InsightfulDiscussion

90 minutes of unfiltered product advice from Snap and Discord’s product chief | Peter Sellis

Lenny's Podcast: Product | Career | Growth1h 36m

Peter Sellis, former head of product at Snap and Discord, discusses building high-performing product organizations by organizing teams like ideological structures with clear decision-making authority, argues that most product managers are mathematically bad due to power-law distributions in skill, and shares lessons on focusing growth on core product value rather than expansion.

Summary

Peter Sellis brings unprecedented candor to this podcast episode, enabled by being between corporate roles without PR constraints. He opens by explaining his organizational philosophy using the metaphor of a "terrorist organization"—not for violence but for combining clear ideological mission with transparent decision-making hierarchies. This allows teams to move quickly and autonomously without excessive collaboration, which he argues carries hidden coordination costs that most executives ignore.

On product management as a profession, Sellis presents a mathematical argument for why the median PM is bad: product management is normally distributed across the population, but only those on the right tail become PMs. This creates a power law distribution where the median falls far below the average. This effect intensified during the zero-interest-rate era when non-technical careers became lucrative, attracting mediocre talent while great PMs exited to founding or executive roles.

Regarding Snapchat's business trajectory, Sellis provides nuanced analysis beyond the common "young users are hard to monetize" explanation. He identifies three specific challenges: the young user base with lower advertiser CPMs and tracking restrictions, the camera-first product design that creates no natural ad placement, and the messaging-app core that lacks proven monetization models. Despite these constraints, Snap achieved impressive growth by launching ads faster than Pinterest, Reddit, and Twitter—a fact he feels deserves more credit.

Sellis emphasizes that growth typically comes from optimizing the core business rather than expansion. He uses DAU-MAU ratios as the primary metric, noting that moving this ratio a few basis points matters more than acquiring new monthly users. At Snap, focusing on performance for existing Android users drove a multi-year growth renaissance after the 2018 redesign. At Discord, despite external focus on cryptocurrencies and AI happening on the platform, the actual growth driver was optimizing the gaming experience for existing users.

On managing talent, Sellis advocates for "riding your best people into the ground" rather than developing average performers. He loads high performers with increasing responsibility until they fail, creating a culture of trust where he defends people publicly but doesn't over-invest in improving mid-tier performers. He references Phil Jackson's ability to mesh disparate talent styles and describes managing someone like Nikita Beer (former head of product at X) as leveraging specialized skills rather than traditional management.

Sellis introduces several frameworks he's developed: "long-term greedy" for ads businesses (prioritizing advertiser ROI today for revenue tomorrow through retention and budget growth), and "core product value" (a three-part phrase combining mission with metrics that guides product decisions and onboarding). He emphasizes the importance of taste in product, evidenced by Evan Spiegel's ability to say no to better ideas than most startups launch—framing good curators as those who don't display their entire collection.

On AI's impact on products, Sellis argues that a new PM skill is pushing back on features AI suggests, since AI is biased toward adding rather than removing. He views OpenAI's advertising infrastructure as "speed-running" proven mechanisms from Meta, and believes the real innovation will be how trust is factored into auction mechanisms rather than just organic relevance.

Sellis identifies three oxymorons that define great PMs: confidence balanced with humility about personal credit, organizational discipline balanced with comfort in ambiguity, and long-term vision balanced with daily bias for action. He also presents his spiciest take—that certain demographic and socioeconomic backgrounds make people better suited for specific consumer products due to their lived experience with certain human needs, though he acknowledges this isn't actionable in hiring contexts.

Finally, he reflects on SpaceX's willingness to sunset the Falcon 9 (arguably the world's greatest launch system) in favor of Starship, demonstrating the organizational courage needed to compete against your own successful products when pursuing a larger mission.

