What you get

  • A 1920x1080, 30 fps MP4 of 16 to 20 seconds
  • An editable HyperFrames project with the device, scroll and reveal
  • A beat sheet with copy budgets for the action, claim and CTA
  • A handoff note on screen source, data scrub and device rights
  • Optional PopNow music, voiceover and studio plate, mixed for social

Good for

  • Launching an app or web product that has a real screen
  • Showing one convincing action instead of a feature tour
  • Launch clips for social feeds and landing pages

Not for

  • Products with no interface, or screens that only exist as mock-ups
  • Step-by-step tutorials with several actions
  • Showing a recognisable branded phone or someone else's app

How to use it

In Claude or ChatGPT

Use the PopNow skill "SaaS Product Video with Device Mockups" (device-mockup-launch-video) for this: Launch film for our notes app from ./capture.png: type 'What did we decide about the launch?', show the answer, claim 'Every note, one question away.'

In a coding agent

/popnow-device-mockup-launch-video Launch film for our notes app from ./capture.png: type 'What did we decide about the launch?', show the answer, claim 'Every note, one question away.'

Then try

Shorten the scroll and keep the action beat as is.

Add a music bed and a voice for the claim. What will it cost?

What to have ready

  • A tall capture of your own product screen, scrubbed of real data
  • The one action, the answer it gives, a one-line claim and a short CTA
  • Your lockup and two colours; Node.js 22+ and FFmpeg

Similar skills

View all →

Questions

Do I need a 3D model of a phone?

No. The device is a neutral slab built from CSS perspective with a bezel, a moving glass highlight and a separate shadow. It deliberately looks like no particular manufacturer's product.

Can PopNow generate the app screen for me?

No, on purpose. The film's whole argument is that the screen is real, so it uses a capture of your own product. PopNow only adds sound and, optionally, an empty studio backdrop.

Why only one action?

One action keeps it a launch film. Two or more turn it into a tutorial, which needs a different length, pace and audience.

What about real customer data in my screenshots?

The checklist makes the agent scrub names, avatars, balances, messages and third-party marks before the screen goes on the device, and record what was removed.