Drew Cohen (21:16)
So this is, you know, a few weeks back and I've talked about Duolingo before, but basically I wasn't fully understanding the reasons why they said growth was going to decelerate. So if we go back a couple quarters to like that first sell off or I guess the second sell off, I don't know which sell off it was. There was a lot of sell offs, apologies to holders, but I think this was around maybe October, something like that. This sell off was, you know, they basically decelerated growth expectations and so they're growing 30 to 40% and then they guided bookings to 20%. And the reasons they gave at the time was they said we're going to do less like unhinged viral marketing and we're also going to focus more on improving user engagement and the experience instead of monetization. I didn't understand why those things would lead to a growth deceleration that much. And so the on the marketing thing, I didn't really understand why the unhinged marketing thing was really a problem for their brand. And if it was a problem for the brand, how come they didn't realize that for like the many years that they're already doing it? I think the more likely thing that ended up happening was the unhinged marketing stopped working. And you know this because by the way, they stopped doing it only in western countries and they're still doing it in other countries that they were newer in. And so it was probably more novel there. I think what was happening was basically there's this issue when you are looking at a business's existing users and they don't have churn statistics. You don't actually know how many users have tried their product and left so if Duolingo, you know, and they say they have 135 million users, that's very possible that they've really had over half a billion people that have used the app. And so that means a vast majority of them have already tried the app and they're not interested in it. So if you hit them with another unhinged, you know, marketing ad that's not going to work on them again because they're not going to be as impulsive and they already tried the app, so now you need to change your marketing to them in order to win them back. So I think that was the first thing that was really happening with the marketing was the reason why it wasn't working as much was because a lot of users have already tried their app, they know what Duolingo is about and the value prop for them wasn't there. So if you're going to win it back, you had to change the messaging. And so I think that was the first aspect of there, the thing that didn't make sense when they talked about it. Then the second one was this kind of idea that we needed to increase engagement. And so that was kind of a weird thing because I don't understand why that wouldn't have always been a focus. Because they're talking about, oh, no, we pushed too much on monetization. You know, we had too many ads, maybe. I think there was a thing where they rolled out energy instead of hearts, where energy penalized you for basically, even if you got answers correct, it would start to run out. And once your energy was out, it booted you out of the app versus hearts, which was it only penalized you if you got something wrong. And so the idea behind, you know, this energy system basically running out was it's trying to push people to the paid tiers. That did kind of make sense to me in terms of that, you know, potentially turning users off. But then the flip side of that is it's also evidence at the same time that there aren't more people that really want to pay for your app. And so that one's kind of mixed. I mean, maybe if they stayed on it long enough, they could be engaged in all that. But on top of that, because they did something called Duolingo Wrapped, where they basically looked at, you know, what Spotify Wrapped was doing, where they, you know, tell you what your most listened to artists are and all that. And there's people that back engineered the data from this Duolingo Wrapped and it showed that, you know, the top 10% of users are spending just a couple minutes a day on the app, two to two and a half minutes a day. And so that's not long enough to of course learn a language. And what this really is, is it's kind of showing that the habit that they built, which is a good thing, especially with the score streaks. But it's a lot of people just getting a lot of notifications, checking in, doing just the bare minimum in order to keep the score streak and moving on. And so on the one hand, like, that's great, they've been able to engineer an app that has got people to open up every single day, spend a couple minutes on it, that's good. But if you're thinking about the actual engagement and what you're really trying to sell, that's not that high of a value added to consumers. And so it makes sense why there's a good portion of them that aren't willing to pay for the app because that's not providing a lot of value to them. You know, this random two minute lessons you're clicking through, getting a score streak. And yes, there are people that use it a lot more and a lot more religiously. But to me, this was kind of speaking to the bigger issues of the business, which is that, that this product is not something people are really using to actually learn and educate themselves. Which is fine, it could be a game and all that, but I don't think as a game it's that fun. And so I think it's just kind of not the strongest value prop for customers. And that is why you're seeing them now once again guide bookings down because now they're, they're really trying to focus on increasing engagement. And I think the thing that happened was a mix between the fact that their customer acquisition engine wasn't working as strong as it was in the past, so it couldn't cover up the fact that user Churn was really kind of higher than maybe people appreciated. And then secondarily, right now they're shifting to trying to increase engagement more to keep users on the app and reduce Churn, which is now actually causing an issue to bookings because they're not going to focus as much on some of those monetization things. I already saw that they're doing tests to reverse the energy thing to bring it back to hearts in some areas and a B testing. And so all of that's just kind of a synopsis of, of what's going on with Duolingo, my take on it and kind of some of the Issues that were overlooked when everyone was just talking about, you know, this AI risk. I think the real risk underlying it all was really engagement and then basically their user acquisition engine not working as effectively, which then in turn meant they couldn't cover up Churn as well.