My side project gets more downloads from Brazil and Japan than from its home market

I built a small mental training app as a side project. German-only, barely any downloads. Nothing crazy.

I localized it into 29 languages to test a tool I was building. Didn't change anything else — no marketing, no new features. Now over 70% of downloads come from outside Germany. Brazil, Japan, China, USA, France, Poland — 25+ countries. For a side project with zero marketing budget, that's wild.

AI translation is production-ready — but not out of the box. Just throwing your strings into ChatGPT doesn't cut it. You'll get broken placeholders, wrong plural forms, translations too long for your UI, words translated without context. Getting consistently good results takes real engineering underneath — context injection, format-aware parsing, constraint handling. That's why I built shipglobal.dev, originally just to solve this for myself. Right now it handles .strings, .xml, .json and other i18n files — upload, translate, download. Costs next to nothing compared to professional translation.

App Store localization was the biggest lever. People search in their own language. If your listing is English-only, you're invisible. I did this part manually for now — it's the next thing I want to automate.

You don't need scale to benefit. Localization isn't something you do once you're big. It's how you get reach when you don't have a marketing budget.
What's holding you back from localizing? Cost? Quality concerns? Maintenance? Would love to hear from you.