Teaser video for app launch: fifteen seconds, sound off

The teaser is the only asset most people will ever see of your launch, and it is usually cut last, from a film built for a page where the sound was on.

August 24, 2026 · 8 min read

Two weeks before the launch, someone asks for the social cut. The 90 second film is finished, approved and beautiful, so the editor trims it to fifteen seconds. The result opens on a logo, the voice-over is gone so the first line of the story is missing, and the product appears at second nine. In a feed, second nine does not exist.

This is the standard order of operations and it is backwards. The launch film gets watched by people who already decided to care: press, investors, your own team, visitors who came to the site on purpose. The teaser is what everyone else gets. It runs in a place you do not control, next to content that is not yours, with the sound off and a thumb already moving.

So the teaser is not a shorter version of the film. It is a different asset with different rules, and it deserves to be designed rather than salvaged.

The teaser has one job: buy three more seconds

Every other format assumes a decision has already been made. A landing page video plays because someone arrived. An explainer plays because someone pressed play. A teaser plays because an algorithm put it in front of a stranger who was looking for something else.

That single difference sets everything. The first frame cannot be abstract, because abstraction reads as an ad. The claim cannot be spoken, because nothing is spoken. The pacing cannot build, because there is nothing to build on. What the teaser has to do in its first second is answer the only question the viewer is actually asking: is this worth three more seconds.

The placement writes the spec

App launches spread the same asset across four or five surfaces, each with its own crop, its own length ceiling and its own audience posture. One build should cover all of them, but only if the ratios were planned before anything was animated.

The row people forget is the app store preview. It is governed by the platform, it gets reviewed, and it is one of the few places where a viewer is already deciding whether to install. A teaser recut for that slot converts better than anything else you will make, and almost nobody briefs it.

What actually goes in the fifteen seconds

Fifteen seconds is roughly four beats. Not four scenes, four beats: things the viewer registers. Spending one of them on a logo animation costs you a quarter of the asset before anything has been said.

1. 01Seconds 0 to 2, the product on screen. Real interface, moving, doing the thing the app is for. Not a phone floating in gradient. Not a hand holding a device. The screen itself, large enough to read on a phone.
2. 02Seconds 2 to 6, the claim, written. One sentence a viewer could repeat to a colleague. On screen, in type big enough to survive a small crop, held long enough to read twice.
3. 03Seconds 6 to 12, the proof. The part of the flow that is not obvious. If the app removes three steps, show the three steps disappearing. This is the only stretch where craft can show off, and it is where most teasers go quiet instead.
4. 04Seconds 12 to 15, the name and the action. Name legible, spelled the way someone would type it, plus one instruction. A URL nobody can read at 9:16 is not a call to action.

Notice what is not on the list: a tagline before the product, a founder speaking, a countdown, a date card that will be wrong in a week. Each of those is a beat spent on something the viewer did not ask for.

Cut the teaser first, not last

The cheapest change you can make to a launch is the order. Build the visual system first: type, motion behavior, color, how the interface is framed, how numbers and states appear on screen. Then cut the teaser from that system, before the long film exists. If the fifteen second version does not hold, the ninety second version was never going to, and you have found that out while it is still cheap to fix.

Working in that order also means the launch film, the store preview and the paid variants are recuts of one build rather than three productions with three different looks. That is the same sequencing behind the four decisions in our post on the product launch video, and it is how our Launch Sprint runs: narrative and visual micro system in the first two days, teaser next, everything else after.

There is also the case where the launch is on Thursday and there is no system to build from. We should be transparent here, because the tool is ours: ShipTeaser was built by the same founder as this studio. It takes a product URL and returns a 15 second 1080p motion graphics teaser made for muted feeds, with an optional music bed and the first one free without a card. What it does not do is give you a narrative spine or a visual system the rest of the launch can inherit, and there is no voice-over, no screen recording and no timeline to edit. For a feed cut you needed yesterday, that trade is usually worth taking. For the launch you have been building toward for a year, it is not the whole job.

The five failures that kill a launch teaser

These come up on almost every project that arrives with an existing cut. None of them are craft problems. They are all decisions made before anyone opened a project file.

• Cut down from the long film. The structure was built for a viewer who chose to watch. Trimming it removes the setup and keeps the pacing, which is the worst of both.
• Type sized for a desktop review. It was approved on a large monitor and it ships to a phone at half the size, inside a crop nobody checked.
• The claim carried by the voice-over. With the sound off, the film says nothing. Most feed views are muted, including the ones that convert.
• Safe zones ignored. The bottom fifth of a vertical feed post is covered by captions, handles and buttons. Anything placed there is gone.
• One master, no variants. A single 16:9 file gets letterboxed everywhere, which reads instantly as an asset that was not made for the platform it is on.

The fix for all five is the same and it is not a bigger budget. It is deciding the surfaces before the storyboard, so the frame is designed for the crop it will live in.

The brief that gets you a usable first cut

Before you talk to a studio, a freelancer or your own designer, write these five lines. They take twenty minutes and they settle the arguments that would otherwise happen during revisions, when changes cost real money.

• The surfaces, named, with the ratio and length ceiling for each one.
• One claim, written as a single checkable sentence, the one that will sit on screen.
• The moment of proof: the exact part of the product you want shown working, with a recording of it.
• The name treatment: how the brand appears at the end, and what the viewer is meant to do next.
• The date you can no longer change anything, which is usually earlier than the launch date.

A team that gets that page back delivers something close on the first pass. A team that gets a deck and a launch date delivers a beautiful film in a format nobody can use, two weeks before you need it, and the recut comes out of your budget rather than theirs. If you want to see how this plays out across a full engagement rather than one asset, our brand video for tech startups post covers the same ground at the level of the whole brand.