About this episode

<p><strong>Peter Sellis</strong> was the first product manager at Snapchat, where he spent seven years building one of the most beloved consumer products in history. He went on to serve as Head of Product at Discord, where he helped drive some of the fastest user growth the platform had seen. Having left his role at Discord, he gets to be especially real and unfiltered — nothing to promote, no PR, no comms team.</p><p></p><p><strong>In our in-depth conversation, we discuss:</strong></p><p>1. Why Peter designs teams like . . . terrorist organizations</p><p>2. What it was really like to manage Nikita Bier</p><p>3. Why the median PM is so bad</p><p>4. Why Snapchat’s ads business never reached its potential</p><p>5. Why growth almost always comes from the core product</p><p>6. Peter’s three “oxymorons” of great product management</p><p>—</p><p><strong>Brought to you by:</strong></p><p><a href="https://workos.com/lenny" target="_blank"><strong>WorkOS</strong></a>—Make your app enterprise-ready, with SSO, SCIM, RBAC, and more</p><p><a href="https://mercury.com/command?utm_source=lennys&#38;utm_medium=sponsored_newsletter&#38;utm_campaign=26q3_brand_campaign" target="_blank"><strong>Mercury</strong></a>—Radically different banking, now with Command</p><p>—</p><p><strong>Episode transcript: </strong><a href="https://www.lennysnewsletter.com/p/90-minutes-of-unfiltered-product" target="_blank">https://www.lennysnewsletter.com/p/90-minutes-of-unfiltered-product</a></p><p>—</p><p><strong>Archive of all Lenny's Podcast transcripts: </strong><a href="https://www.dropbox.com/scl/fo/yxi4s2w998p1gvtpu4193/AMdNPR8AOw0lMklwtnC0TrQ?rlkey=j06x0nipoti519e0xgm23zsn9&#38;st=ahz0fj11&#38;dl=0" target="_blank">https://www.dropbox.com/scl/fo/yxi4s2w998p1gvtpu4193/AMdNPR8AOw0lMklwtnC0TrQ?rlkey=j06x0nipoti519e0xgm23zsn9&amp;st=ahz0fj11&amp;dl=0</a></p><p>—</p><p><strong>Where to find Peter Sellis:</strong></p><p>• X: <a href="https://x.com/petersellis" target="_blank">https://x.com/petersellis</a></p><p>• LinkedIn: <a href="https://www.linkedin.com/in/disgruntled" target="_blank">https://www.linkedin.com/in/disgruntled</a></p><p>• Website: <a href="https://pjs.lol" target="_blank">https://pjs.lol</a></p><p>—</p><p><strong>Where to find Lenny:</strong></p><p>• Newsletter: <a href="https://www.lennysnewsletter.com" target="_blank">https://www.lennysnewsletter.com</a></p><p>• X: <a href="https://twitter.com/lennysan" target="_blank">https://twitter.com/lennysan</a></p><p>• LinkedIn: <a href="https://www.linkedin.com/in/lennyrachitsky/" target="_blank">https://www.linkedin.com/in/lennyrachitsky/</a></p><p>—</p><p><strong>In this episode, we cover:</strong></p><p>(00:00) Introduction</p><p>(02:48) Why Peter avoided podcasts until now</p><p>(05:44) Designing teams like terrorist organizations</p><p>(10:35) Cult versus terrorist org: the key distinction</p><p>(11:19) Managing Nikita Bier</p><p>(12:56) Building orgs that handle spiky talent</p><p>(15:16) Why the median PM is probably bad</p><p>(18:28) Is the PM profession getting better?</p><p>(20:49) Snap’s three ads monetization challenges</p><p>(25:03) Snap’s underrated product track record</p><p>(27:57) Systems thinking</p><p>(32:19) Instagram copying Snapchat’s worst features</p><p>(33:32) Software moats</p><p>(35:27) Taste, museum curators, and the art of saying no</p><p>(40:05) OpenAI, ads, and the trust auction mechanism</p><p>(46:25) Counterintuitive lesson: ride your best people hard</p><p>(50:13) Why collaboration is often overrated</p><p>(50:50) Growth lives in the core product</p><p>(57:50) When to bet on something genuinely new</p><p>(01:02:18) SpaceX and sunsetting the Falcon 9</p><p>(01:07:01) Ambition as a learnable muscle</p><p>(01:10:26) Long-term greedy: the ads philosophy</p><p>(01:13:36) The Core Product Value framework</p><p>(01:16:33) The three oxymorons of a great PM</p><p>(01:20:29) Contrarian corner</p><p>(01:26:00) Lightning round and final thoughts</p><p>—</p><p><strong>References: </strong><a href="https://www.lennysnewsletter.com/p/90-minutes-of-unfiltered-product" target="_blank">https://www.lennysnewsletter.com/p/90-minutes-of-unfiltered-product</a></p><p>—</p><p>Production and marketing by <a href="https://penname.co/" target="_blank">https://penname.co/</a>. For inquiries about sponsoring the podcast, email <a href="mailto:[email protected]" target="_blank">[email protected]</a>.</p><p>—</p><p><em>Lenny may be an investor in the companies discussed.</em></p> <br /><br />To hear more, visit <a href="https://www.lennysnewsletter.com?utm_medium=podcast&#38;utm_campaign=show-notes-no-free-preview-language">www.lennysnewsletter.com</a>

