App Store Localization Rejected: It's Usually One of These
App Store localization rejections usually trace back to a few metadata mistakes; Google Play follows different rules. Here is how to check both.
It usually goes like this. Your main listing has been sailing through review for months. One day you decide to push the app into Japan, so you add a Japanese title, subtitle, and keyword field, keep the English screenshots, and hit submit. Monday morning there is a rejection note citing metadata, and you cannot tell which part actually failed.
Localization rejections are a wall almost every team going overseas runs into. The rejection message is usually vague, but the causes behind it repeat. This piece walks through the common failure points on both stores — App Store and Google Play do localization differently, and plenty of teams fill in Google's forms using Apple's playbook, then cannot figure out what went wrong.
The short version
When App Store localization gets rejected, the cause is usually one of two things: metadata that does not match the actual content, or keyword field abuse. On Google Play it is more often an incomplete default-language store listing, or translations that no longer match what the app does.
The fix is the same idea on both stores: treat every locale as a standalone product and check it as one. Title, subtitle, keywords, screenshots, description — they all have to be coherent in the same language. Now the details.
Where App Store submissions usually get stuck
The first repeat offender: screenshots and preview videos that do not match the listing language. You translated the title and description into Japanese, but the screenshots still show an English interface. To review, that reads as content shown to Japanese users that does not match reality. Screenshots are part of your metadata — localization is not just the text fields.
The second: keyword fields stuffed with other apps' names, competitor brand terms, or strings of words glued together with symbols. The keyword field is per-locale, and some teams assume a new language means a fresh chance to stuff keywords. It does not. The field exists to describe your app, not to borrow someone else's traffic.
The third: new-language descriptions containing claims you cannot back — "number one in the market", prize draws, or prices, awards, and press quotes irrelevant to that market. Different markets apply different standards here, and these get flagged.
For exactly which wording triggers rejection, check the current official guidance before you ship.
Google Play is not the same system
Many teams assume Google Play localization is just a translated copy of the App Store setup. It is not, and the mechanics differ.
Google Play store listings have a default language, and that default listing must be complete. Everything else is layered on top of it. If the default language is missing a description or screenshots, adding ten more languages will not get you through.
Another common trap: each language's translation is entered separately in Play Console, often by different people on the team. The German description ends up describing a feature you shipped three versions ago. Google does not tolerate description-reality mismatches well, and text with obvious raw machine-translation artifacts can get flagged too.
One more distinction people miss: custom store listings aimed at specific audiences or regions, and localized listings aimed at languages, are two different tools. Using the wrong one for the job is a classic beginner detour. For field definitions, check the current official guidance in the Play Console help documentation before you ship.
Machine translation is the biggest trap
The single most common root cause I see behind rejections comes down to two words: raw machine translation.
Machine translation itself is not against the rules — nobody forbids you from starting there. The problem is what unreviewed machine text does. It translates things that should never be translated, like product names, button labels, and model numbers, turning them into something odd in the target language. And it reads stiff enough that a reviewer can tell in seconds this team does not actually serve this market.
My rule: machine translation can produce the first draft, but a native speaker of the target language must review it before release. If you cannot find one, at minimum lock down the words that should stay untranslated — brand names, feature names, UI labels. This step cannot be skipped. The cost you save comes back as rejections and rework.
Should you redo keywords per language
Yes, and this is the highest-leverage part of localization work.
The App Store keyword field works per locale. A keyword that proved itself in English, translated literally into Japanese, is usually not what Japanese users actually type — search habits, autocomplete behavior, and input methods all differ. Running keyword research fresh for each language beats pasting machine-translated English terms every time.
Google Play has no keyword field. Its search signals come mostly from the text in your title, short description, and full description. So "keyword localization" on Google Play is really copywriting localization: write the phrases your target-language users search for, naturally, into the description. Keyword stuffing does nothing there and can be flagged as spam.
Screenshots and preview videos count as metadata
Screenshots are the most neglected piece. Both stores let you upload screenshots per language, but many teams consider the workload too heavy and ship one English set worldwide.
That does not automatically get rejected — if your interface is genuinely in English, English screenshots under any language's listing are usually defensible. But if you claim to support a language and your screenshots show a different one, the rejection risk on the App Store climbs noticeably. On Google Play you may pass review and still lose conversions, because users who cannot read the screenshots do not download.
Preview videos follow the same logic. If the captions or voiceover language does not match the listing, that is another flavor of metadata mismatch.
When you should not localize yet
The other side of the tutorial. Three situations where I would tell you to wait:
First, the product has not stabilized in its main market yet. Localization is an amplifier. If the numbers at home are weak, translating into eight languages just copies the problem eight times.
Second, you have no plan to maintain multiple languages. The listing gets translated, the product iterates through ten versions in a quarter, and every translation goes stale. That hurts more than not localizing at all — users download based on an outdated description, find something different, and leave one-star reviews.
Third, the budget covers "translating once" but not "translating well". In that case, do two or three priority markets deeply instead of thirty languages of raw machine output. Breadth does not convert. Depth does.
How we prioritize markets and languages
My ordering method: market first, language second, field last.
Market first: whichever country or region is already sending you organic users goes to the front of the queue. The regional data in your dashboards has the answer — no guessing required.
Language second: one language often covers several markets, but the same language reads differently across them. Priority markets deserve separate proofreading.
Field last: within one language, title and subtitle carry more weight than the long description, and screenshots move conversion more than copy. When resources are tight, fill the gaps in that order.
What to do next
1. Open App Store Connect and Play Console and audit every locale — title, subtitle, keywords, screenshots, description — for coherence within the same language. List the mismatches first. 2. Redo the keyword field per language instead of reusing machine-translated English terms; on Google Play, write the way your target users actually search into the description. 3. Get a native speaker to review machine-translated text, and lock down brand names, feature names, and UI labels that should stay untranslated. 4. If you are unsure which market to start with, pull downloads and conversion broken down by country, and localize the two or three markets already bringing organic traffic.
Localization rewards pacing more than coverage. If you are planning a specific market push or stuck on a rejection right now, bring the details — we read these forms every day and can help you locate the problem first.
Want your app to rank for the keywords that matter?
Get a Quote