App Store screenshots and previews: your conversion levers
Your screenshots and app preview video are not search-ranking factors. Apple does not index them for keywords, the way it indexes your app name, subtitle, and the hidden keyword field. What they do is decide whether someone who already reached your product page actually taps Get. They are pure conversion levers, and the first two to three frames carry almost all the weight, because that is what shows above the fold before anyone scrolls. Think of the visual assets as the second half of an ASO system: metadata wins the impression in search, the visuals win the install on the page. The two halves are measured differently and tested differently. You optimize the keyword field, name, and subtitle for visibility against Apple Search Popularity (treated as a relative index, never fabricated search volumes); you optimize screenshots, captions, and the preview against download rate in App Store Connect. This guide stays on the visual half: which frames matter, how orientation and captions change the read, what the autoplaying muted preview can and cannot do, why localizing the whole set beats translating captions, and where the App Review rules quietly constrain what you can put in frame one.
Why the first two to three screenshots matter most
On the App Store product page, only the first two to three screenshots (or the preview video plus the next frame) are visible before a user scrolls. In search results the preview is even tighter: users see a sliver of your leading frames inline. Most visitors never scroll the full gallery. So your single strongest value proposition, your most legible UI, and your clearest 'this is what the app does' moment all belong in those leading frames. Treat positions one through three as your headline and positions four onward as supporting detail for the minority who keep looking. If your best frame is sitting in slot five, you are spending it on almost no one. A practical test: cover everything past frame three with your hand and ask whether a stranger could state what the app does and why it is better. If they can't, reorder before you redesign. Sequencing matters too: lead with the core benefit, then a differentiator, then proof (a chart, a result, a recognizable integration). Don't open with onboarding or a settings screen; those are slots three-plus content at best. You can supply up to ten screenshots per device size, but the back half is for the committed reader, not your pitch.
How portrait vs landscape changes what shows above the fold
Orientation is a layout decision with conversion consequences. Portrait screenshots show roughly three at a time on an iPhone product page, so a portrait set front-loads more frames above the fold, usually the right call for most apps, because it gives you three chances to land the value instead of one. Landscape screenshots display larger but typically show only one to two at once, which suits games and video apps where the wide composition is the point and a single cinematic frame does more than three cramped ones. Pick the orientation that puts the most persuasive content in the visible region, and keep the whole set consistent so the page doesn't feel broken; a portrait set with one landscape frame mid-gallery reads as a mistake. Remember Apple requires assets sized for current device displays (6.9-inch and 6.5-inch iPhone classes, plus iPad if you ship there); the iPad page has its own gallery and its own above-the-fold math, so don't assume an iPhone-tuned sequence translates. If you support both, design the iPad set deliberately rather than letting it scale down an iPhone layout.
Should screenshots have captions?
Yes. Captioned screenshots almost always outconvert raw UI grabs, because a bare screen makes the visitor do the work of figuring out the benefit. Add a short caption (three to six words) stating one concrete benefit per frame: 'Track every workout', 'Offline maps included', 'No ads, ever'. Keep captions readable at thumbnail size, since the above-the-fold frames render small in search results and on the product page before tap; a 40-character sentence in thin gray type is illegible at that scale. Favor high contrast, a heavy weight, and short lines anchored top or bottom away from device chrome. Each frame should make exactly one claim; stacking three benefits onto one screenshot dilutes all three. These captions are marketing copy, not metadata: they are not indexed for search, so write them to persuade a human, not to chase a keyword. There is no ASO penalty or benefit to the words you choose here, which frees you to use natural, punchy language ('Plan trips offline') instead of the comma-separated keyword discipline you'd apply to the hidden 100-byte field. Avoid claims you can't show on the screen behind the caption; a benefit the UI visibly delivers is more convincing than a slogan.
What the app preview video does (and its first seconds)
The app preview is a 15-30 second video that autoplays muted in the first screenshot slot. Because it autoplays without sound, the opening two to three seconds must communicate the core experience visually: real in-app footage, no long logo intro, no title card burning your runway. Apple requires previews to be made mostly of captured device footage, so show the app actually working rather than a marketing sizzle reel; a preview that is all motion graphics risks rejection and, worse, tells the viewer nothing about the product. You can supply up to three previews per localization, but only the first occupies the prime autoplay slot, so make that one carry the message. A poster frame (the still shown before playback or when autoplay is unavailable) should itself be a strong screenshot, because some contexts show the frame, not the motion. A strong preview can lift conversion, but a slow or abstract one wastes your most prominent slot; if you can't make those first seconds land, a sharp first screenshot may convert better, and an empty first slot is never the answer. When in doubt, lead with the single action that delivers the app's payoff.
