Which languages to localize your app into first (2026)

Which languages to localize your app into first comes down to three things per market: how much people spend on the App Store there, how many of them won't read an English listing, and what the translation costs you to ship and keep current. For most indie apps in 2026 that points to Japanese, German, French, Spanish (Mexico) and Portuguese (Brazil) first, with Simplified Chinese at the top if your app can legally ship in China mainland, and Korean moving up if you make games.

Below is the data behind that order, the exceptions, and a tiered plan for all 50 App Store Connect languages. Every revenue figure is dated, because they move every quarter.

How to decide which languages to localize: a four-factor score

We rank a market on four factors and multiply them in our heads rather than in a spreadsheet:

1. Spend. How much money flows through the App Store in that market.
2. English gap. The share of buyers there who won't search, read or convert on an English listing.
3. Fit. Whether your category sells in that market. A Japanese RPG audience and a Japanese to-do-list audience are different sizes.
4. Cost. The one-off translation cost plus the ongoing cost of updating What's New, promotional text and keywords in that language every release.

A big market with a high English gap and a low cost goes first. A big market where everyone reads English goes later than its revenue suggests. A market you legally can't ship to doesn't go at all, however big it is. That last case is China mainland for a lot of apps, which gets its own section below.

None of this produces a precise ROI number, and we're wary of anyone who claims one. It produces an order, which is what you need.

Top App Store markets by consumer spending

The US is the largest market by a wide margin, then China mainland, Japan, South Korea and the UK. Public country-by-country figures for the App Store alone don't exist, so the best available data mixes iOS and Google Play.

Rank (Q2 2025 IAP revenue)	Market	What the data says	Storefront default language
1	United States	Nearly $60 billion in 2025; more in-app revenue than the next five markets combined	English (U.S.), en-US
2	China mainland	$22.1 billion in 2025, iOS only	Chinese (Simplified), zh-Hans
3	Japan	Only top-20 market with a year-over-year decline in Q3 2025	Japanese, ja
4	South Korea	Ranked fourth in Q2 2025	Korean, ko
5	United Kingdom	Up 24% year over year in Q2 2025	English (U.K.), en-GB

Sources: the ranking and UK growth come from Sensor Tower's Q2 2025 Digital Market Index (August 2025, iOS and Google Play combined, iOS only for China). The US full-year figure is from Sensor Tower's State of Mobile 2026 (January 2026), the China figure from GameDev Reports' summary of the same report, and the "next five markets" and Japan notes from the Q3 2025 Digital Market Index (November 2025).

Three other numbers from the same Q2 2025 report shape the plan. Europe added nearly $1.5 billion in in-app revenue year over year, with the biggest contributions from the UK and Germany. Brazil grew 33%. And iOS took nearly 70% of all in-app purchase revenue while Google Play took 77% of downloads, which is why App Store markets skew toward richer countries.

Apple's own figure is bigger but less useful here. Its 2025 ecosystem study by Analysis Group (published June 2026) counts $1.4 trillion in billings and sales, but most of that is physical goods and services. Digital goods and services were $149 billion, and the newsroom post doesn't break that down by country.

How far does an English-only listing already reach?

Further than you'd think in northern Europe and almost nowhere in East Asia or Latin America. EF's English Proficiency Index 2025 scores 123 countries and regions. It's based on adults who chose to take EF's test, so treat it as a ranking, not a census, but the gaps are large enough to act on.

Market	EF EPI 2025 score	EF band	What Apple's storefront shows by default
Netherlands	624	Very high	Dutch, with English (U.K.) as an additional language
Germany	615	Very high	German (also the default in Austria and Switzerland)
Norway, Denmark	613, 611	Very high	English (U.K.), with Norwegian or Danish additional
Sweden	609	Very high	Swedish, with English (U.K.) additional
Finland	603	Very high	English (U.K.), with Finnish additional
France	539	Moderate	French
South Korea	522	Moderate	Korean
Brazil	482	Low	Portuguese (Brazil)
China	464	Low	Chinese (Simplified)
Japan	446	Very low	Japanese, with English (U.S.) additional
Mexico	440	Very low	Spanish (Mexico)

Scores from EF's 2025 country fact sheets and the EF EPI table; storefront languages from Apple's App Store localizations reference.

Germany is the odd one: very high English proficiency, but a German-default storefront in three countries and a large paying audience. People still search in their own language even when they can read English, and a German keyword field gives you 100 more characters of search terms. So German stays in the first tier, while the Netherlands and the Nordics drop to the long tail. Japan is the opposite case: Japan scored 446, ranked 96th, the lowest it has ever placed. An English listing in Japan is close to invisible.

