Treat Every Country the Same (Mostly)
Contrary to the instinct to heavily localize per country, Gina found the opposite works: humans are far more similar than we like to believe. Duolingo treated the world as one, assuming what worked in one market would work everywhere, with rare exceptions like China and India. Over-localizing multiplies code, org, and time complexity — and time matters more than money for a startup.
- People love to believe their country is different, but core human behavior is largely the same
- Every market told her 'here it's different, this won't work' — listening would have caused problems
- Duolingo assumed what worked in one market would work everywhere and rolled out once
- Over-localizing multiplies versions, code complexity, and test time — and time beats money for a startup
“in our country we we think that green is is a negative color like owls are like poorly seen as an animal or here people…”