soon as the viewer knows why they should listen. Promethee names itself at ten seconds ("Welcome to Promethee"). InCard does it at 21 seconds ("Meet InCard"), after a short opening about its audience. Wait much longer and the viewers who are still watching start wondering what they are looking at.

3. The product in action

This is the middle of the film and its longest part. Show the product doing three to five real things, one action per shot. In the Promethee film, from 0:30 to 1:20, the voiceover walks through one action at a time: a session starts, the user enters deep focus, climbs the leaderboard, receives a quest, earns XP, shares the session.

Real product states beat abstract animation. If the interface is not ready to be filmed, redesign the screens for video first: larger type, fewer elements, real content. For Promethee, we rebuilt the design system from the product's codebase before producing the film.

4. The payoff: what changes for the viewer

End on what the features add up to, not on one more feature. Promethee's last line turns a focus tracker into a community: "Then you look up and realize that you're not playing alone."

5. The end card

Close on the product name, the address and one next step: try it, join the waitlist, read the announcement. Hold it long enough to be read, because viewers often pause on it.

How long should a product launch video be?

Long enough to show the product doing what matters, and no longer. Our three films run 53 seconds, 1 minute 24 and 1 minute 45, and the difference comes from how much the viewer needs to see.

Film	Length	Voiceover	Structure
Promethee (2026)	1:45	Most of the film	Hook, name, product walkthrough, payoff
InCard rebrand (2024)	1:24	First 25 seconds	Audience, promise, name, new identity to music
RockFi Intelligence (2026)	0:53	None	Picture and music only

A new kind of product, like Promethee's mix of focus tracker and game, needs time to be shown working. A rebrand like InCard's needs less explanation and more atmosphere. A product for a specialist audience, like RockFi Intelligence for private bankers, can lean on picture alone when the viewer already knows the job it does. Past two minutes, the hero film starts competing with a product demo, and that is a different video.

Voiceover, on-screen text or music only

The three films make three different choices, and each one fits its product.

RockFi Intelligence launch video (2026), by Brand Appart. No voiceover: picture and music carry the film.

Voiceover across most of the film (Promethee): for a product that needs explaining, and a story with a character the viewer can step into.
Voiceover, then music (InCard): a few lines set up the audience and the promise, then the visuals take over. It suits a rebrand, where the change is something you see.
Music only (RockFi Intelligence): suits an audience that already knows the problem, and a restrained brand register. It also plays the same with the sound off.

InCard rebrand launch video (2024), by Brand Appart. About 25 seconds of voiceover, then the new identity to music.

Whatever you choose, put the key lines on screen as text or captions. In a feed, many viewers decide whether to keep watching before they turn the sound on.

The launch video script template

Use this as the first draft of your script. The timings are for a film of about 90 seconds; scale them to your length.

0:00 to 0:05, the hook. One line about the viewer's world: "You've probably…", "Every [audience] knows…".
0:05 to 0:10, the turn. The question or tension your product answers: "What if…".
0:10 to 0:20, the name. The product name and a one-line promise.
0:20 to 1:10, the product in action. Three to five lines, each one a verb the product performs on screen.
1:10 to 1:25, the payoff. What the viewer's work or life looks like after.
1:25 to 1:30, the end card. Name, address, one call to action.

Read the script aloud against a timer before anything is designed. A line that takes longer to say than its shot lasts will be cut later, when cutting costs far more.

Choose how to produce it, shot by shot

A launch video does not have to use one technique. It can mix three: product UI animation for anything the viewer must read, motion design for the brand system (type, logo, transitions), and generative AI for shots that would be expensive or impossible to film, like environments, atmospheres and visual metaphors.

The Promethee and RockFi Intelligence films were made with generative AI in the pipeline. We still design interfaces and brand elements by hand, because they have to be exact. A director makes the choice shot by shot, which is how our AI creative studio works.

Cut it for every channel

The hero film is the source, not the only deliverable. Plan the cutdowns in the script, not after the final cut.

Homepage: the full film, with a first frame that works as a poster.
X and LinkedIn: a shorter cut that reaches the product faster, with captions.
Product Hunt: the hero film or a one-minute cut, alongside the screenshots.
Vertical formats: 9:16 versions for Reels, Shorts and TikTok, reframed rather than cropped.
Ads: 6 to 15-second edits built around a single product moment.
Investor update: the hero film, with a line on what launched and what comes next.

Launch video checklist

One idea, written as one sentence.
A hook about the viewer in the first ten seconds.
The product named in the first quarter of the film.
Three to five product actions, one per shot.
Interface shots designed for video.
A payoff line about the viewer, not the product.
An end card with one call to action.
On-screen text or captions for silent autoplay.
Cutdowns planned for each channel.

Mistakes that sink launch videos

Opening on the logo. The viewer does not know the brand yet, so the logo means nothing in the first second.
Reading the feature list. Features are a list; a launch video is a story about one change.
Unreadable interface shots. Desktop screenshots shrunk into a phone feed cannot be read.
No product at all. A mood film for a product nobody has seen leaves viewers guessing.
One length for every channel. The homepage film is too long for an ad and too slow for a feed.
No next step. The film ends and the viewer does not know where to go.

If you want this blueprint applied to your own launch, our launch video team writes, designs and produces the film and its cutdowns.

How long should a product launch video be?

Under two minutes for the hero film. Our launch films run between 53 seconds and 1 minute 45, depending on how much of the product the viewer needs to see. Cutdowns for social and ads are shorter, from a few seconds to about a minute.

What should a product launch video include?

A hook about the viewer, the product name early, the product doing three to five real things, a payoff about what changes for the viewer, and an end card with one call to action.

Does a launch video need a voiceover?

No. A voiceover helps when the product needs explaining. A film can also work with a short voiceover followed by music, or with picture and music only, as long as on-screen text carries the key lines for people watching without sound.

Can you make a product launch video with AI?

Yes, in part. Generative AI is useful for environments, atmospheres and shots that would be costly to film. Interfaces and brand elements still need to be designed precisely. Two of the three launch films in this article were made with generative AI in the pipeline.

How do I write a launch video script?

Start from the one sentence the viewer should remember, then write the five beats: hook, name, product in action, payoff, end card. Read it aloud against a timer before anything is designed. The template above gives timings for a 90-second film.

Good to know

Many people will first see your launch video muted, in a feed. Design the first frame and the on-screen text for that viewer before you polish the soundtrack.