Scaling Duolingo, embracing failure, and insight into Latin America's tech scene

We constantly looked at every copy we wrote and asked, 'Could this have been written by other companies? Could this have been any other company or is it Duolingo? What makes this Duolingo?'

Why do I know this, but not this entrepreneur who honestly is smarter and more ambitious than me? Oh, it's because I've had a career in Silicon Valley. It's because I've gotten to work with some of the best people in the world and I have access to this information just free flowing in my way all the time, and we take it for granted.

There is always a quirk, like it's unexpected. The way we talk to you is a little bit funny. It doesn't take ourselves too seriously and it makes the person receiving this message feel something. Again, it's about how you make people feel.

I had one learning there that has stuck with me, and I think can work for other people too, is that it's not just about doing things that actually matter, and learning. It's about being able to tell the story, and it's about understanding what other people perceive as valuable.

If as a startup, whether it be early stage, but as a startup you focus too much on those marginal differences between groups of people, you can run the risk of making too many changes too soon and learning very little and adding crazy amounts of code complexity and overall organizational complexity to what you're building.

You build this thing, you made it, and then you tried to convince people to buy it. That was it. It's no longer at all that because products are constantly in motion and evolving. Unless you're able to own that data and understand what it can tell you, you're never going to win.

I had one learning there that has stuck with me, and I think can work for other people too, is that it's not just about doing things that actually matter, and learning. It's about being able to tell the story, and it's about understanding what other people perceive as valuable.

Fake it till you make it. I don't mean lie. I don't mean pretend to be something that you're not, but it's almost like a way to overcome imposter syndrome and by opening rooms for yourself to be who you want to become so that you enable opportunities to become that.

Retention, I don't think of it in terms of like, wow, I must retain this user. It's like, is this thing valuable or not? That's what retention is to me. Either it's actually providing real value or it's not. If it's providing real value, people stick around. It's as simple as that.

I didn't let the team run this experiment for six months so that we focus on lower hanging fruit. We ran this very simple experiment that was like, 'You sign up and then you get a badge.' And of course, in retrospect, it led to no results, because no one is proud of signing up.

Humans are very similar and we think we're very different. I think that maybe we were brought up differently, and so for example, in Brazil we like to dance. In Mexico, we like to dance salsa, in some places it's okay to hug and other places it's not okay to hug. Yes, there are cultural differences and it's important to love everyone and to adapt, et cetera, but a lot of these differences when it comes to core human behavior are that final 5% of understanding people.

Even if you have a low CAC, I was just talking to an entrepreneur this morning, he was like, 'Oh, we were able to hit a 10-cent CAC.' And I'm like, 'Yeah, but that grows.' As you start targeting more people, the low hanging fruit dry out, and you end up having to pay more money.

If people only read the title and the button, they got it. It can't be like, 'Wouldn't you like the next President of the United States to blah blah blah?' And then the button is like, 'apply.' It's like, apply for what? Now I have to go read the thing and I didn't get it.

My mandate was make Duolingo grow. Here is no budget. I had no budget. People are shocked. They're like, 'Oh, but when did you turn on the ads?' We're like, 'We just didn't.'