Skip to content

Guides

How to record a product demo video on Windows

Plan one job per video, script it in sentences, clean the screen and record in sections, then cut, zoom, caption and export. The method I learned one bad take at a time, with Focal as the example.

By Updated 8 min read

To record a product demo video on Windows, pick the one job the video shows, write what you will say as full sentences, clean up the screen, and record it in one take with a marker at each section. Then cut the mistakes, add zooms so every click is readable, add captions, and export in the shape the place you post it wants. Most of that works with any recorder. The worked example here is Focal, a screen recorder for Windows, which does the zooms, captions and chapters for you after you stop.

My name is Vince, and I make Focal. The first audience for every demo I record is me, watching it back at my desk, and I am a tough room. Here is what that room has taught me.

Plan one job per video

A demo works when it shows one job from start to finish: invite a teammate, connect a payment provider, send the first invoice. Pick the job a new user tries first, or the feature you are announcing, and leave the rest for another video. I know the pull to show everything. It is your product, and all of it was hard to build. A tour of every screen still only proves that your product has screens, which nobody doubted. One finished job tells the viewer what it is for.

Then write one line: who is watching, and what they should do when it ends. That line decides what stays and what goes.

How long a feature demo should be

For one feature, 60 to 90 seconds is enough. People speak at roughly 150 words a minute, so that is a script of 150 to 225 words. A whole product can take three to five minutes. Past that, split it into chapters or several short videos. Nobody has ever reached the end of a demo wishing for a sixth minute.

Write a script of sentences, not bullet points

Bullet points make you improvise, and improvising is where “so, um, what I’m going to do here is” comes from. I have said that exact sentence, on camera, about software I wrote myself. Write the sentences you will say, one per action, in the order you click. Then read them aloud with the product open and rewrite any you trip on.

Say what happens, not what the control is called: “Pick a date range, and the chart redraws” beats “This is the date range selector.” Open with the problem or the result, show the steps, and end on what the viewer now has. Skip “Hi, in this video I’ll show you”. The title already said it.

Prepare the screen before you record

I go through this list before every take, because the one notification I forget to silence always arrives during the good one.

  • Turn notifications off. On Windows 11, switch on Do not disturb in Settings > System > Notifications. On Windows 10 it is called Focus assist. Close chat and mail apps too.
  • Clean the desktop and the browser. Hide the desktop icons, close the tabs you do not need, and use a browser profile with no bookmarks bar, no extension icons and no saved form entries. Autocomplete is how a personal email address ends up in a launch video.
  • Make the text larger. Zoom the browser to 125% or 150%, or raise your app’s font size. Text that is comfortable on a 27-inch monitor is hard to read on a phone.
  • Use test data that is not embarrassing. Believable names, sensible numbers and a few weeks of history look like a product people use. “asdf” and a real customer’s name do not.
  • Size the window sensibly. About 1600 x 900 to 1920 x 1080 has the shape of a 16:9 video and keeps the interface readable. A maximized browser on a 4K monitor makes everything tiny.
  • Get to the starting screen. Sign in and load each page once so nothing is slow during the take.
  • Check the microphone. Record a sentence and listen back.

Record it with Focal

Record the whole demo in one go, section after section, with a marker at the start of each. Later you can jump straight to any part instead of scrubbing back and forth like a safecracker.

  1. Press Print Screen. If another program owns that key, Focal uses Ctrl+PrtSc and tells you so. You can rebind it in Settings.
  2. Click your app’s window and the recording snaps to it. A window recording can leave out the app’s own title bar, and in Chrome, Edge and other Chromium browsers everything but the page, which suits a web app. Focal remembers the choice per app. On Windows 10, Windows draws a yellow border around a recorded window that only Windows 11 can hide.
  3. Choose the audio. Turn on the microphone, and system audio only if your product makes a sound worth hearing. On Windows 11, system audio can record only one app or leave one out, so a chat ping stays out of the demo. Focal keeps these choices for next time. There is more in the guide to recording system audio and your microphone.
  4. Turn on the camera before you start, if it helps. A live bubble shows you the camera and never appears in the recording.
  5. If the demo uses keyboard shortcuts, set keystrokes to Shortcuts, and combinations like Ctrl+K show as a pill on the video. Avoid Everything if you will type a password: the input log then keeps all typing, and Focal warns before such a project is exported or shared.
  6. After the countdown, a small control bar shows the time, with pause and stop. Press Ctrl+Alt+M at the start of each section. Every marker becomes a chapter.
  7. Press stop at the end of the script.

When you fumble a line

You will. Everyone does, and the people whose demos look effortless are mostly good at cutting. If you lose your place, click pause on the control bar, find your spot in the script and resume. It stays one recording, and clicks on the bar never cause a zoom. If you trip over a word, keep going: stop, wait a second, and say the whole sentence again. The silence and the repeat are easy to spot in the transcript and quick to cut. Start over only when the product itself goes wrong in a way you cannot cut around.

Edit the recording

When you stop, the editor opens with zooms already made from your clicks and bursts of typing. The camera eases in and out like a spring, waits for each click rather than drifting to the next, and joins nearby zooms into one that pans. Edits save as you work. This is the order I go in, which I settled on after trying most of the others.

