My Two New Big Bets: Hims and Duolingo.
Update
Edited by Brian Birnbaum and an update of my original Hims deep dive and Duolingo deep dive.
Hims and Duolingo are Singularity Scalers: companies likely to grow by orders of magnitude as the AI revolution continues. If the AI revolution doesn’t play out, they still get much bigger.
As I announced on X last week, I’ve started a position in Duolingo. Duolingo now joins Hims in the next-generation segment of my portfolio, which I’ve constructed in accordance with the mental framework I teach in my Tech Stock Goldmine course. I believe these two companies are likely to grow orders of magnitude larger over the coming decades as their free cash flow per share explodes exponentially. Both companies operate without true competition, extraordinary management teams, and an infinite runway with two layers of asymmetry.
When I say they have no competition, I don’t ignore that there are other companies trying to do what they do. But as the Spotify case has taught us, in the digital space a marginal and initially imperceptible user experience advantage translates into an exponential advantage and adoption over the long term. Duolingo and Hims offer a better user experience than similar companies, which they continue to compound at an unmatched rate. Their moats will soon be insurmountable, if they aren’t already.
Both business models boil down to acquiring more subscribers and making more money per subscriber. The discussion around tariffs, policy, and other uncontrollable factors may be weighing on these two stocks–it’s difficult to say for sure during widespread forced selling–but they have zero impact on the evolution of their fundamental value creation process.
What matters is that the Hims and Duolingo management teams have demonstrated a world-class ability to delight and acquire more end consumers. They’ve also demonstrated the ability to translate that growth into higher free cash flow per share, as you can see in the two graphs below. Going forward, my impression is that both management teams can be trusted to continue growing their subscriber bases and print more free cash flow per share.
Duolingo’s and Hims’ asymmetrical opportunity contains two layers. First, there’s plenty of upside left ahead of them simply by continuing to grow the subscriber base and increasing revenue per subscriber. Second, they can accelerate that process dramatically if the AI revolution lives up to its potential. In such a case, Hims transforms from a direct-to-consumer healthcare platform to an AI doctor and healthcare platform. Duolingo goes from being an education platform to an AI teacher and education platform. The externalities of both events are infinite and promise to position both companies as operating systems for their respective industries.
This can play out along a spectrum. In effect, I believe both Hims and Duolingo are Singularity Scalers: a breed of company that is positioned to grow their earning power exponentially and at a marginal cost, in tandem with the evolution of AI models. These companies, day after day, relentlessly aggregate proprietary datasets that enable them to create AI models that others simply can’t. The value created by these datasets is growing exponentially as AI continues to improve, with marginal investments on behalf of these companies.
With every linear step forward, the odds of exponential success increase in kind. Every time Hims deploys a new vertical, they aggregate more data that enables them to understand how to best treat a condition for individuals with a broad range of characteristics. Every time Duolingo launches a new course, they gather similar signals as to what, when, how, and why students are learning any given set of material in a manner they can then optimize. Having more and better data enables them to stay ahead of the competition and gives them the option of training an AI that others can’t, furthering the first agenda in order to generate outsized value.
Hims and Duolingo have two slightly different risk/reward ratios: Hims is greatly undervalued trading at just over 5 times sales, while Duolingo is likely fairly valued at just over 18 times sales. The jury is still out on whether Hims has a moat, and the jury (market participants) likely won’t be in until the judge (the market itself) has decided for us. This is precisely what happened with Amazon, Spotify, and other network-defined juggernauts. These moats are highly abstract, largely intangible, and requiring imagination to understand, much more foresee.
That being said, if you believe a management team can be a moat in its own right, they may already have one. A moat is an elastic concept. Network- and data-defined moats are built when these things reach critical mass. When critical mass is assured, the moat essentially exists. Thus the presence of the moat becomes a debate over whether and when it's assured. With the current management team in place, we are, in my view, nearing or perhaps already beyond this event horizon.
Hims’ moat purgatory goes to explain much of its relatively compressed valuation. Per my last writeup, I believe what Hims is doing is sufficiently difficult enough that Amazon and Walmart have already failed in attempting to replicate it. Hims is building a moat akin to Spotify’s, with the added feature of personalized treatments at scale, as depicted below.
Both companies have one characteristic in common: they are becoming the top of the funnel for their respective industries by fundamentally resetting consumer expectations. This is a dynamic that we’ve seen play out time and time again over the past two decades. Top of the funnel companies start as toys that no one takes seriously and they suddenly rise to dominate industries. At the beginning no one thought movies could debut on Netflix or that Spotify could become the lifeblood of music labels, but here we are.
Both Hims and Duolingo are compounding tremendous customer goodwill by delivering much more value per dollar spent on behalf of customers than the rest of their respective industries. Both are still operating across a small number of verticals, but are branching out rapidly. Hims is launching new healthcare verticals at an accelerated pace and Duolingo recently launched the math and music verticals. As they continue to launch more verticals, I believe they will reset consumer expectations across the board.
The most impressive example of this sort of dynamic is Amazon. They provide such exceptional convenience, low prices, and wide selection to customers that a growing volume of merchants simply have no option but to default to the platform. In the event that Hims should prevail as depicted across my writeups, healthcare providers will likely have to operate through Hims to be competitive. As for Duolingo, teachers likely won’t be needed once/if Agentic and Physical AI models enable them to turn Tesla’s humanoid robots into physical teachers.
At their axiomatic core, healthcare and education industries optimize and turn information into actions that generate the best possible patient and student outcomes. It makes sense that, if you own the relationship with the end customer and achieve critical mass, you ultimately end up capturing the industry’s most treasured data. With today’s technology, this puts you in pole position to process that data and ultimately sell the intelligence that the rest of the market needs to make good decisions.
These two investments are an additional test of the mental framework that I teach in my course. The framework has so far led to my 20X+ AMD and Tesla picks over the last decade and my 11X+ Palantir and 5X+ Spotify picks over the last few years. What’s interesting is that Hims and Duolingo are smaller companies. It’s therefore a test of how predictive the mental framework is in earlier stages, where much higher returns can be achieved. Palantir and Spotify were already an intersection of venture capital and public markets investing, but Hims and Duolingo lean far more heavily into the realm of venture.
It’s taken me a decade of intense work to hone my ability to spot extraordinary companies early, but you can get a head start by taking my course. In under two hours–I teach you everything I know in an elegant and powerful mental model that over 450 students have happily integrated to date.
Until next time!
⚡ If you enjoyed the post, please feel free to share with friends, drop a like and leave me a comment.
You can also reach me at:
Twitter: @alc2022
LinkedIn: antoniolinaresc