App Store localization in 2026: the indie dev's guide to translating screenshots

If you've ever looked at your App Store Connect analytics and noticed that 60% of your impressions come from countries where the listing is still in English, this is for you.

App Store localization — translating your screenshots, title, subtitle, and description into other languages — is one of the highest-ROI growth levers indie devs almost never pull. Not because they don't know it works. Because they've never had a fast way to do it.

This guide covers what actually moves the needle: which parts of your listing matter most, which languages return the most installs per hour spent, and how AI tools have collapsed the cost of doing this from "weeks of translator coordination" to "thirty seconds per screenshot".

What App Store localization actually means

Apple lets you provide a separate version of every piece of your App Store listing for each of 34 supported languages:

• App name + subtitle
• Promotional text
• App description
• Keywords
• Screenshots
• App preview videos
• What's New text (per release)
• In-app purchase names + descriptions

If you don't provide a translation for a given language, Apple falls back to your primary language (usually English). Which means a user in Tokyo browsing the Japanese App Store sees your English text — and converts at a fraction of the rate a Japanese user would on a fully localized listing.

How big a fraction? Apple has published case studies showing +10% to +200% download lifts from full localization, with screenshots being the single biggest contributor. Text-only translations (description, keywords) usually deliver less than half what screenshot localization does, because the App Store carousel is what most users actually look at.

The myth that's stopping you

The default mental model for indie devs:

"Localizing means hiring translators for every language. Each one charges $0.10–0.30/word. I have 50 words per screenshot × 5 screenshots × 10 languages = $250–$750 just to translate the text. Then I need to redesign each screenshot in each language. Total: maybe $3k–$10k of work. Skip."

That math was correct in 2020. In 2026 it's not — for two reasons:

1. Apple ranks based on download velocity in each storefront. A listing that converts well in German earns higher rank in the German App Store, which earns more impressions, which compounds. Skipping localization isn't a "save money" decision; it's an actively rejected growth lever.

2. AI removed the per-language cost. A good AI screenshot localization tool can swap English headlines to German, Japanese, Korean, etc. while keeping the design identical — for cents per screenshot. The translator+designer pipeline isn't required for most indie apps anymore.

Which languages actually pay back

You don't need to localize into all 34 supported languages on day one. The data on App Store revenue by country tells you exactly where to start.

In rough order of revenue contribution for most indie apps:

| Tier | Languages | Why | |---|---|---| | 1 — must-have | Simplified Chinese, Japanese, Spanish, German | High ARPU, large user bases, low English fluency relative to spend | | 2 — high-ROI | French, Korean, Italian, Brazilian Portuguese, Russian | Strong markets, App Store cultures that value local-language listings | | 3 — depends on category | Dutch, Polish, Turkish, Arabic, Hindi, Indonesian, Thai | Strong in specific verticals (gaming, utility, social) | | 4 — completeness | Nordic languages, Eastern European, smaller markets | Small marginal lift each but cheap to add once you have a workflow |

If you only do four, do Chinese (Simplified), Japanese, Spanish, German. These four cover roughly 50% of non-English App Store revenue.

A rule of thumb: the smaller your app, the more important localization is. Established apps already get organic global discovery from being well-known. An indie launch competing for the same keyword in the German App Store with a German-language listing will beat a competitor with an English-only listing 9 times out of 10, regardless of which app is better.

What to localize: ordered by ROI

If you've never localized anything, do these in this order:

1. The first 2–3 screenshots — biggest single ROI

On the App Store search results page, Apple shows the first 2 or 3 screenshots in a horizontal carousel before users tap into the listing. Those are the only thing most non-converting users see. If those screenshots are in English to a German speaker, you lose. Localizing just these 3 images per language typically captures 60–80% of the total localization lift.

2. App name + subtitle

These show up in search results next to your icon. They're indexed for keyword search in each language's App Store. If you don't localize them, you're invisible to native-language searches. Cheap to translate (10–20 words). Massive impact on discoverability.

3. Remaining screenshots

The 4th-through-10th screenshots. People who tap into your listing read more. Doing all 10 in each language doubles the per-language lift over just the first 3.

4. Description + promotional text

Modest impact on conversion, big impact on App Store keyword ranking. Worth doing eventually but not until you've done the above.

5. What's New + IAP localization

Marginal. Skip until you have revenue from a language to justify the time.

Common mistakes (and how to avoid them)

Translating literally instead of idiomatically

"Track expenses in 5 seconds" translates literally to German as something stiff and unnatural. A native German speaker would write "In 5 Sekunden den Überblick behalten" — different verb, different cadence, sounds like a German marketer wrote it. Headlines in particular need an idiomatic pass, not a literal one. AI tools that ask for the "natural" or "idiomatic" translation (not the verbatim one) do this well by default.

