Duolingo: Speaking Growth Fluently, Despite the AI Noise

This article first appeared on GuruFocus.

The technology sector is one of the most interesting in my view, and at the same time, it offers the prospect of strong growth and the creation of solid moats, but it also carries high risks of disruption due to its dynamic nature. The case of Duolingo (NASDAQ:DUOL) is somewhat mixed, the company is growing rapidly, but uncertainties continue to grow along with its KPIs.

Apple (NASDAQ:AAPL) recently announced a new feature for its AirPods, live translation, which, along with several other initiatives such as Meta Glasses, LLMs, and the like, makes the need for an app like Duolingo increasingly uncertain for the coming years.

However, I believe that at least part of this narrative is flawed. This is because this AI-related fears ends up disregarding some interesting moats that Duolingo has managed to build, in addition to the potential addressable market and expansions for the coming years. Therefore, although the uncertainties are real, I also believe that Duolingo is already approaching a more compelling valuation, which makes the compounding of its growth even more appealing.

The AI Threat For Duolingo:

First, let's not downplay the threat. The current environment shows us several real-time translation tools emerging, and they are real. Apple's Airpods with Live Translation, Meta's Smart Glasses, all of them may be targeting the casual learner, such as a tourist who just needs to order a coffee or find out where a tourist attraction is in a city where they don't speak the language. This segment of people who are using Duolingo exclusively for this purpose is indeed more vulnerable. If your sole goal is transactional communication, these AI tools end up being a more practical and faster choice than spending a week or two learning basic words. This could slightly reduce demand at the top of Duolingo's funnel.

Still, this argument misses the fundamental motivation of an engaged audience, which is Duolingo's core.

Translations serve a purpose, but it's a completely different need. If you just want to visit a few foreign tourist spots, these tools might help. But if you really want to engage with local people, experience the culture, whether for pleasure or business, speaking the language and understanding it is very different. It is a skill that is desired and appreciated by many people. No one spends weeks learning a language just to read signs or menus; in fact, it is to increase their connection with the culture, waiter, client, or a potential partner. It is a desire for genuine human connection, self-improvement, career advancement, and the like. That is Duolingo's real moat.

Story Continues

Furthermore, I don't think the Duolingo app should be seen as a tool, but rather as an experience. It is an educational app, but it also has gamified features, features similar to social networks, and even entertainment.

Duolingo's management knows that AI is a powerful trend, and therefore already embraces initiatives and includes features such as Duolingo Max, which has advanced AI features, like Video Call, which ends up being more efficient and simpler than a generic LLM.

Duolingo: So Much Untapped Potential

Despite the section above, I believe that the main point of the thesis is not its ability to defend AI or monetize these AI features, but rather the untapped potential of the user base. It is a case that in some ways resembles the early days of Netflix (NFLX), a company that has a massive TAM and also has the potential to better monetize its subscribers/users.

The Q2 results show this enormous potential. DAUs grew 40%, reaching 47.7 million. MAUs grew 24%, reaching 128.3 million, and paid subscribers grew 37%, reaching 10.9 million.

Let's take some hard numbers to verify this monetization opportunity. With 10.9 million paid subscribers and around 128 million MAUs, the current pay rate is 8.5%. The company already sees 6% YoY growth for ARPU due to a better mix of higher prices (such as Family and Max plans). A relatively simple analysis already shows the potential of this conversion. For every 100 basis points that the pay rate increases (such as from 8.5% to 9.5%), more than 1 million new subscribers could be added. At current ARPU, this means a very interesting increase in revenue, and that with a high margin. This dynamic is further enhanced by some very interesting tailwinds, such as the ability to bypass App Store fees through the web, which can further boost the profitability of new subscriptions.

Of nearly 130 million users, only 10.9 million are paid subscribers, meaning there is a huge base to be monetized. So, it is not crazy to think that this KPI could double or even triple in a few years, since even considering only DAUs, which are the most engaged, if penetration is 50%, we are talking about nearly 24 million subscribers.

Considering the total addressable market, there are currently more than 1.5 billion people learning a foreign language. There are trends such as increased travel, specialization, immigration, and multinational companies, which can further drive the need to learn a second language.

Another interesting fact is that only 1.5 billion people speak English in the world, while another billion speak Chinese. Think about how much potential there is for both numbers to increase, since China influences many countries around it, while the US influences several countries around the world, meaning there is room for increased penetration of these languages around the world.

Many people perceive that India has many English speakers, but even though the number is close to 130 million, that is not even 10% of the population.

Not to mention other initiatives, such as music and chess education. For example, chess.com has over 220 million members.

Duolingo's Valuation and the Power of Max

The company's financials are growing rapidly, with revenue reaching $250 million in the quarter and adjusted EBITDA reaching $78 million, following impressive growth.

Annualizing this revenue, we already have something close to $1 billion. But note that if the company increases, for example, by 20 million subscribers over the next 10 years, that already adds more than $6 billion in annual revenue (considering an average ticket of $25/m).

The adjusted EBITDA margin is already 31%, but with scale, this can continue to increase much more.

Even though an increase of 20 million subscribers seems possible to me in the long term, let's assume that revenue reaches something close to $3 billion. 35% (a little above the Q2 adjusted EBITDA margin) of that is already more than $1 billion, and Duolingo is currently worth only $14 billion.

Another key driver of future profitability will be adoption of Duolingo Max. This is a premium tier, priced at a much higher price than the traditional subscription and accounts for less than 10% of subscribers. Let's model the impact.

Duolingo Max costs around $29.99 per month, almost $350 per year. A 15% penetration in MAUs means around 19 million subscribers in this tier, which alone could be around $6.6 billion in revenue. With a 35% margin, that would already be more than $2 billion alone.

In a more bullish scenario, where the company achieves 20% penetration and a 40% margin, that gives $8.9 billion in revenue and $3.5 billion in EBITDA for this category alone.

In other words, it seems to me that there are several scenarios that point to high asymmetry if the company manages to unlock all this potential.

Things Can Change For Duolingo, But It Has Some Moats

Obviously, there are risks to this growth story. There is a degree of cyclicality in users, which means that management must be able to keep users engaged and learning more than one language or, after learning a language, start using the Chess features, and so on, so that users maintain their subscription and/or remain engaged with the app.

Another point is competition, since AI is likely to be a trend that greatly lowers the barrier to entry.

But these uncertainties are mitigated by the moat, such as the strong brand power that the company has already built, the gamified experience, and the initiatives to diversify revenue.

Gurus

When we look at the Gurus, there is nothing too surprising, but some movements are curious. Although Ron Baron (Trades, Portfolio) (Trades, Portfolio) is the largest shareholder among them, he owns only 0.08% of the assets managed. Despite this, moves in Q2 such as those by Jefferies Group (Trades, Portfolio) (Trades, Portfolio) and Joel Greenblatt (Trades, Portfolio) (Trades, Portfolio) , who added significant exposure to Duolingo (adding 117% and 270% respectively), allow for greater conviction in the company's momentum.Duolingo: Speaking Growth Fluently, Despite the AI Noise

GuruFocus

The Bottom Line

In short, I believe that fears of AI disruption and uncertainties about Duolingo's model, while real, also seem overly exaggerated, as the company has already managed to build a profitable business model with some moats and interesting prospects.

The good part is that this market fear makes the valuation very asymmetric if the company succeeds in continuing to grow and monetize its base.

View Comments