China mainland: the ICP filing and game approval gate

Simplified Chinese is the right second language only if your app can be on the China mainland storefront at all. Apple's App Information reference says China's Ministry of Industry and Information Technology (MIIT) "requires that some apps possess a valid Internet Content Provider (ICP) Filing Number", and the number plus your app metadata must match what was filed with MIIT, checked against your Simplified Chinese localization. If it's missing, App Store Connect shows the status "ICP Filing Number Missing" and the app isn't available in China mainland.

Games have a second gate. The same page says "games must secure an approval number" from the National Press and Publication Administration (NPPA), verified with your game's ISBN Issuing Letter or Approval Reply and a copy of your Business License. Books and magazines need an internet publishing permit, news apps a permit from the Cyberspace Administration of China, and apps with religious content one from the National Religious Affairs Administration.

The app filing regime has applied to new apps since 1 September 2023, and existing apps had until 31 March 2024 to file, per China Briefing. The filing asks for a business license and identity verification of the people in charge, in Chinese, which in practice means a Chinese entity or partner. We couldn't find an official statement that a foreign developer with no Chinese presence can file directly.

China is also crowded at home. Sensor Tower's State of Mobile 2025 found that more than 90% of downloads and consumer spend in China mainland went to apps from China-based publishers. If you can't file, skip China mainland and treat zh-Hans as a mid-priority language for the US and Singapore storefronts, where Apple lists it as an additional language. zh-Hant covers Taiwan, Hong Kong and Macau, which are separate storefronts outside Apple's China mainland requirements.

Games vs productivity apps: how category changes the order

Category moves languages up or down a tier. In 2025, consumers spent more on non-game apps than on games for the first time, per Sensor Tower's January 2026 release, with non-game in-app revenue up 21%. The year before, Sensor Tower's State of Mobile 2025 put the strongest growth in North America and Europe and described "gaming-focused markets in Asia" seeing modest growth or slight declines, with Japan's mobile game spending down 7% in 2024, while game spending grew fastest in Turkey (+28%), Mexico (+21%), India (+17%), Thailand (+16%) and Saudi Arabia (+14%).

What we take from that:

• Games: Korean and Japanese go to the very top. Turkish, Thai and Arabic earn a place in the second tier. If you have Game Center achievements and leaderboards, translate those too; see Game Center localization.
• Subscription and productivity apps: German, French and the English variants matter more, because Europe and North America are where in-app spending has grown fastest. Korean can wait.
• Utilities with tiny listings: translation is so cheap that the order matters less than just shipping all of tier 1 in one go.

One localization can cover several storefronts

Some locales are worth more than their home country because Apple shows them in many storefronts. From Apple's storefront table:

Locale	Shown in
Spanish (Mexico), es-MX	16 storefronts: Mexico, Argentina, Chile, Colombia, Peru and 11 more in Latin America
German, de-DE	Germany, Austria, Switzerland
French, fr-FR	France, Cameroon, Côte d'Ivoire, Gabon
Arabic, ar-SA	Additional language in 16 Arab storefronts and the US
English (Australia), en-AU	Australia and New Zealand

The US storefront also lists nine additional languages, including Spanish (Mexico), Chinese (Simplified), Korean, Portuguese (Brazil) and Vietnamese. That matters for search as well as display: some storefronts index keywords from their additional locales, so a localization can widen your keyword coverage in a market it was never written for. We cover which ones and how to plan for it in App Store keyword localization, and the full list of codes is in App Store Connect languages.

What app localization costs per language

Metadata is small. A listing with a near-limit 4,000-character description, plus name, subtitle, keywords, promotional text and What's New, comes to roughly 700 English words. That makes the translation itself cheap; the ongoing upkeep is the real cost.

Approach	Rough cost per language, full listing	Notes
Apple's on-device Translation	Free	Literal; doesn't know the 30- and 100-character limits
Your own LLM API key	A few cents	At $2 input and $10 output per million tokens (Claude Sonnet 5.5, October 2026), a listing in and out is a few thousand tokens
Human translation, Gengo Standard	From about $63	Gengo lists Standard from $0.09 per word
Human translation, Gengo Advanced	From about $126	From $0.18 per word
Native keyword research	Hours, not dollars	No translator does this by default; keywords need search phrasing, not translation

The per-word prices are Gengo's published starting rates; the per-language totals are our arithmetic at 700 words. Loci translates with Apple's free on-device Translation or your own Claude, OpenAI or Gemini key, so the machine-translation rows are what we see in practice.