Why you should localize the screenshot set, not just translate
Screenshots can be localized per storefront, and this is one of the highest-return ASO moves available. Localizing captions into the storefront's language removes friction for that audience, but the bigger win is cultural and contextual fit: the example data, currency, units, date formats, and imagery that make the app feel native. A user scanning a localized set in their own language with familiar context (yen prices, kilometers, local place names, faces that look like their market) converts at a meaningfully higher rate than one squinting at English captions and dollar amounts. Prioritize localized screenshots for your highest-traffic storefronts first; you don't need all 40-plus localizations on day one. A pragmatic tiering: full localization (captions plus in-frame data) for your top revenue markets, caption-only translation for the mid tier, and the default set for the long tail. Watch right-to-left languages (Arabic, Hebrew) where layout, caption alignment, and even reading direction of the frame sequence flip. Localization also interacts with the preview: an English voice-free preview travels, but on-screen English text inside the footage does not, so localized text overlays inside the video matter for top markets.
App Review and content rules that constrain what you show
Your visual assets are subject to App Review, and rejections here cost release days. Screenshots must reflect the actual app: no mockups of features you don't ship, no fabricated content, no devices or platforms you don't support shown in frame. Don't show pricing claims you can't honor, don't imply rankings or awards you don't hold, and don't include other companies' trademarks or competitor names as comparisons. Status bars, notches, and device frames should be clean and current; outdated device chrome signals a stale app. Avoid placing promotional badges ('Editor's Choice', 'Top App') unless Apple actually granted them; fabricated badges are a common rejection. If your app has age-rated content, your screenshots must stay within that rating; a frame too explicit for the app's rating gets flagged. Keep marketing text in captions truthful and non-comparative. Treat the first frame especially carefully, since it surfaces in search and lists where scrutiny is highest. Getting these right up front avoids the cycle where a rejected screenshot set blocks an otherwise-approved binary, delaying a launch you've already scheduled.
How screenshots affect conversion, but not search rank
Keep the mental model clean: keyword fields, app name, and subtitle drive whether you appear in search; screenshots and previews drive whether appearing turns into a download. The weight order for text relevance is app name, then subtitle, then the hidden keyword field, then categories, none of which your images touch. They influence rank only indirectly: Apple factors conversion rate into ranking, so a page that converts better can rank better over time, but the screenshots themselves are never indexed for text relevance. This is why you A/B test screenshots against download rate, not against keyword position. Use Product Page Optimization to split live traffic across up to three variants and read the lift against product-page-view-to-install, and let the test reach a meaningful sample before calling it. Optimize the metadata for visibility and the visuals for the tap. One more loop worth naming: because conversion feeds back into rank, a screenshot redesign that lifts install rate can, over weeks, also nudge your search position, but you should attribute and measure that through App Analytics, not assume it, and never confuse it with editing indexed text.
FAQ
Do App Store screenshots affect keyword search ranking?
No. Screenshots and the app preview video are not indexed for search. They drive conversion (whether a visitor downloads) which can influence rank indirectly because Apple factors conversion rate into ranking, but the images themselves carry no text-relevance weight. Your text relevance comes from the app name, subtitle, and the hidden 100-byte keyword field, in that weight order. Write captions to persuade humans, not to chase keywords; there is no indexing benefit to the words you place on a frame.
How many screenshots show above the fold on the App Store?
Roughly two to three portrait screenshots (or the preview video plus the next frame) are visible before scrolling. Landscape sets usually show only one to two. Most users don't scroll, so put your strongest frames in the first two to three slots and treat positions four through ten as supporting detail for the committed minority. A quick check: cover everything past frame three and confirm a stranger could state what your app does and why it's better.
How long should an App Store app preview video be?
15 to 30 seconds, autoplaying muted in the first slot. Make the opening two to three seconds communicate the core experience visually with real in-app footage, with no long logo intro, since it plays without sound. Apple requires previews to be mostly captured device footage, so show the app working rather than a motion-graphics reel. You can supply up to three previews per localization, but only the first gets the prime autoplay slot, so make that one carry the message.
Should I localize my screenshots or just translate the captions?
Localize the whole set where you can. Translating captions helps, but adapting example data, currency, units, dates, and imagery to the storefront's context converts noticeably better than English data with translated text. Prioritize your highest-traffic storefronts first: full localization for top markets, caption-only for the mid tier, the default set for the long tail. Watch right-to-left languages, where layout and reading order flip, and localize any English text baked into the preview footage for top markets.
Can my screenshots get rejected by App Review?
Yes. Screenshots must reflect the actual app: no mockups of unshipped features, no fabricated content, no unsupported devices in frame, and no awards or 'Editor's Choice' badges Apple didn't grant. Keep marketing claims truthful and non-comparative, avoid competitor trademarks, and stay within your app's age rating. A rejected screenshot set can block an otherwise-approved build, so vet the assets (especially the first frame, which surfaces in search) before submitting to avoid losing release days.
appusula puts these checks in one self-hosted dashboard: keyword ranks, metadata validation, competitor tracking, and AI visibility.
Get started