Cory Levy

The through-line

Almost everything Cory Levy does runs through the same conviction: that the most interesting information sits with people at the two ends of the age curve, and the way to get it is to put yourself physically next to them and ask. As a teenager he cold-called venture capitalists and pitched himself as "the teen in your office," offering the perspective of the user they were investing behind, and spent short shadowing stints at Founders Fund, Union Square Ventures and Tech Stars [3]. He then built a product for high schoolers on the premise that the founders should design what they themselves would have wanted in high school, and later went undercover into high schools with permission to watch teenagers use it [3]. The same instinct shows up in his interviewing, where he asks accomplished operators to introduce themselves twice, once as they were at 22 and once as they are now, and then works forward through the practical mechanics of how they actually run things [1].

The subject matter has moved from teenagers to founders to institutions, and the method has not. He probes Mark Pincus on taste, hiring and which conventional startup advice is wrong [1]; he presses Balaji on how success is measured at Network School and on what a diagnosis of Western decline actually means for a person living inside it [2][5]. He is candid that the public-facing part is not the appeal: "I don't really like being on podcast, but here we are," he said, before turning the microphone around and interviewing his host instead [2].

On building where teenagers actually are

After School was, in his own compressed pitch, "a private social network for teenagers" whose function was blunt: "after schools replacing Facebook for teenagers" [3]. The design logic was that teenagers needed a bounded space rather than a general one, so a student picked their school from a nearby list, verified attendance with a school ID or Facebook or one of several other mechanisms, and landed in a network consisting only of classmates [3]. Users described it back to him as "a Twitter just for your high school," and it grew from anonymous posting into a general purpose network with matching on activities and topics, a Q&A polling game, and photos [3].

He is explicit that adoption is a student decision rather than an institutional one. The company worked with students the way Facebook, Snapchat and Twitter worked with users, and treated school outreach as a separate safety and information function, including a dedicated team in San Francisco and a parents guide produced with ConnectSafely [3]. The generational argument behind the product is one he has made directly to adults, in a piece addressed to old people arguing that today's teenagers will drive enormous change and that adults need to understand them in order to teach or influence them at all [6]. His own framing of the product question was simply what he and his co-founder would have wanted when they were in high school, given that his co-founder's own high school experience had been extreme [3].

On safety as a product decision

He treats moderation as the core of the product rather than an afterthought. "Every single post goes through proactive moderation," he says, meaning nothing reaches the school community until it is approved, which he claims no other social network was doing at the time [3]. Negative posts are blocked before anyone sees them, so that in his description "there's no negativity" on the network, and what remains is mostly the relationship talk teenagers actually want, crushes and how to ask someone out [3]. The stack is deliberately hybrid: human moderation came first, running around the clock in America and other countries, with moderators trained, tested and spoken to weekly and their numbers scaling with the audience, while engineers in San Francisco built technical moderation that hands anything it cannot decide confidently back to humans [3]. Community reporting sits on top, with a one-and-done policy where a report removes the post immediately, no questions asked [3].

The second pillar is intervention. Through a partnership with Crisis Text Line, a teenager who posts that they are struggling is offered help from a trained crisis counselor at the moment they post it, a feature he says over 100,000 teens have used [3]. That commitment costs money and the app is free, which he answers with a straightforward sequencing argument: investors are supporting the product now, monetization comes later, and brands, studios and movie producers were already approaching them because their teenagers had told them this was where to advertise [3].

On iterating until something spreads on its own

He is unromantic about how long it took. He and his co-founder had been building social products since 2011, starting with One, an app connecting people around interests and introducing strangers, which they iterated on dozens and dozens of times [3]. Some college versions got real penetration in a couple of schools, but in his words "it was a lot of fighting to get those users," and that friction is what he uses to distinguish traction from noise [3]. After School itself took only three or four months from summer 2014 to a November beta, which he concedes is only meaningful on the theory that they had already built it four times before [3].

The signal he trusts is unforced spread. Testing in a handful of San Francisco high schools, the beta went from zero to over 50% of one high school inside a week, purely word of mouth [3]. Put into the App Store through Apple's beta program hoping for a few thousand testers, it reached 200,000 teenagers in three weeks and touched roughly half of US high schools, and eighteen months on it was in over 80% of them and a top ten social networking app [3]. He credits the category as much as the execution, noting that networking applications are extremely viral by construction, and he is candid about the operational cost of that, including a period where they were doubling infrastructure every two days and taking the service down to do it [3]. His curiosity about how others compress that loop is why he asks founders directly what the first thirty days of a new project look like [1].

