Armando Braun

The thing I don't get about Duolingo has nothing to do with AI. The market sees a stock that's down or thinks there's some tenuous AI risk, and that drove the sell-off (imo). I don't think AI apps or AI translation are the real threat. My concern is simply why growth has slowed so much. For context, they guided bookings to about 20%, or half of prior revenue growth. CEO Luis von Ahn says it was purposeful: 1) curtailing "unhinged" marketing and 2) focusing more on product development and teaching efficiency. The first reason suggests their marketing was problematic for the brand long-term or had stopped working as it once did. I don't fully understand this, but I don't see how it's positive. The second should be a cost pressure, and it's unclear why it would impact topline—unless it ties to the energy issue (explained below). My suspicion is they noticed something in engagement numbers that signaled longer-term churn risk, so they focused more on engagement. The energy system they introduced (which penalizes you even for correct answers) seems to have hurt them more than appreciated. It limited time spent on the app, hoping to push users to higher-priced tiers with unlimited energy. Some users report the energy system reverting to the old heart system, suggesting this change was an issue. You can spin that as positive—it's a reversible decision if it was problematic. Or it's proof that users' willingness to spend is fairly limited. Duolingo Wrapped data showed the top 10% of users spend just 2–2.5 minutes a day on the app. This shows the vast majority aren't deeply attached. Increasing engagement and deploying AI for more A/B tests, features, and courses could change this. But it appears the app as it stands has a value prop problem, which is the real reason they guided bookings down. They are still growing, though. And with the stock down ~80% since May, it materially resets the assumptions an investor needs to make. They now trade at 20x trailing FCF (loaded for SBC). This valuation only makes sense if they continue on a trajectory of growth—but slowing growth isn't out of the question. With 135M MAUs and simple churn assumptions, it's possible they've burned through over 500M users who downloaded the app, tried it, and left. These users aren't likely to redownload just because they see an "unhinged" marketing post—which may explain why they curtailed that marketing (at least in Western markets). They now need a new message to win users back. Doesn't mean they can't do it, but the old formula is changing. The stock sell off may be attributed to AI, but I think their set back is more prosiac--growing at high rates becomes harder the larger you get. And markets can become saturated. But again, at 20x, you need to assume far less in order to make the math work. $DUOL