Key Insights

  • Sellis argues that effective teams should be organized with clear ideological missions and transparent decision hierarchies (like successful decentralized organizations), rather than relying on consensus-based collaboration which carries hidden coordination costs that most executives fail to internalize.
  • He presents a mathematical explanation for why most PMs are bad: product management skill follows a normal distribution, but only right-tail performers become PMs, creating a power-law distribution where the median PM is substantially below average—an effect amplified during low-interest-rate eras that attracted non-technical talent.
  • Snap faced three specific monetization headwinds beyond young users: tracking restrictions for under-18 audiences, a camera-first design with no natural ad placement, and a messaging-app core that lacks proven ad models—yet still achieved faster ads growth than Pinterest, Reddit, or Twitter.
  • Sellis claims growth comes primarily from optimizing core product metrics like DAU-MAU ratios for existing users rather than acquiring new users, pointing to Snap's Android performance optimization and Discord's gaming focus as examples of moving this ratio a few basis points for significant returns.
  • He advocates managing high performers by increasing responsibility until they fail rather than developing mid-tier performers, arguing this creates a culture of trust and leverages asymmetric talent (like Nikita Beer's specialized skills) better than traditional development approaches.
  • Sellis introduced 'long-term greedy' as a framework for ads businesses: prioritizing advertiser ROI today (even at short-term revenue cost) because tomorrow's revenue is causally linked to today's performance, enabling mathematical justification for long-term thinking during earnings pressure.
  • He describes great product curators as those with the discipline to not display their entire collection, measured by the high-quality ideas they've built, held in their hands, but never shipped—a form of taste that AI systems struggle to replicate.
  • Sellis argues that certain demographic and socioeconomic backgrounds make individuals materially better at building consumer products for specific cohorts because understanding certain human needs in Maslow's hierarchy requires lived experience, though he acknowledges this insight isn't directly actionable in hiring.
  • He claims a new PM competency is needed: pushing back on features that AI suggests adding to products, since AI is biased toward additive rather than subtractive design, requiring human taste to maintain product focus.
  • Sellis observes that OpenAI is 'speed-running' proven advertising infrastructure mechanisms from Meta (particularly modified VCG auction mechanisms), but the real innovation will be how trust is factored into auction design rather than organic relevance.
  • He emphasizes that the most important career decision a product manager makes is choosing their life partner, claiming this has knock-on effects on growth and intellectual development that no amount of external learning can match.
  • Sellis identifies three oxymoronic traits of great PMs: confidence balanced with humility about personal credit, organizational discipline balanced with comfort in ambiguity, and long-term strategic thinking balanced with daily bias for action that sets company pace.

Topics

Product organization design and team structureProduct management as a discipline and careerGrowth strategy and core product optimizationAdvertising and monetization modelsSnapchat's business challenges and successesDiscord's growth and gaming focusTalent management and high-performer developmentProduct taste and curationAI's impact on product developmentLong-term thinking in business

Transcript

I asked you what are some counterintuitive lessons that you've learned that go against common startup wisdom? I'm no longer a beholden to any corporate comms team right now, so I can let it rip a little bit. Do you like to design teams like a terrorist organization? A good terrorist organization essentially has two things. One is it has this ideological culture that everyone understands why they are doing what they're doing. And then it has a very, very clear organizational structure of who can be trusted with which decision. You like to spend time riding your best people into the ground versus improving the average people. If you see somebody drowning, you want to save them. I did…

Full transcript available for MurmurCast members

Sign Up to Access

More from Lenny's Podcast: Product | Career | Growth

Get AI summaries like this delivered to your inbox daily

Get AI summaries delivered to your inbox

MurmurCast summarizes your YouTube channels, podcasts, and newsletters into one daily email digest.