Short answer: around 70 seconds is best for a product launch, 40-45 seconds for a feature launch. Most launch videos you see run 60-90 seconds. A story-led or filmed video can go up to about 120, but even then your time is limited, so how you split it matters more than the total.
Why shorter wins
The longer a video gets, the more boring it becomes. That includes 90 seconds, which I used to think was fine.
A launch video has one job: make people understand the product and want it. Every extra second is a second where they can scroll away. If a part doesn't move the story forward, it's costing you viewers.
There's no real industry standard. But most of what you see runs 60-90 seconds, and we'll make a 50-second or a 95-second video when it's right. If you ask what's best, it's 60-70.
What makes a video slow
It's almost always the same idea, repeated.
Say your product solves ten problems. You want people to relate, so the script lists them: you've got this on your plate, you've got that problem, you're doing this but missing that. Each line is roughly the same idea. That section starts to feel slow, and people turn the video off. The same thing happens with features.
The second cause is tempo. A one-sentence hook, two lines of problem and ten lines of features is bad planning. The parts need balance. On a recent script, one new beat added about 23 seconds of problems before anything good happened. We pushed back, because it would sound boring later.
The third is a line that comes back. The same sentence at the start, in the middle and again at the end. Cut it, or rewrite it.
Product launch, feature launch, or a story
- Product launch: around 60-70 seconds. Enough to set up the problem, turn to the answer, show two or three concrete things the product does, and close.
- Feature launch: 40-45 seconds. People already know who you are. Skip the setup and show the new thing.
- A story or a filmed video: up to about 120 seconds. A strong story or a scenario holds attention the way a short film does. But even the longest version is limited, so be careful how you split it.
How the time splits
Every launch video we make has the same parts. A shorter video keeps all of them, each one smaller.
- Hook. Grab people in the first 3 seconds. People and the problem, felt, not explained.
- Turn. One or two lines that change the mood from problem to answer.
- Product. The longest part. Two or three concrete things it does, not a feature list.
- Proof. One number, or logos. Short.
- Close. The last line people remember, then an end card with logo and URL, no voiceover, 3-5 seconds.
The product part should tell one story. Follow one user, or one agent, from start to finish, and the viewer remembers it.
For a 90-second video, roughly: hook and problem 0:00-0:15, turn to 0:25, product to 1:05, proof to 1:15, close to 1:30. A 60-70 second video keeps the same order, each part shorter.
Count words, not seconds
Length is decided in the script, long before anything moves. The rule we use: about 2.3 words per second of voiceover.
But the voice never fills the whole video. Pauses, music-only moments and the end card take 10-15 seconds. So a 90-second video holds about 170-190 words, not 207. For a 60-70 second video, that's roughly 105-135 words.
Our Stacks launch video ran at 2.35 words a second, because the client wanted it fast. Most videos should aim lower, so the words have room to land. A video packed wall to wall with voice feels like an ad. Leave a short pause between parts and a moment with only music.
Two more checks: one voiceover line runs about 3-8 seconds, and if it runs longer, it's two lines. And keep one idea per line, one number per beat at most.
No voiceover? Keep it under about 70 seconds. In our experience, type-only videos work best short. Reading is slower than listening.
Where it plays changes the length
- Website, social posts, email: the main launch video. The website hero doesn't need its own length if you're happy with the main one.
- Teaser cuts for LinkedIn and X: 10-15 seconds. Think of a movie trailer. For Stacks, the main video was built around the new Stacks Workspace. The cut drops the general hook and problem, keeps one line to set it up, shows the exciting part, and closes on the logo.
- A booth screen at an event: 45-70 seconds, built to work with no sound. More in our conference booth video guide.
- Product Hunt: use your main launch video. More in our Product Hunt video guide.
Short-form content made to go viral is a different job from a launch video. It lives or dies on the first 3 seconds and each platform's format, and it needs its own scope.
Agree the length before the script
Videos grow. Someone adds a line, then a feature, then a customer quote. On two launch videos in a row, ours got longer than we planned. So now length is in the proposal.
By default, we plan for about 60 seconds. If you already have a length in mind, we start from that. When a client is rebranding an existing video, we usually keep its length. If the video grows, we talk about it. Not only the price: does the extra time make the story better, or is it just more seconds? If it's just more seconds, we'd rather cut. More on how runtime affects price in how much a launch video costs.
What to agree on before writing starts:
- Type: product launch, feature launch, or a story-led video.
- Target length: written down.
- Must-say list: the 2-3 things the video has to land. Everything else is a candidate to cut.
- Where it plays: feed, website, booth screen, stage. It changes the brief.
Proof: Corti's launch video - "It ran the conference circuit, and held up for a public launch later." - Ryan Ackell, Growth Lead at Corti
Got a launch coming? Tell us the date and we'll plan the length, the script and the timeline with you.