How to make a product demo video

A good product demo video shows one product doing one job, in about a minute, with a voice that says what the viewer is looking at. Here is the whole process — the decisions that matter with any tool, and how they map onto DemoScreen.

Before you start: decide what the video is for

Write one sentence that finishes “After watching this, the viewer should…”. Sign up, understand the new feature, know how to file an expense — each one produces a different video. A launch video sells the outcome; an onboarding video teaches the clicks; a feature-update clip shows the one thing that changed. If you cannot finish the sentence, the video will try to do everything and land nothing.

Then pick where it will play, because placement decides length and shape. A homepage hero, an App Store preview and a LinkedIn post are three different videos, even when they show the same product. The length guide has the numbers per placement.

Step 1 — Choose the screens

A demo is a sequence of screens, so the screens are the storyboard. For a one-minute video, 6 to 10 screenshots is the comfortable range: fewer and each one lingers, more and the viewer cannot read anything before it moves on. Close-ups on highlighted parts of a screen add scenes on top of that.

• Follow one path through the product. Start where a new user starts and end where they get the result. Skip settings, empty states and anything you would have to explain.
• Use realistic data. A dashboard full of “Lorem ipsum” and zeros looks unfinished. Fill it with plausible names and numbers before you capture.
• Capture at full resolution. Close notifications, hide bookmarks, use one browser width for every shot so the frame does not jump between scenes.
• Order them as they should play. The order of the screenshots is the order of the story.

Step 2 — Mark what the viewer should look at

A full-screen screenshot shown for five seconds is a wall of interface. The viewer needs to know where to look. Pick the one or two regions per screen that carry the point — the chart that went up, the button that does the work, the field that fills itself in — and zoom the camera onto them.

Close-ups matter even more on small screens. On a vertical 9:16 frame a whole desktop dashboard is unreadable; a close-up of one card is not.

Step 3 — Write the script

Write one line per screen, in the order the screens play, and say what the viewer is seeing and why it matters. Keep sentences short enough to say in one breath. Lead with the problem the product solves, show three things it does about it, end with one clear next step.

Budget words against time: spoken English runs about 2.5 words per second, so a 60-second video holds roughly 140–150 words — less if you want room to breathe. The script template gives a beat-by-beat layout you can copy.

Step 4 — Pick a voice and music

• Voice: choose one that suits the audience — calm and clear for training, a little brighter for a launch. Listen to a sample reading your own script, not a demo sentence.
• Music: a quiet bed under the voice, never over it. Pick by mood (calm, upbeat, tech) and keep it well below the narration.
• Loudness: platforms normalise audio. A mix levelled to a streaming loudness target (around −14 to −16 LUFS) sounds consistent next to everything else in the feed.

Step 5 — Choose length and format

Render the same screens in more than one shape rather than cropping one video later — a crop cuts off the parts the close-ups were aimed at.

Step 6 — Add captions and an end card

Most social video plays muted, so ship a caption file (SRT) with every video. Finish on an end card with the product name, one line on what it does and the address to go to — that is the frame people pause on.

Doing it in DemoScreen

DemoScreen turns the steps above into five screens in the studio: Screens → Script → Sound → Brand → Video. The tutorial below walks through all five on a real project.

1. Screens. Drop in your screenshots in order (PNG or JPG). Pick a look — Midnight, Keynote, Glass, Terminal or Paper — and a format: 16:9, 9:16 or 1:1. Drag on a screenshot to mark up to four highlights; each becomes its own close-up scene.
2. Script. AI reads the text on your screens and drafts one line per screen. Edit it, or type notes and rewrite. Choose a length from 0:30 to 3:00 (Free up to 1:00, Starter 1:30, Pro and Team 3:00) and the narration is written to fit.
3. Sound. Pick a voice and a music mood, with a sample for each. The mix is levelled to streaming loudness automatically.
4. Brand. Add your logo, accent colour, headline and website to the end card.
5. Video. Review every scene in the live preview, then make the video. A 60-second demo usually renders in about 10 minutes, longer if other videos are ahead of it in the queue; you get an email with a 1080p MP4, an SRT file, a GIF of the opening and a share link.

You can try the same edit loop on a sample project in the sandbox without an account, or see finished renders in every format in the gallery.

Where to publish it

• Landing page: autoplay muted above the fold, with captions burned in or a visible play button. Keep it under a minute.
• App Store / Google Play: vertical, 15–30 seconds on the App Store; on Google Play only the first 30 seconds autoplay, muted.
• LinkedIn and other feeds: square or vertical, captions on, the point in the first three seconds.
• Onboarding emails and help centre: landscape, a little longer, one task per video. Link to the share page instead of attaching a file.
• Sales and internal training: landscape, up to a few minutes, one workflow per video.