Forgetting that German is 30% longer than English

Translated text changes length. German runs 20–30% longer than English. Japanese can be 30% shorter. If you don't adjust font size or layout, German headlines overflow your design; Japanese ones look small and lost in whitespace.

A serious tool either auto-adjusts the typography or, ideally, generates the same design with appropriately-sized text for each language. Tools that just paste new text into the same template look amateur.

Localizing screenshots without localizing the app

If the user taps GET expecting a German app and opens it to find English-only UI, they leave. Localize your in-app strings before (or at least simultaneously with) your App Store listing.

Using machine translation for keywords

App Store keywords are 100-character compressed lists ranked for SEO. Machine-translating them word-for-word usually picks wrong synonyms and loses ranking. Either use a native speaker, an ASO consultant for that language, or skip keywords and lean on app name + subtitle.

Mixing scripts when you shouldn't

Don't put English brand names inside a Japanese-localized screenshot unless your brand is genuinely English (like a product name nobody translates). It looks like you forgot to finish localizing. Either keep brand names romanized OR fully transliterate them — consistency is what reads as intentional.

The hard part: maintaining localizations

Here's the trap nobody warns you about. You ship version 1.0 with screenshots translated into 8 languages. Three months later, you redesign your app. Now you have:

• 1 new English screenshot set
• 8 outdated localized sets you have to redo

Most indie devs hit this wall and quietly let the localizations get stale. After a year, your German listing is showing screenshots from a UI you ripped out two versions ago, which makes you look unmaintained.

The way around this: build a workflow where re-localizing is cheaper than the original localization. Source your screenshots in one language, run them through an AI localization tool every time you ship, replace the App Store assets in one batch. If a re-localization run takes 10 minutes, you'll actually do it.

What AI screenshot localization gets right (and where it fails)

Honest assessment of where the tech is in 2026:

Wins:

• Translation quality from gpt-image-2 and similar models is at native-speaker quality for major languages (Spanish, French, German, Japanese, Korean, Chinese). Reads idiomatically, not machine-translated.
• Layout preservation is solid for headline-only swaps. The model keeps the typography, palette, photography, in-app UI mockup all identical.
• Cost per screenshot is well under $1, often well under $0.50 with the cheaper models.
• Speed: seconds per screenshot. A full 10-screenshot localization to one language takes ~2 minutes.

Open problems:

• Layout shifts on full-text-replace (when you translate ALL visible text, not just headlines). The model occasionally nudges fonts or spacing to fit longer translated text. Less reliable than headline-only swaps.
• Right-to-left languages (Arabic, Hebrew) sometimes need the layout mirrored, not just the text translated. Current tools don't do this automatically.
• Languages with complex typography (Thai, Japanese vertical, Arabic ligatures) occasionally render imperfectly at small sizes.

For most indie dev workflows, the wins dramatically outweigh the open problems. If a translation isn't quite right, you re-run with adjusted prompting. The cost of trying is so low that experimenting is cheap.

A 30-day localization plan

If you're starting from scratch and want to be in 5 languages by next month:

Week 1: Localize your first 3 screenshots to Chinese (Simplified), Japanese, Spanish, German. Headlines only (most stable AI output). Estimated time: 1 hour with a good AI tool. Submit the update.

Week 2: Localize your app name + subtitle for the same four languages. Use a native speaker for these — they're short, high-leverage, worth getting right. Update App Store Connect.

Week 3: Localize screenshots 4–10 to the same four languages. Submit the update.

Week 4: Watch your App Store Connect analytics. The four target storefronts should show measurable conversion-rate lift. Pick one or two more languages from Tier 2 based on which storefronts show the most impression growth. Localize and submit.

After 30 days you'll have data telling you whether to keep going (most likely yes, in which case do Tier 2 next month) or whether your category just doesn't pay back from a specific language (some niche utility apps don't, and that's fine).

TL;DR

• Localization is a 10–200% download lift indie devs leave on the table because the old workflow was expensive.
• AI tools collapsed the cost. Modern screenshot localization is seconds and cents per image.
• Screenshots are the single biggest contributor. Do the first 3 screenshots in 4 languages (Chinese, Japanese, Spanish, German) first. That's the 80/20.
• Don't translate literally — translate idiomatically. The way a native marketer would write it.
• Build a workflow that makes re-localization cheap so you actually maintain it across app updates.
• Localize screenshots even if your in-app text is still English — but plan to catch up on the app.

---

Screenshot Roast localizes your existing App Store screenshots into any of 34 App Store-supported languages while keeping your design pixel-equivalent. First roast is free; localization is part of the Pro plan ($20/mo).