Scaling Duolingo, embracing failure, and insight into Latin America’s tech scene | Gina Gotthilf

Gotthilf warns that teams working at computers can overlook how their page feels on a phone. The first check is a usable mobile experience. Put the central message and intended action where they are easy to find rather than expecting a visitor to work through the page.

She expects people to skim rather than read every word. A headline and button that make sense together can communicate the offer with fewer steps. Review whether someone seeing only those pieces would know what the action means.

Gotthilf treats retention as a signal of whether the product is useful to the audience it reaches. Acquisition alone can hide a mismatch if people do not stay. This is her growth perspective, not a claim that retention is the only metric every business should optimize.

Her badge experiment gave people a reward for signing up, without the achievement, collection or sharing that could make badges meaningful. A flat result did not adequately test the fuller idea. Define which mechanism an experiment preserves before interpreting its result.

She says the growth team initially wrote hypotheses and prioritized tests without enough dogfooding. Actually using the experiment would have made its weak badge experience more obvious. A well-written plan does not replace checking what a customer will see and do.

Gotthilf describes asking whether copy could belong to any company or whether it feels like Duolingo. That makes voice a repeatable review question rather than a vague request for personality. Choose a tone suitable for the audience and preserve clear meaning while making it distinctive.

She describes sharing learnings so others can avoid repeating mistakes, and notes that successful stories are easier to remember than failed efforts. Make failure review part of the record. This helps a team keep the context behind decisions instead of only retelling the wins.