Make captions first, because they are also your transcript. Focal makes them on your computer with a speech model that runs there (about 190 MB, downloaded the first time), and it reads the words on screen as hints, so your product’s name comes out right. Correct any line that is still wrong. The guide to adding captions to a screen recording covers the looks and the timing.

Then cut by the transcript. Select the words of a fumbled sentence in the captions list and cut them from the video. Cut every “um” and “uh” in one click, then tighten long pauses. Pauses where you click or type are kept.

Trim the start and end on the timeline, and split out what the transcript cannot reach, like a page that took five seconds to load. If the viewer needs to see it happen, speed the clip up instead. Your voice keeps its normal pitch.

Now check the zooms. Delete any that land on nothing. Drag on the zoom lane or press Z to add one, and press K to add a keyframe that moves the focus or changes the level. Speed is Slow, Smooth or Snappy, for all zooms or one, and zooms you edited stay as you left them if you change the automatic settings later. The guide to zoom effects in a screen recording goes deeper, and the zoom section of the home page sums it up.

Place the camera once you know where the action is: a corner the demo does not use, at any size, as a circle, a rounded square, or a wide or tall rectangle, with an optional border and shadow.

Last, turn on voice cleanup in the Audio tab. It removes background noise and evens out loudness to the level web video is mixed at. Background music added under the recording gets quieter while you talk.

Make it look like your product

Raw screen footage looks like a screen, which is accurate in the way a passport photo is accurate. A background, padding and a soft shadow make it look like a product shot, and a style saves all of it so every demo matches the last.

A style in Focal stores the background, padding, corners, shadow, cursor, zoom feel, camera bubble, captions and more under a name. Give it to any recording in one step, make it the default for new ones, or export it as a file with its images inside. The background can be one of twelve ready-made ones, a gradient of up to five of your brand colors, an image blurred or whole, or the recording itself blurred.

An optional macOS or Windows title bar frames the recording like an app window, which suits a page recorded without its browser. Your logo can sit in a corner of every frame. Intro and end title cards name the video and give a next step, such as your web address, and chapter titles can appear on screen. The styles section has the full list.

Export and post it

Focal exports MP4 in H.264 or HEVC at Original, 1080p or 4K, up to 60 fps, or a GIF. H.264 plays everywhere, which makes it the dull, correct choice: pick it unless you know HEVC is accepted where you post. Presets cover YouTube at 1920 x 1080, Shorts and Reels, Instagram posts, HEVC and GIFs, plus your own. The export section has the details.

For YouTube, use the YouTube preset. Your markers come out as chapters in the file, and a click copies them as YouTube timestamps for the description. YouTube shows chapters when the list starts at 0:00 and has at least three entries, so check the first line before you publish.

Captions can go out as a track viewers switch on, drawn into the picture, or both, and on YouTube you can upload a saved .srt with the video. Drawn captions suit social feeds, where many people watch with the sound off.

For Shorts, Reels or TikTok, change the aspect ratio to 9:16 and Focal re-aims the zooms and captions. Use Save a copy first, so the 16:9 version stays, then export the copy with the Shorts and Reels preset. The guide to turning a screen recording into a vertical video has more, and 1:1 and 4:5 work the same way.

Focal makes a file and stops there: no share link, no hosting. The video goes wherever you post it. Exporting that file needs a license. Recording and editing are free, with every feature, no time limit and no card or account, so you can download Focal, make your demo and watch it in the preview before you buy. Until you enter a key, the preview shows a “Made with Focal” watermark.

Questions

How long should a product demo video be?

For one feature, 60 to 90 seconds, which is a script of about 150 to 225 spoken words. A walkthrough of a whole product can run three to five minutes, and anything longer works better as chapters or as several short videos.

Do I need a camera in a product demo?

No. A camera bubble helps when the viewer is deciding whether to trust you, as in a founder's launch video or a sales demo, and adds little to a how-to. In Focal the camera has to be switched on before you record, and its size, shape and position can be changed afterwards.

Do I need a script for a demo video?

Yes, written as the sentences you will say rather than bullet points, one sentence per action in the order you click. Bullet points make you improvise, and improvising is where most filler words and restarts come from.

How do I hide notifications while recording on Windows?

On Windows 11, turn on Do not disturb in Settings > System > Notifications. On Windows 10 the same idea is called Focus assist. Close chat and mail apps as well before you start, so nothing pings in the middle of a good take.

What resolution should I record a demo at?

Record a window of about 1600 x 900 to 1920 x 1080 with the text enlarged, and export at 1080p for most places. Focal exports up to 4K at 60 fps, even from a smaller recording, whose text and captions are then drawn at 4K while the recording itself is scaled up; 4K is worth it when the recording was made at that size.

Can I use one recording for several video shapes?

Yes. In Focal one recording can be exported as 16:9, 9:16, 1:1, 4:5, 21:9 or its own shape, with the zooms and captions re-aimed for each, so the YouTube video and the Short come from the same take.

Read next