On small teams and hiring

After School ran with five or six people at the start and about twelve or thirteen at millions of users, which he sums up as "you can do a lot with a little," adding that this would have been unlikely five years earlier and impossible ten [3]. His hiring stance follows from that: recruiting is the standard Silicon Valley challenge, and his answer is a narrow filter, "we interview tons of people and only let a few in," with an explicit preference to "hire prodigies" and very smart people, on the view that an amazing team makes everything else work [3]. He is interested enough in the opposite position to go looking for it, pressing Pincus on team building, disagreement and letting people go, and drawing out the argument that a wider hiring funnel and speed of filling a seat can beat slow, careful selection [1].

On raising money by asking for advice

His fundraising lesson is a formulation he repeats as the main takeaway from the seed round: "when you ask for money you get advice," and asking for advice is what gets you money [3]. It started with a mentor who looked at their side project, asked whether they were raising, and told them they always should be, which led to a seed round closed when he was 19 in his freshman year, after which he dropped out and moved west [3]. The company raised roughly a million in seed, a second seed, then a Series A after launch, seventeen or eighteen million in total [3]. He is equally clear about the value of not needing it, observing that they generally were not raising when the conversations happened [3]. That instinct for founder leverage runs alongside the warning he surfaces from Pincus, that in an era of abundant venture capital an investor's preference for growth over profitability may not match the founder's own interest [1].

On the questions he puts to founders

His interviewing has a consistent shape and reveals what he thinks matters. He asks people to place themselves at 22 and today so the gap becomes visible, and he treats continued hands-on building as the tell of a serious person, telling Pincus he found it "really cool that you're still working on projects and still playing" [1]. From there he goes after transferable mechanics: what someone is genuinely expert in, whether taste for a mass-market consumer hit can be taught, how you lead a team when your instinct conflicts with theirs, how you handle firing, and, pointedly, what widely repeated startup advice they think is simply wrong [1]. He is also willing to be the skeptic in the room. Confronted with a sweeping thesis about institutional decline, his follow-up was the practical one: "so what," for the ordinary person living through it [2]. Over years of conversations he has taken the same approach to a founder of Twitch, a partner at Union Square Ventures, a self-made billionaire running a restaurant and casino empire, and the founder of the Prison Entrepreneurship Program [9][12][11][4].

On mental health, honesty and second chances

The Crisis Text Line integration is not an isolated gesture; it sits with a sustained interest in the parts of ambition that go unspoken. He has given entire conversations to three young people talking openly about depression, anxiety and suicidal thoughts and how they came back from their lowest points [10], and to a psychiatrist whose clients are top writers, actors, producers and CEOs, on tools for defeating inner obstacles [7]. He has also gone to a musician's account of building a career and giving back [8], and to the work of transforming the lives of people with criminal histories through entrepreneurship [4]. The common thread is a belief that the interesting material is the part people usually keep off stage.

On building alternatives rather than reforming institutions

His more recent attention is on where talent should go when existing institutions stop working, covering the decline of legacy institutions and the shift from a rules-based order toward a code-based one, and pointing people toward building parallel structures [5]. He engages with this the way he engages with a product, by turning up and looking. Arriving at Network School expecting something quiet from what he had seen online, he found instead that "it seemed to be like a bustling sort of place," and his first question was how the people running it define success and KPIs a year in [2]. He also pushed on the psychology behind the thesis, asking what else people are in denial about beyond the examples already marked to market, and whether the recognition, when it comes, arrives as a sudden flight or a gradual dawning [2].

Takeaways

• Treat moderation as the product: on After School "every single post goes through proactive moderation," with humans first, machine classification handing ambiguous cases back to people, and one-tap community reporting that removes content with no questions asked [3].
• Wire crisis support directly into the posting flow rather than into a help page; the Crisis Text Line partnership offered a trained counselor the moment a teenager posted distress, and over 100,000 teens used it [3].
• Distinguish real traction from forced traction: earlier products meant "a lot of fighting to get those users," while the After School beta went from zero to over 50% of a high school in a week and to 200,000 users in three weeks on word of mouth [3].
• Fast builds rest on slow groundwork; the beta took three or four months only because the team had spent three or four years iterating on social products since 2011 [3].
• Keep the team small and the filter tight: "you can do a lot with a little," and "we interview tons of people and only let a few in" [3].
• Fundraise indirectly, because "when you ask for money you get advice," and the reverse is what actually raises rounds [3].
• Get proximity before you get permission: cold-calling investors to be the teen in their office, shadowing at Founders Fund, Union Square Ventures and Tech Stars, and later going undercover into high schools to watch the product in use [3].
• Adoption belongs to users rather than institutions; After School worked with students directly and handled schools through a separate safety and information effort, including a parents guide made with ConnectSafely [3].