Screenshotify
· ASO · 5 min read

App preview video vs screenshots — which converts?

TL;DR. Video wins where motion sells (games, fitness, video editors, music creation). Screenshots win where the install decision happens in 2 seconds off the first panel (banking, weather, notes, utilities). For most apps the right answer is one short video on top plus a strong static carousel — not video instead of screenshots.

Apple added App Preview videos in 2014 and most apps ignored them until ~2020. By 2026 the feature has earned its weight: the autoplay poster sits at the top of the App Store product page, larger than it used to, and a 30-second muted clip plays on first scroll. For some categories the video has become the single highest-conversion creative asset.

For other categories it does nothing.

Where video wins

Categories where motion is the product or close to it:

In these categories, a strong 15-30 second video can lift tap-to-install by 20-40% over screenshots-only setups (consistent finding across StoreMaven, AppTweak, Mobile Action benchmarks 2023-2026).

Where video loses (or doesn't help)

Categories where the install decision is fast and static:

In these categories, the time spent producing a 30-second video would have been better spent on a stronger static carousel.

Where it's category-dependent

Mixed verdict:

If you do ship a video

Apple's video specs (covered in detail on the video requirements guide) — 15-30 seconds, H.264, per-device-class dimensions, no marketing voiceover, no external URLs. The constraints are tighter than YouTube/Instagram ad video.

Three rules that hold up:

The realistic plan for most apps

  1. Ship strong static screenshots first. Carousel does most of the lifting.
  2. Add a video if your category benefits from it. Skip the video if it doesn't.
  3. Localize the carousel before localizing the video. Translated captions on static screenshots are cheap and high-impact. Localized videos are expensive and less impactful.
  4. A/B test video vs no-video if you have install volume for significance — categories vary.

What to read next

Open the editor →