> ## Content Index
> Fetch the complete content index at: https://www.signalful.com/llms.txt
> Use this file to discover other available public pages before exploring further.

# Hand Your Best People More Rope
- URL: https://www.signalful.com/hand-your-best-people-more-rope/
- Published: 2026-10-04T15:49:00.000Z
- Updated: 2026-10-07T05:15:33.000Z
- Description: Snap's first product manager designs teams around two traits he borrowed from an unlikely place. He explains why it works.
- Author: Signalful Editorial
- Tags: Builders, Snap, Peter Sellis, Product Management, Org Design, Hiring, #sf-0036, #yt-97LRJUUPy_w, #Import 2026-10-06 15:50

Peter Sellis was the first product manager at Snapchat, where he spent 7 years and ran much of the company's ads product, and later served as head of product at Discord. He started his career in 2009 at Ustream, an early live streaming company, under the mentorship of Josh Elman. At Discord he managed Nikita Bier after the company acquired Bier's startup, and he led the 2024 refocus on people who play multiplayer games with friends. He recorded this, his first podcast, while between roles and free of any corporate comms team. These signals are from Peter Sellis's interview with Lenny Rachitsky on Lenny's Podcast.

1. **The Terrorist Org Chart.** Sellis designs product teams around two traits he sees in effective terrorist organizations: an ideological culture where everyone knows why they're doing what they do, and a clear structure of who can be trusted with which decision. His reason is cost. Collaboration brings coordination costs that leave a team moving "as slow as the slowest node," and he notes no executive ever tells people to stop collaborating. At high-growth companies, PMs make calls the founders would have made 6 months to 2 years earlier, so he turns the founders' vision into short phrases people can repeat in place of the whole strategy. Discord's chief people officer, he says, was horrified and intrigued at once.
2. **Room for Spiky Talent.** Sellis managed Nikita Bier at Discord and compares him to Liam Neeson in Taken: a rare set of skills that wins in the right situation and goes badly when you need him to buy groceries. He once had to message Bier on 4 different apps to get a one-on-one, and threatened to book a paid call with him on Intro. Sellis calls a product org that can't absorb this kind of person a failure, and models his approach on Phil Jackson, who got players who didn't get along to win titles. He treats creative PMs like bands: if he still believes in the artist, he sticks with them through a dud album.
3. **The Median PM Problem.** Sellis argues that most PMs are bad, and the math explains it. PMs come from the right tail of a skill distribution, which behaves like a power law where the median sits far below the average. During the zero-rate era, product management became the best-paid tech career for nontechnical people, so weak PMs fight to stay while strong ones leave to become founders and executives. Snap went a long time without hiring PMs, and its bar for adding one was that "this person has to provide value over nobody."
4. **Who Complains About PMs.** Great engineers rarely complain about product managers, Sellis says, because they bend PMs to their will or refuse to work with bad ones. Bad engineers don't complain either, since they can hide behind PM decisions. The complaints come from engineers in the middle. That's why he evaluates each PM together with the team they sit on, to check that the pairing is net positive.
5. **Why Snap's Ads Stalled.** Sellis names three limits on Snap's ads business. Its audience skews under 18, and ad prices jump sharply at 18 and again at 21 because targeting and measurement rules for teens are strict. The app opens to the camera, a creation tool with no natural slot for an ad, and at its core Snap is a messaging app, a category no one has monetized well per minute of use. He calls camera monetization one of his biggest failures. Even so, Snap reached about $4 billion in ad revenue faster than Twitter, Reddit or Pinterest despite starting after them, and he thinks Snapchat+ launching 4 years earlier would have eased the pressure on ads.
6. **Taste Is Restraint.** Sellis says he watched Evan Spiegel personally say no to better ideas than most consumer startups ever launch. He wanted to help founders become the richest museum curators in the world, where the shipped product is a small slice of the collection. "If a museum has all of their collection on the walls then the curator hasn't done anything." To test taste in an interview, he asks candidates about prototypes they built, held in their hands, and chose not to ship.
7. **Trust as the Auction Currency.** Sellis believes ads are still the best way to monetize consumer internet products, and says OpenAI is speedrunning the ads infrastructure Meta worked out between 2010 and 2020\. He backs a modified Vickrey-Clarke-Groves auction, where an ad's organic value to the user factors into the price an advertiser pays. In chat he would swap organic value for trust: how far an answer to "what mic should I buy" bends toward advertisers should depend on how much trust it costs. He points to Amazon search, where 16 ads can appear before the first result, and calls ad formats at OpenAI one of the coolest PM jobs today.
8. **Ride Your Racehorses.** When someone is crushing it, Sellis keeps "handing them more and more rope to hang themselves with," loading on decisions until they start making bad ones. He defends his people in front of executives and never criticizes them in front of others, because confidence makes people better. He admits he has no good method for helping weak performers improve, so he lets them fail on their own. The caveat for managers who copy this is that a person getting little attention may simply be trusted to execute.
9. **Money in the Banana Stand.** Sellis says growth in network products comes from the core user. A monthly user is worth 1/30 of a daily one, so moving DAU/MAU a few points beats chasing new signups. After Snap's 2018 redesign, the only time its DAU went flat quarter over quarter, the team focused on Android performance and got years of growth. At Discord, Midjourney was invisible in usage data; the 2024 push on friends playing multiplayer games, including fixing bugs from Reddit threads, moved users from 10 active days a month to 11, 12 and 13.
10. **Backcasting the Next Bet.** For new bets, Sellis watches for users jumping through hoops, as Discord users did when they stretched a 20-person gaming group into large communities. The harder method is backcasting from the company's philosophy with no reference to today's product: ask what the best way to share a moment would be, and you arrive at glasses. Few people can do this, and he jokes that most PMs take an ayahuasca trip or visit Japan and come back sure they've solved it. He points to SpaceX phasing out the Falcon 9, the best launch system ever built, for Starship as rare proof a company will make its own winner obsolete.
11. **Long-Term Greedy.** Ads teams chase quarterly revenue goals, so Sellis gave his teams at Snap and Discord the phrase "long-term greedy." Tomorrow's ad revenue depends on the ROI advertisers get today, even when delivering it costs short-term revenue. The phrase gave teams cover to think long term during missed quarters and layoff threats. He also defined "core product value" in words and in metrics: Snap was "the fastest way to share a moment with your best friends," and he delivered the deck explaining it to every new hire for 7 years.
12. **Three Pairs of Wolves.** Sellis sees three contradictions in every great PM. They are confident to the edge of arrogance, which lets them give credit freely to the designer who made a last-second change or the engineer who stayed up all night. They love checklists and good process yet are comfortable throwing the plan out, close to despising authority. And they hold a 2-year picture of the product while showing a bias for action he calls "borderline annoying," so the team wakes up when they do.

Watch the full video at [https://www.youtube.com/watch?v=97LRJUUPy\_w](https://www.youtube.com/watch?v=97LRJUUPy%5Fw&ref=signalful.com).