Two consequences. First, with machine translation, the cost of adding a language is mostly the time to review it, so the long tail is cheap enough to do eventually. Second, the bigger the market, the more a human pass on the name, subtitle and keywords pays back. Those 160 characters drive search; the description mostly doesn't. Hand-check those three fields in every tier 1 language, even if you machine-translate the rest.

How to measure whether a language paid off

App Store Connect gives you enough to judge each language without third-party tools. In App Store Connect Analytics, the Acquisition section shows unique impressions, unique product page views, total downloads and conversion rate, which Apple defines as how often unique impressions turn into total downloads. You can filter all of it by territory.

The routine we use:

1. Write down the date each localization went live with an approved version.
2. Filter Acquisition to the territory (Japan for ja, the 16 Latin American storefronts for es-MX) and compare the 28 days before with the 28 days after.
3. Read impressions as search visibility. If they rise, your localized name, subtitle and keywords are being found.
4. Read conversion rate as persuasion. If impressions rise but conversion doesn't, the screenshots or the first lines of the description are still doing the work in English.
5. Compare against a territory you didn't touch in the same window to separate the localization from seasonality or a feature.
6. Check proceeds by territory in Trends a month later. Downloads that don't pay aren't a win for a paid app.

Territory in Analytics is the storefront from the customer's billing address, per Apple's filters and dimensions reference. A German speaker with a US account counts as the US. For a small app, a week of data in a single territory is noise; give it a month.

A tiered plan: the first 5, the next 10, then the long tail

Here's the order we'd use for a typical paid or subscription iOS app with an English (U.S.) primary language. Adjust with the category rules above. For a wider view of the whole process, start with the App Store localization guide.

Before tier 1: the English variants

English (U.K.) en-GB, English (Australia) en-AU and English (Canada) en-CA are separate localizations, each the default in its storefront, each with its own 100-character keyword field. The UK was the fifth-largest market in Q2 2025 and the fastest-growing of the top five. The cost is spelling conversion ("color" to "colour") and a fresh set of keywords, so do these the same day.

Tier 1: the first five

Locale	Why
Japanese, ja	Third-largest market, very low English proficiency (446)
German, de-DE	Europe's growth leader alongside the UK; default in three storefronts
French, fr-FR	Moderate English (539); default in four storefronts
Spanish (Mexico), es-MX	One localization, 16 storefronts; Mexico scores 440
Portuguese (Brazil), pt-BR	33% revenue growth in Q2 2025; low English (482)

Swap rules: if your app can get an ICP filing (and an NPPA approval number for games), put Simplified Chinese at the top of this tier. If you make games, put Korean in place of French.

Tier 2: the next ten

Locale	Why
Korean, ko	Fourth-largest market; moves to tier 1 for games
Chinese (Simplified), zh-Hans	Additional language in the US and Singapore even without China mainland
Chinese (Traditional), zh-Hant	Taiwan, Hong Kong and Macau, outside the China mainland gate
Italian, it	Moderate English (513); a large European market
Spanish (Spain), es-ES	Spain's storefront default; es-MX reads as Latin American there
Turkish, tr	Fastest game-spending growth in 2024 (+28%); low English (488)
Arabic, ar-SA	16 Arab storefronts plus the US; Saudi Arabia scores 404
Thai, th	Very low English (402); game spending +16% in 2024
Vietnamese, vi	Additional language in the US storefront as well as Vietnam
Russian, ru	Additional language in the US storefront as well as its own; check your own territory data first

The long tail: everything else

The remaining 31 languages: Dutch, Swedish, Danish, Norwegian, Finnish, Polish, Czech, Portuguese (Portugal), French (Canada), Hebrew, Greek, Hungarian, Romanian, Slovak, Ukrainian, Croatian, Malay, Indonesian, Catalan, Hindi, and the 11 Apple added in March 2026 (Bangla, Gujarati, Kannada, Malayalam, Marathi, Odia, Punjabi, Slovenian, Tamil, Telugu and Urdu).

These are either markets where English already reaches most buyers (the Netherlands and the Nordics, where three storefronts default to English (U.K.) anyway) or markets where in-app spending is small relative to downloads. With machine translation each one costs minutes, so add them in batches once tiers 1 and 2 are stable, keep them on name, subtitle, keywords and description, and don't feel obliged to translate every What's New.

A last check before you push any tier: run every field against its limit (30 characters for name and subtitle, 100 for keywords) in the target language, since German and French run long and Japanese packs more meaning per character.