The keyboard everyone already knows

The socio’s instinct was immersion: use the keyboard people actually use in the target region, AZERTY for France and Belgium, so the experience feels native. The result was a keyboard where typing “hola” produces “Hulk” ▸ 8:33, and the reversal arrives from the numbers: some 88% of the world types on QWERTY ▸ 6:18, so an authentic layout would fight the very fluency the keyboard study set out to add.

The evidence is a self-experiment Juan already ran. Studying Chinese, his phone once switched to a predictive numeric keyboard, press a digit, cycle to a character ▸ 6:36, and he tried to keep going with it to see if the immersion paid off. It didn’t: it was “otro peso cognitivo que no era necesario” ▸ 9:00. The failure mode is quantified: his lessons shrank to one a session because typing took so long ▸ 9:16, and then he stopped, reverting to the QWERTY he’d used his whole life ▸ 9:33.

no gastes voluntad enseñando un teclado →

The principle underneath is a budget, not a preference: learning a language already spends willpower, and forcing the learner to also learn a keyboard withdraws from the same scarce account until the habit itself goes bankrupt. So Severo ships one QWERTY for everyone ▸ 9:37, with the language-specific characters, the ñ, the accents, the German ß, reachable by long-press rather than by relearning where every key lives ▸ 10:00. The authentic choice would have been more faithful to France and less faithful to the only thing that matters, whether the user comes back tomorrow…