Localized screenshots,
without the spreadsheet spiral.
The App Store serves about 40 locales; Google Play serves over 80. A user browsing in Tokyo sees your English screenshots only because you never made Japanese ones, and listings localized into a browser's language convert meaningfully better than English fallbacks in nearly every non-English market. Screenshots are the visible half of that. Here's how to ship them without drowning.
What "localized screenshots" actually means
There are three depths, and you don't have to pick the deepest one to get most of the value:
- Tier 1, captions only. The marketing text on the screenshot is translated; the app UI inside the frame stays English. An hour per language, and it moves the needle because captions carry the argument.
- Tier 2, captions plus UI. The screenshot inside the device frame is captured with the app running in that locale. Looks native, costs a capture pass per language, and requires your app to actually be localized.
- Tier 3, culturally adapted. Demo data changes too: names, currencies, dates, content. A Tokyo weather screenshot shows Tokyo. This is where big-budget listings live; indies rarely need it before the first two tiers are done.
The honest default for an indie: tier 1 for every locale you localize at all, tier 2 for your top two or three markets once the app itself speaks those languages. A localized caption over an English UI is imperfect but useful; an English caption over an English UI is invisible.
Which locales are worth it
Not a ranking of countries by size, a ranking by return for your app. Three signals, in order:
- Your own analytics. Store impressions and installs by territory. A market already finding you in English converts far better once greeted in its own language.
- Category gravity. Productivity and utilities do disproportionately well in German, Japanese, and French markets; games shift toward Japanese, Korean, and Simplified Chinese; finance apps track wherever your product can legally operate.
- Willingness to pay. If you monetize, Japan, Germany, and the Nordics punch far above their download counts.
The common indie ladder: ship English, then add German, Japanese, French, Spanish, and Brazilian Portuguese, then let analytics argue for the next five. Five languages at tier 1 is a weekend, not a quarter.
The math, before you commit
Screenshot localization scales multiplicatively, and it's worth doing this arithmetic while you can still say no. Ten screenshots × two required App Store sizes × six locales is 120 rendered images, per release whose screenshots change. Play adds its own set. The workflow question isn't "can I make these once", it's "can I regenerate all of them in an afternoon when the UI changes".
Both stores fall back to your primary locale for any language you skip, missing locales show English, not blanks. Localization is always additive; there's no penalty for doing five languages instead of forty.
Text expansion, CJK, and RTL, the layout tax
Translated captions don't stay caption-sized, and layouts designed at English length break quietly:
- German and French run long. Plan for 30 to 35% more characters than English. "Track your habits" becomes "Verfolge deine Gewohnheiten", nearly double. If your English caption already fills the width, the German one won't fit at a legible size; write shorter English or budget a smaller cap.
- CJK runs short but dense. Japanese and Chinese captions use fewer characters at larger optical weight, they usually fit easily, but line-break rules differ; a caption that wraps mid-word in Japanese reads as broken. Break lines manually for these locales.
- Arabic and Hebrew mirror. RTL locales flip reading order: captions align right, and a composition that "points" left-to-right (text on the left, device on the right) reads backwards. Mirror the layout for RTL sets, not just the text.
A workflow that survives release two
- 1, Freeze the design once. Build the full screenshot set in your primary language: composition, backgrounds, frames, caption positions. Save it as a template. Design decisions are now done for every locale.
- 2, Keep captions in one flat file. A single sheet: rows are screenshots one through ten, columns are locales. This is the only artifact translators ever touch, and the only thing that changes per language.
- 3, Translate meaning, not words. Captions are marketing claims, and word-for-word translation reliably flattens them. Give translators the screenshot image and one line of intent, "this one promises privacy", and let them write the native version of the claim.
- 4, Get one native read per language. Machine translation plus a native speaker's ten-minute review beats agency-grade translation with no review. The failure mode you're guarding against isn't grammar, it's tone-deaf phrasing on a public billboard.
- 5, Render per locale, upload per locale. Load the template, paste the locale's column, export the sizes, upload to that locale in App Store Connect or the Play Console. Repeat. Boring on purpose.
The mistakes that ship most often
- Verbatim caption translation, grammatically right, persuasively dead. See step 3.
- Locale-mixed screenshots, a Japanese caption over an English UI is a fine tier 1 shot; an English caption over a Japanese UI is a mistake, because the marketing layer should always match the browsing locale first.
- Visible artifacts of the wrong locale, an English keyboard, "$" prices, or MM/DD dates inside a shot that claims to be German. Reviewers don't care; users do.
- Localizing screenshots for a locale the app doesn't support, a German listing for an English-only app buys installs that become one-star reviews. Match listing locales to app locales.
- Skipping re-renders after a redesign, stale localized screenshots outlive UI changes by years. The one-flat-file workflow exists precisely so regeneration is cheap enough to actually happen.
Localized captions are the cheapest conversion lift most indie listings never ship. Write the English set properly first, the copywriting guide, then spend the weekend.