A product launch video script names the product and what it does in the first eight seconds, shows the specific old way it replaces, demonstrates one job end to end, and closes with a single next step plus availability. For a 60-second launch video that is roughly 150 to 160 spoken words across four beats. The discipline that makes it work is subtraction: one job proven properly beats six mentioned quickly, every time.
Launch videos fail in two predictable ways. The first is the mood film, sixty seconds of slow motion and swelling strings that never says what the product is. The second is the feature tour, a breathless list of everything the thing can do, which leaves the viewer with a vague sense of capability and no idea what it is for. Both come from the same root cause: the team knows the product too well to remember that the viewer is starting from zero. This guide covers the four-beat structure, the beat everyone skips, how to pick the one job to prove, the launch cut kit, and a finished example you can copy.
What is the best structure for a product launch video?
Four beats, arranged so the viewer's questions get answered in the order they actually ask them: what is it, why does it exist, does it work, what do I do now. That order is not negotiable, because a viewer who does not know what the product is cannot evaluate a demo of it.
The change (0-8s)
Say the product name and what it does, plainly, in one sentence. "This is Setter. It prints a label in two seconds without a phone." No slow build, no logo animation eating the opening. If your first line could describe three other products, it is a tagline rather than a description, and it is costing you the most valuable seconds of the video.
The before (8-22s)
Show the specific old way. This is the beat that gets cut for time and it is almost always the wrong cut, because "new" only means something against a "before". Make it concrete and slightly uncomfortable: the handwritten label peeling off, the six taps through an app, the spreadsheet with the merged cells. One recognisable moment of the old way does more persuading than a paragraph of benefit copy.
The proof (22-48s)
Demonstrate one job, start to finish, in real time if you can. The temptation is to cover the range. Resist it. A launch video that shows one task completed convincingly makes the viewer believe the rest; a launch video that shows six tasks halfway makes them believe none of it. If a feature is not part of the one job, it belongs on the product page, not in this video.
The next step (48-60s)
One action, plus the two facts people need to act: when they can get it and what it costs. Launch videos lose an astonishing number of interested viewers by ending on a logo with no availability. If it ships in six weeks, say so. If there is a waitlist, that is the action.
Why does the product video carry so much of a launch?
Because it is doing work that used to be done in a room with a person. Buyers evaluate before they ever speak to anyone, which means the launch video is frequently the whole first meeting, and the questions it fails to answer mostly go unanswered.
That capability has also moved in-house, which changes what a launch video can reasonably cost. Wistia's 2026 report found 54% of companies now have an in-house video team, up from 36% two years earlier. A launch film no longer has to mean an agency and a five-figure invoice, which in practice means you can afford to re-cut it when the positioning shifts.
How do you choose the one job to prove?
Pick the job that is both the most common and the most visibly changed. Most common, because it is what the majority of buyers will actually do with the product. Most visibly changed, because a demo only lands when the viewer can see the difference without being told about it. Where those two overlap is your proof beat.
"It syncs, it exports, it has six templates, it works offline, it connects to your other tools, and it is available in four colors."
One unbroken shot: type the word, press once, peel the label, stick it on the jar. Done in eleven seconds, no cuts.
The unbroken shot matters more than it sounds. Cutting away mid-task reads, correctly, as hiding something. If the job genuinely takes ninety seconds, show it compressed but say so on screen, because an honest speed-up costs you nothing and a concealed one costs you the trust the whole video was built to earn.
Should the founder appear in the launch video?
Only if they are doing a job the script needs done. A founder is the right choice for the "before" beat, because the story of why the product exists is genuinely theirs to tell and it is hard to fake. A founder is usually the wrong choice for the proof beat, where a pair of hands and the product on a table is more convincing than a person describing what you are watching.
The failure mode is the founder-thank-you launch video, four minutes of gratitude to the team with the product appearing at the end. That video has a real purpose, but it is an internal and investor artifact, not the thing you put on the launch page. Cut it separately.
What does a product launch video script look like (example)?
Here is a prompt you can paste into PlanThatVideo, and the timed script it turns that paragraph into. The product is a cordless label printer, chosen because the "before" is instantly recognisable and the one job takes seconds to prove.
And the script it produces, beat by beat with timecode, voiceover, and on-screen action:
Count it: about 155 words, roughly 58 seconds at a dry, unhurried pace. The product is named and explained by second five, the "before" is two specific images rather than a claim, the proof is a single shot with a visible timer, and the close carries a price, a date, and one action. From the same prompt, PlanThatVideo also generates the matching shot list:
Four setups, one afternoon, one table. If you want the mapping from script lines to shot rows in detail, see what is a shot list.
What else should one launch shoot produce?
Four cuts, planned before the shoot rather than salvaged after it. The beats above are deliberately separable, which is what makes this possible at no extra shoot cost:
- The hero cut (60s). All four beats. This lives on the launch page and the press kit.
- The demo cut (90-120s). Beat three extended, with two more jobs shown properly. This is for the product page and for anyone already interested.
- The social cut (15-20s). Beat one plus the unbroken proof shot, vertical, captions burned in. No "before", because in a feed the demo is the hook.
- The founder cut (60-90s). The "before" beat told as a story, to camera. This is the LinkedIn and email version, and it is the one where a person on screen genuinely helps.
Capturing all four costs perhaps forty extra minutes on the day and is impossible to add later. The full method is in how to repurpose video content.
Plan for the re-cut too, because your launch-day positioning is a hypothesis rather than a finding.
"I've launched a lot of products, almost 20 brand new products in the market. Not once did we correctly guess all those things." — April Dunford, positioning expert and author, How to Web
Which is the practical argument for separable beats. When you discover eight weeks in that the product is being bought by a different person for a different reason, a video built from four independent blocks needs one new beat, not a new shoot.
A reusable product launch video script template
Copy this and fill every line. If a beat runs over its word budget, cut it rather than speeding up the read.
FAQ
How long should a product launch video be?
Sixty seconds for the hero cut. That is long enough for all four beats at a natural pace and short enough that people finish it, which matters because the availability and price live at the end. Make a longer demo cut separately rather than stretching the hero video to fit more features.
Should the product name go in the first line?
Yes, along with what it does. A launch video is often the first time anyone has heard the name, so withholding it for a reveal wastes the attention you will never have again. Name it, describe it in plain words, then earn the rest of the minute.
What if the product is hard to demo?
Then demo the outcome instead of the mechanism. Software with no visible surface, services, and infrastructure products all have a before-and-after somewhere: a report that took a week now takes an hour, a screen with nine fields now has two. Find the thing a viewer can see change, and shoot that.
Do I need a different launch video for each channel?
One shoot, four cuts, as above. The structure holds across channels but the trim does not: a feed cut leads with the demo and drops the "before", while the launch page cut keeps all four beats. Plan the cuts in the shot list so you capture the vertical framing on the day.
How do I know if the launch video worked?
Decide the number before you shoot: waitlist signups, pre-orders, or trial starts within a set window, measured against a written baseline. Views will be unusually high on launch day and tell you almost nothing. The method is in how to measure video marketing ROI.
Launching something? Start from a script, not a blank page.
PlanThatVideo turns a one-paragraph product description into a timed, beat-by-beat launch script, plus the matching shot list and storyboard. Edit, export, shoot.
Try it free →