Welcome to the eighteenth edition of Black Box. Last time, I tried to end the Reddit blackout. This time, I walk through a failure of my own.
Longtime readers (thank you, if any exist) know that I usually write about a thesis that I believe reflects how tech or people actually works. Of course, I come up with a lot more wrong ideas than right ones, and I wanted to share this one because I think the economic intuition behind it is interesting and it highlights the gap between theory and practice which is always my point of contention as a thesis-driven investor. (Putting together a thesis is also a lot of work, so if it turns out to be false the least it can do is serve as content for a post!)
Late last year, a startup called Zette caught my eye. Its product is a browser extension that unlocks paywalled news content on a per article basis using prepaid credits. I signed up immediately as I ran into paywalls constantly while doing research and it wasn’t practical to subscribe to dozens of media outlets. This was a problem that many folks have, and I suspected that, like me, they bounced the second they were asked to pay up. Indeed, less than 0.5% of people convert at a paywall. Even among those who entered their payment information, nearly 10% churn within a day— probably cancelling after getting the article they needed.
The idea of a universal subscription, of which Zette is for news, is not new. The family of startups based on this idea include the infamous MoviePass and the equally controversial ClassPass. All of them take advantage of the fact that club goods are non-rivalrous, meaning they can offer the product or service in question to an arbitrary number of people (many or few) for the same cost. This allows movie theaters and gyms to maximize profit by trading off price for quantity, which is attractive since both are really busy on nights and weekends but well below capacity all other times.
Students of economics may recognize the smoothing of this lumpiness as moving towards Pareto efficiency, the dream state of resource allocation. But what is good for society is not necessarily good for business, as the two Passes showed. Before I brought Zette to our investment committee, I had to understand whether it could be good for both.
The key for me was Zette was expanding access to a non-physical product and therefore could scale infinitely at effectively zero marginal cost. This meant that any new users were pure upside, which also spared Zette from the unit economics and cannibalizing that drove MoviePass and ClassPass into the ground. And I believed news sites would gain new users due to the Jevons effect as people consumed more content on outlets from which they previously bounced because of paywalls. This doubled as a credible reason for platforms to partner with Zette.
To be explicit, three things could happen when a person starts to use Zette from the perspective of the news companies:
A full-time subscriber downgrades to a part-time subscriber
A non-user converts to a part-time subscriber
A part-time subscriber upgrades to a full-time subscriber
Investing in Zette is equivalent to a bet that the incremental gain from the latter two would outweigh than the incremental loss from the first. Given the abysmal paywall conversion rate, I felt it was worth taking.
In hindsight, there were two things that I missed. First, as pointed out by our investment committee, is that news is increasingly commoditized and free thanks to the platform previously known as Twitter. I questioned this objection then as many in the room subscribed to The Information, but I see what they mean now as a slightly more active Twitter user (for my sanity, I use X mostly to source companies). The second is the pair of statistics that I cite above are symptomatic of a Goldilocks Problem. People either rely on a media outlet enough to subscribe or don’t; if almost everyone bounces at a paywall, then no one article is sufficient to convert them. The tenth who do pay churn right after because they do not expect to return, which is exactly the point of subscribing. There are simply not many “just right” users.
Although we didn’t invest, I chalked these issues up to Zette’s specifics and felt the logic behind digital universal subscriptions was still sound. So I was excited to come across Keye a few months later. The product is essentially a generalization of Zette — it gives access to categories beyond news including data (e.g., Crunchbase), productivity tools (e.g., Loom), and education (e.g., Codecademy). I thought the addition of tools was particularly intriguing as companies try to consolidate their SaaS providers, which have ballooned to an average of over 250 applications. Perhaps Keye could enable companies to keep their current setup at a lower price or optimize their stack without worrying about whether their volume of usage on a given software justified subscribing to it.
Upon further diligence, I concluded Keye is not that solution. The common theme among Keye’s partners is their usage is self-contained, whether it be exporting data from Crunchbase or recording a Loom. Offering integrated tools like Calendly or Zendesk would require Keye to extract a tremendous amount of data from the user and populate them into a provider such that the software would work on demand despite being “disconnected” the rest of the time. I wasn’t sure if this was technically possible, and if it were, the economics probably wouldn’t make sense for Keye’s business model. And speaking of business models, SaaS companies are already moving towards usage-based pricing, so Keye’s value would diminish over time. I would be better off looking into SaaS management platforms.
With that, I dismissed the idea of universal subscriptions in both atoms and bits. My thesis was wrong, but I convinced myself that startups of this class don’t work in a way that I think makes sense instead of relying on the hand-wavy “pattern matching” that VCs often cite. For me, that’s the real value of forming and testing theses. The best investors are not purely truth-seekers, but they do understand things that others don’t. Thanks to Zette and Keye, they now have one fewer leg up on me. ∎
What is something you thought made sense but turned out to be false? I’d love to hear about it at @jwang_18 or via LinkedIn.