Sep 20, 2026 · 1h 37m
Peter Sellis argues product leaders need more restraint and ambition
90 minutes of unfiltered product advice from Snap and Discord’s product chief | Peter Sellis
As AI makes feature creation cheaper, product leaders face greater pressure to choose what deserves to exist and connect it to durable business value.
- 1Strong organizations give exceptional people autonomy, responsibility, and room to work differently.
- 2Durable growth usually comes from strengthening the core product rather than chasing every adjacent market.
- 3AI can accelerate building, but taste and restraint determine whether added features improve the product.
Don't miss
Sellis makes the case that AI’s ability to add features will make restraint and taste more important, not less.
The brief
Peter Sellis, whose product leadership spans Snapchat and Discord, explains why he avoided podcasts before offering unusually candid views on teams, talent, and the limits of product management.
Sellis argues that autonomous teams need shared ideology and clear decision rights, while exceptional people such as Nikita Bier may require organizations willing to accommodate unconventional ways of working.
His critique extends to growth: Snapchat built a powerful consumer product but struggled to monetize its young audience, showing that product strength and business success can diverge.
The central lesson is restraint. AI can generate features quickly, but deciding what to remove—and what core value to strengthen—requires taste, ambition, and judgment.
Sellis closes by tying mission, user behavior, retention, and business metrics together, arguing that the best product leaders make those connections memorable and causally clear.
What was said on this episode
22 statements · 15 positive · 7 negative
Snap’s young audience is less valuable to advertisers than Pinterest, Instagram, or Twitter audiences.
“Snap's audience is definitively like less valuable from that perspective versus like a Pinterest or an Instagram or a Twitter.”
Listen at 22:58
Snap faced structural difficulty building a multibillion-dollar advertising business.
“Snap was trying to build a multi-billion dollar ads business very quickly on a lower value audience, a very tough format for ads, and a service that was more utilitarian than consumption oriented.”
Listen at 24:46
Snapchat Plus has been successful as a subscription product.
“I think Snapchat Plus has actually been quite successful”
Listen at 28:52
Instagram’s version of Stories is better than Snapchat’s version.
“the version of it on Instagram is now much better.”
Listen at 33:00
Software features no longer provide a durable competitive moat.
“software is no longer a moat”
Listen at 33:46
Product restraint is one of the few ways leaders can exercise product taste.
“Exercising that kind of muscle of like saying no, uh, of restraint is probably one of the easiest, like one of the few like muscles of taste you can exercise.”
Listen at 36:29
Advertising is the strongest consumer monetization model on the internet.
“ads are, ads are the product market fit for the internet.”
Listen at 40:41
Consumer subscriptions are generally overrated compared with advertising.
“subscription products and things like that are kind of overrated versus ads from a consumer side of things.”
Listen at 40:56
Modern advertising marketplaces and auctions are essentially solved problems.
“What drives a modern ads marketplace and auction is essentially a solved problem”
Listen at 41:34
A modified VCG auction is the correct mechanism for ChatGPT advertising.
“I do think that the mathematical mechanism, which is essentially a modified VCG, I think Vickrey-Clarkesgrove auction mechanism, this mathematical mechanism is the correct way to build an ads auction in even like a chat environment, in ChatGPT.”
Listen at 41:43
Maximizing trust likely improves user retention and advertiser value.
“trust maximization will probably be correlated with user retention, if not causal, and with advertiser value.”
Listen at 44:20
Designing ad formats at OpenAI is an especially important product role.
“one of the coolest jobs right now for a PM, uh, and designer would probably be like ad formats at OpenAI.”
Listen at 44:40
Network-effect businesses grow more effectively by strengthening existing users’ experience.
“what you can do to drive growth much more effectively is focus on people who are essentially already using or wanting to use the product”
Listen at 51:19
Almost every consumer product should optimize for daily active use.
“I think almost every consumer product, maybe it's also a controversial take, like needs to be optimizing for daily active use.”
Listen at 51:44
Improving Discord’s core gaming experience drove exceptionally fast growth.
“As we made Discord better, as we focused on performance, we were actually able to drive some of the fastest, I think if not the fastest growth”
Listen at 55:12
Companies should continually assess adjacent markets for their core strengths.
“you should honestly always be assessing that”
Listen at 58:43
A radically superior competitor can eliminate an improving business almost overnight.
“everything's getting better and better and better until something comes along that's literally 10x better than what you're doing and it takes you to zero overnight.”
Listen at 1:01:56
SpaceX is currently the world’s greatest company.
“SpaceX is the greatest American company right now, greatest company in the world, truly phenomenal on every single level.”
Listen at 1:02:49
Falcon 9 is becoming obsolete relative to SpaceX’s future launch systems.
“the Falcon 9 is the past.”
Listen at 1:04:09
SpaceX is willing to replace its successful products with superior successors.
“we're completely willing to build the thing that will compete against ourselves and make the thing that we built irrelevant.”
Listen at 1:04:35
SpaceX’s current activities were outlined in plans published roughly ten years earlier.
“Nothing SpaceX is doing is not something that they didn't talk about 10 years ago.”
Listen at 1:06:06
People who repeatedly raise organizational ambition are rare, especially in large organizations.
“the type of people who can do that within an organization, especially a large organization, are pretty rare.”
Listen at 1:08:58
Statements are attributed to the speaker as said on the episode and reflect their view at the time, not PodLume's. They are not advice.
Featuring
Discord
Meta
ChatGPTBooks & mentions
Listen to the full episode and explore every guest, topic, and moment on PodLume.