Localising an iOS app properly costs about $0.20 per word, plus translator review, plus screenshot redesign, plus QA. For a small app, that is hundreds to low thousands of dollars per language. You should not localise everywhere by default. You should localise where the data says yes.
The two-stage approach
Stage 1 — localise the App Store listing only. Title, subtitle, description, keywords, screenshots. No app strings.
This costs perhaps $50–150 per language, takes a week, and tells you something invaluable: does the listing convert in that language? If installs from a market grow noticeably, the demand is real and stage two is justified.
Stage 2 — localise the full app. Strings, paywall, error messages, support docs.
This is the expensive step. We do not enter it for any language until stage one has shown a clear signal.
The languages worth doing first
For most utility apps in 2026, in this order:
- English (US default).
- Spanish (Latin America split from European Spanish if you can; Latin American Spanish is the bigger market for most app categories).
- Portuguese (Brazil — large, growing, often underserved).
- German (high purchasing power, picky reviewers, willing to pay).
- French (similar to German on willingness to pay).
- Japanese (high revenue per user, expensive translation, do it well or not at all).
- Italian, Dutch, Polish, Russian depending on category.
Korean and Simplified Chinese are large markets but each carries category-specific risks (regulatory, competitive, distribution). Approach them only with category research.
What rarely pays back
- Languages with very small App Store revenue per user unless your category is extremely strong there.
- Heavy localisation of every micro-copy in an app where the audience is largely English-comfortable. Localise the listing; consider not localising the in-app copy.
- Auto-translated strings. Always worse than English for most users in the target language. Hurts conversion. Genuinely.
A note on the App Store algorithm
Apple's Search algorithm respects the language of the localised metadata. A well-localised listing ranks for native-language keywords in that locale. This is a real, free distribution boost — but it requires hand-translated metadata, not auto-translated. Translators who know your category translate keywords better than translators who do not.
A note on screenshots
If you localise the listing but use English screenshots, the conversion drop is substantial in non-English locales. Localised screenshots are not optional. They are the visible promise.
What we do at Tappa
For each app, we localise the listing into the top six languages by category-specific revenue. We localise the app strings into the top three by signal. We do not localise into more than that until one of two things happens: a market shows runaway organic growth, or a partner asks for it.
This discipline frees up budget for things that actually move the needle — better screenshots in the languages we already cover, more thoughtful onboarding, and one more shipping engineer.