Apple added 11 languages to the App Store. In most of them, your keyword field just shrank by two-thirds

Since March 2026, App Store Connect supports 50 metadata languages instead of 39, nearly all of them aimed at India. Nine of the eleven new languages use scripts that Apple’s byte-based keyword field charges three times the price for, and nobody warns you before you hit the limit.

What Apple actually changed

On 31 March 2026, Apple announced that App Store Connect now supports metadata localization in 11 additional languages: Bangla, Gujarati, Kannada, Malayalam, Marathi, Odia, Punjabi, Slovenian, Tamil, Telugu and Urdu. That takes the total from 39 to 50 supported languages and locales for app names, subtitles, descriptions, screenshots and the rest of a product page.

Ten of the eleven are official languages of India. The eleventh, Slovenian, is the odd one out, an Eastern European addition riding along with an update that is otherwise unmistakably aimed at one market. Localized metadata in the new languages can be added starting with a developer’s next version submission.

Why now: India is the largest app download market on earth

India accounted for over 25.5 billion app downloads across iOS and Android in 2025, more than double the 12.6 billion recorded in the United States, according to Sensor Tower data reported by Business Standard. Indian users also spent over 1.2 trillion hours in mobile apps that year.

The catch is monetization. Sensor Tower put India’s in-app purchase revenue at roughly $0.03 per download, against over $0.20 in Southeast Asia and Latin America. The United States generated nearly $60 billion in IAP revenue in 2025 on about half of India’s download volume. Reach and revenue are pulling in opposite directions in this market, and a localization decision here is closer to a brand and distribution bet than a near-term revenue one.

The catch: Apple’s keyword field is priced in bytes, not characters

Apple’s hidden 100-character keyword field is, in practice, a 100-byte field. Apple does not spell this out plainly in its documentation, but it shows up the moment you type into the field in a language whose script needs more than one byte per letter, which is most of what just got added.

Plain English and Slovenian text sit in the part of Unicode that UTF-8 encodes as one byte per character, the same as always. Urdu, written in Arabic script, needs two bytes per character. Bangla, Gujarati, Kannada, Malayalam, Marathi, Odia, Punjabi, Tamil and Telugu are all Brahmic scripts whose letters sit further along in Unicode, where UTF-8 spends three bytes on every single character.

What this means if you are localizing into one of these

How Google Play compares

Play has no hidden keyword field. Everything indexed there, the title, short description and full description, is text a real user reads, so there is no separate byte-priced field to run out of space in. Play has also offered store-listing translation in well over 100 languages for years, so this specific squeeze, a store abruptly adding a script that costs more per character, is an App Store problem rather than one shared across both stores.

Blog · Free ASO tools · Products · AsoTheory