App Preview Videos That Convert: Patterns From 12 Categories

App preview videos are the single most over-debated and under-tested asset in the App Store. Teams agonize over them, ship one weak version, never test, and conclude "videos do not work." When the video is right, it lifts product page conversion 10 to 25%. When it is wrong, it drops conversion 5 to 15%. The difference is not budget. It is structure.

This article catalogs the video structures that win in 12 of the most common app categories in 2026, the production rules that hold across all of them, and the testing approach that tells you whether your video is helping or hurting.

The format constraints that decide everything

Before category-specific patterns, the constraints to design around:

30 seconds maximum on both App Store and Play
Autoplay with sound off by default — the video has to read with no audio
15-second mental cutoff — viewers decide to install or bounce inside the first 15 seconds
Portrait for iPhone, landscape for iPad, separate Play variants
No prices or app store interface in the video itself

The single most important constraint is the 15-second cutoff. Treat the video as two halves: the hook (seconds 0–7) and the proof (seconds 7–15). Everything after 15 is bonus.

The 12 category patterns

1. Subscription fitness (Strava, Peloton, Strong, Tonal)

Pattern: result first, then routine. Open on a transformed metric (a personal record, a graph trending up, an unlocked badge), then show one or two seconds of the in-app moment that produced it. The user is shopping for a result, not a UI.

What loses: Long pre-workout countdown screens. Stock footage of generic athletes.

2. Language learning (Duolingo, Babbel, Drops)

Pattern: progress and streak. Show a streak counter ticking up, a milestone unlock, then 2–3 seconds of the lesson loop. The category is about consistency, not depth — the video should convey "I will stick with this."

What loses: Quiz screens that look like school.

3. Productivity and task managers (Todoist, Notion, Things)

Pattern: chaos to order. Open on a busy state (a cluttered inbox, a chaotic list), show a fast transition to organized state, end on a clean screen. The product moment is the transformation.

What loses: Tutorial-style walkthroughs of every feature.

4. Note-taking and second-brain (Bear, Obsidian, Apple Notes)

Pattern: thought to artifact. Show the friction of capturing an idea elsewhere (browser tab, sticky note, scribbled paper), cut to capturing it in the app, end on the linked or organized artifact.

What loses: Showing the formatting toolbar. Users do not buy on formatting features.

5. Personal finance (YNAB, Copilot, Rocket Money)

Pattern: money found. Open on a number (subscription cost, savings unlocked, debt paid down), show one screen of the app surfacing the insight, end on the projected savings. Trust matters; avoid anything that looks like a sales pitch.

What loses: Generic graphs without specific numbers. Pie charts.

6. Mental health and meditation (Calm, Headspace, Insight Timer)

Pattern: feeling, not feature. Show ambient visual (water, breath, color shift), one screen of the app's core moment, and a single text overlay describing the feeling. Less is more — meditation users dislike high-energy editing.

What loses: Rapid cuts. Voice-over scripts. Selling on "1000+ meditations."

7. Dating (Hinge, Bumble, Raya)

Pattern: outcome, not interface. Show a positive interaction (a match, a message, a date that happened), keep the interface in the background. Users are shopping for connection, not for swipes.

What loses: Demonstrating swipe mechanics. Showing other people's faces longer than 2 seconds.

8. Mobile games — casual and puzzle (Royal Match, Block Blast, Wordscapes)

Pattern: satisfying mechanic on loop. First 3 seconds must show a satisfying core game moment — a chain combo, a level-clear animation, a Match-3 cascade. Repeat the mechanic, escalate the stakes, end on a level-unlock or near-miss.

What loses: Story setup. Loading screens. Anything that delays the satisfying moment past second 3.

9. Mobile games — mid-core and strategy (Clash of Clans, Marvel Snap, AFK Journey)

Pattern: hero and stakes. Open on a character or unit reveal, show progression (a base upgrading, a card collection growing, a unit leveling), end on a competitive moment (a battle, a leaderboard).

What loses: Pure gameplay with no progression. Cinematic intros that are not actual game footage.

10. Shopping and marketplaces (Amazon, Depop, Vinted)

Pattern: search to checkout. Show a desired item, the search or filter that surfaced it, the speed of buying. The video sells the speed of the experience, not the product catalog.

What loses: Generic stock photos of products. Showing the cart icon for more than a second.

11. Streaming and media (Spotify, Apple Music, Netflix)

Pattern: discovery moment. Show a recommendation that hits, a personal-playlist generation, or a familiar artist's cover art. Category is mature; users know what the app does. The video should communicate what makes this app's discovery better.

What loses: Showing the player UI. Users have seen a music player.

12. Utility and tools (Photo editors, scanners, translators)

Pattern: before and after. Show an unedited photo, a one-tap transformation, the result. Or: a piece of foreign text, a one-tap translation, the translated text. Utility apps sell the speed and quality of the output.

What loses: Showing the editing toolbox. Users buy results, not tools.

Production rules that apply to every category

The category pattern dictates the structure. The production rules below are constant.

Rule	Detail
Hook in 3 seconds	First 3 seconds must contain the most compelling visual of the entire video
One core moment per video	Pick one product moment, not five. The video that tries to show every feature shows none
Captions on, audio off	Every video must read silently with on-screen text
No App Store UI	No price tags, install buttons, or platform chrome inside the video
Native footage, not stock	Real app footage beats high-production stock footage in conversion tests
Test against current screenshots	The video competes with screenshot 1; if the screenshot is stronger, the video can hurt
Update every 90 days	Stale videos are often the cause of conversion decay on otherwise healthy pages

How to test the video against the rest of the page

A common mistake is treating the video as a standalone asset. The right test is:

Holdout 50% of traffic on the current page (screenshots only).
Run the video variant against the other 50% with no other page changes.
Measure product page conversion rate over 14 days.
If the video variant lifts conversion 3% or more, ship it. If it is within 2% of control, keep iterating. If it drops conversion, the video is wrong, not "videos do not work."

This sequence isolates the video's contribution. Skipping it produces conflicting test data.

Frequently asked questions

Do I need different videos for App Store and Play? Yes. Format and viewer behavior differ enough that one master video is rarely optimal for both. Plan a shared base concept with platform-specific cuts.

What length is optimal — 15 seconds or 30? The full 30 seconds rarely beats a tight 15. Most winning videos in 2026 run 18 to 22 seconds. The shorter the video, the less room to lose viewers.

Should the video use the same color and style as the screenshots? Yes. Visual consistency between screenshots and video lifts conversion 5 to 10% over visually inconsistent assets. The viewer should not feel they are looking at two different products.

How much should an app preview video cost to produce? $1,500 to $8,000 for a strong custom production in 2026, or under $500 for a competent in-house edit using existing footage. The cost-to-conversion-lift ratio favors in-house production unless the category requires high production value (mid-core games, premium subscription apps).

Is voice-over worth the cost? Rarely. Voice-over is off by default, so most viewers never hear it. If you do use voice-over, it should reinforce captioned text, not replace it.

If you are planning a new app preview video or testing why an existing one is not converting, the Semnexus mobile app marketing team handles preview video strategy and testing. For app teams who need both the video and a coordinated screenshot refresh, the app development team often takes on the asset production alongside store-listing work.