App Store search keywords by locale: a 2026 ASO primer

Keyword research is not a single global exercise. This primer explains how App Store keyword fields work per locale, where search behavior diverges most by language, and how to build native keyword lists for the markets that reward ASO investment most.

Keyword research is the foundation of App Store Optimization, but most guides treat it as a single global exercise. In practice, App Store search is fragmented across more than 175 storefronts, each with its own language, search vocabulary, and user intent signals. A keyword strategy built in English and machine-translated into Japanese, German, or Brazilian Portuguese will underperform — sometimes dramatically — compared to one built natively for each locale.

This primer walks through how keyword fields work per locale, where language-specific search behavior diverges most sharply, and how to build a market-by-market keyword list that actually moves the needle on impressions and conversion.

How App Store keyword fields work across locales

App Store Connect allows developers to submit separate metadata for each language supported in a storefront. The fields that are indexed for search are the title (up to 30 characters, with the highest weighting in Apple's ranking algorithm), the subtitle (up to 30 characters, secondary weighting), the keyword field (up to 100 characters of comma-separated terms, not visible to users), and in-app purchase display names (partially indexed for relevant searches).

Apple does not index the long description for keyword ranking purposes. This is a common misconception that wastes developer effort. The 100-character keyword field is your precision instrument; every character is a ranking signal.

Because all metadata fields are per-locale, you have the opportunity — and the obligation — to build a separate keyword strategy for each language you support. Running your English keyword list through a translation API and submitting the result is one of the highest-leverage mistakes in App Store Optimization. It wastes character budget on terms that may not reflect how native users actually search, and it misses the structural differences between languages that dictate which keyword forms actually get matched.

How search behavior diverges by language

Users in different locales don't just search in different languages — they search differently. Research and published case studies from ASO practitioners at Phiture, AppFollow, and MobileAction consistently highlight several structural patterns worth understanding before building any keyword list.

Japanese. The App Store Japan indexes hiragana, katakana, and kanji separately. A keyword entered in romaji (romanized Latin characters) will not match a user searching in hiragana or katakana, even for the same phonetic word. Effective keyword lists for Japan must include native-script terms. Japanese users also tend to search with shorter, high-intent queries rather than multi-word phrases, which influences how you should structure and prioritize terms.

German. German forms compound nouns that function as single search terms — Haushaltsbuch (household budget book) rather than Haushalt and Buch separately. A keyword strategy targeting only individual component words will miss many high-intent searches. ASO practitioners report that compound-word coverage is one of the biggest gaps in machine-translated German keyword sets, and closing that gap often produces meaningful impression gains in a market that offers some of Europe's highest App Store revenue per user.

Brazilian Portuguese vs. European Portuguese. These are distinct locales in App Store Connect (pt-BR and pt-PT). Vocabulary, spelling, and even product category names differ substantially. Apps localized for one market routinely fail to rank meaningfully in the other. Given that Brazil is one of the fastest-growing App Store markets globally, treating both Portuguese locales as interchangeable is a significant missed opportunity.

Arabic. Arabic is a right-to-left language with a morphologically rich root system. A single English keyword might map to a dozen Arabic root-form variants. Because App Store search in Arabic generally matches on root forms rather than exact strings, keyword research for Arabic requires native-language expertise to identify the canonical search terms users actually type. Direct word-for-word translation produces keyword lists that are technically correct but pragmatically useless for search matching.

Korean. The Korean App Store is one of the highest-revenue markets in Asia. Korean users are active app searchers, and the App Store indexes Hangul (Korean script) precisely. Published reports from MobileAction and other ASO platforms suggest that apps with properly localized Korean keyword fields consistently outrank apps relying on English or auto-translated metadata. As with Japanese, Latin-alphabet keywords have minimal effect on Korean search rankings.

Building locale-specific keyword lists without a dedicated ASO team

The goal is not to translate your keyword list — it is to rebuild it from first principles in each target language. Here is a workflow that scales without requiring a full-time ASO hire.

Start with category leader analysis in-locale. In each target locale, set your device or App Store Connect preview to that storefront and search for your app's primary use case. Note the exact terms that surface top-ranking competitors. Those terms represent the baseline vocabulary for that locale. ASO tools like AppFollow, MobileAction, and Sensor Tower expose estimated keyword volume and difficulty scores by storefront, which helps you prioritize between overlapping terms once you have a raw list.

Mine competitor titles and subtitles natively. Look at the titles and subtitles of the top three apps in your category within the target locale. The vocabulary patterns visible in their user-facing metadata represent terms they have found worth prioritizing — and terms the Apple algorithm has rewarded with rankings. This is where German compound-word construction, Japanese script-specific terms, and regional vocabulary splits in Spanish (Mexico vs. Spain vs. Colombia) become visible without requiring a subscription to a keyword tool.

Use autocomplete as a free signal. The App Store's search bar autocomplete is powered by real search volume data. Typing partial terms in the target language on a device set to that locale reveals high-intent long-tail terms that keyword tools frequently miss. This takes 15-20 minutes per locale and is one of the highest-signal, lowest-cost inputs available to an indie developer.

Validate with a native speaker. Machine translation tools are adequate for understanding meaning, but they are not reliable for identifying the colloquial terms users actually type when searching for apps. A 30-minute review with a native speaker — even a freelancer on Upwork briefed on the specific task — will routinely surface five to ten high-value terms that a translation engine misses or distorts. This is the step most developers skip, and it is often where the largest ranking gains are found.

Pack the keyword field efficiently. Apple does not require spaces between comma-separated keywords; commas alone are the delimiter. Singular forms are generally sufficient — Apple's algorithm handles pluralization in most Latin-script languages. Numbers as numerals rather than spelled-out words save characters. Avoid repeating terms already present in your title or subtitle, since Apple's algorithm already indexes those fields and keyword field duplication wastes your 100-character budget.

Measuring keyword performance by locale

Apple provides impression and tap-through data in App Store Connect under Analytics → App Store → Search. Filtering by territory and comparing before and after a metadata update lets you track whether keyword changes have moved impression counts in a specific locale. What Apple does not provide natively is keyword-level attribution — you cannot see which specific keywords drove which impressions without a third-party tool. AppFollow, MobileAction, and Sensor Tower all offer keyword rank tracking by storefront, which is the primary feedback loop for locale-specific keyword experimentation.

A minimal measurement protocol: record your current keyword rankings for ten to fifteen target terms in each locale before submission, wait four to seven days after an App Store Connect metadata update for indexing to stabilize, then compare storefront-level impressions and tap-through rates before and after. Track conversion rate per storefront separately from your global conversion — a keyword change that expands impressions can temporarily affect conversion if the new searchers have different intent than your previous audience.

Keyword updates are metadata-only submissions — no binary is required — and Apple typically reviews them within 24 to 48 hours. This makes keyword iteration significantly faster than many developers realize. You can run a localized keyword experiment, measure results, and adjust within a single week, without waiting for a full app review cycle. For apps with IAP or subscription offerings, the IAP display name is an additional indexed field worth localizing carefully; see iOS in-app purchase localization: the 5 fields most devs forget for detail on how that field affects both search and conversion.

Sources and further reading

• Apple Developer — App Store Search
• Phiture Mobile Growth Stack — ASO Keywords
• AppFollow Blog — ASO and keyword research
• MobileAction Blog — App Store keyword data and market analysis
• Sensor Tower Blog — App Store market intelligence
• App Store Connect Help — Add keywords