ExpertsFM90 minutes of unfiltered product advice from Snap and Discord’s product chief | Peter Sellis
Lenny's Podcast: Product | Career | Growth · September 20, 2026
5 Selected Lessons:
heuristic · 2 case · 2 framework · 1
Core Product Value: a three-part phrase + metrics definition that aligns mission with growth levers
Source
Lenny's Podcast: Product | Career | Growth — 90 minutes of unfiltered product advice from Snap and Discord’s product chief | Peter Sellis (September 20, 2026) · at 50:52 ▸ · at 1:13:36 ▸
When this applies
when a team needs a shared, actionable definition of why users come back to a product that can guide both roadmap prioritization and data science metric design
The takeaway
Define a product's core value proposition as a short, decomposable phrase paired with retention-linked metrics for each component, so prioritization decisions and growth hacks can be tested against a single shared standard.
Why it works
A pure mission phrase is too abstract to action day-to-day, while a pure metric is meaningless without a narrative; combining a decomposable three-part phrase (e.g., 'fastest way to share a moment with people you care about') with matched, retention-causal metrics lets every function — PM, data science, engineering — test any roadmap idea against the same shared standard and onboard quickly into a shared culture.
One level up
Alignment at scale requires translating mission into both language (for buy-in) and metrics (for execution) simultaneously — neither alone suffices.
Where it stops holding
This stops working when the metrics chosen aren't actually causally linked to retention (just correlated), misdirecting the team's effort, or when the phrase is too vague to decompose into testable components.
Detail
procedure:
- Write a short phrase describing why users return daily, decomposable into ~3 components
- Break each component into a testable concept (e.g., 'fast' = time to load/share)
- Define metrics for each component that are causally (not just correlationally) linked to retention
- Use this phrase+metrics combo in onboarding (Sellis delivered this presentation to every new Snap hire for 7 years) so new data scientists/PMs/engineers can action it immediately
- Test every roadmap item against whether it strengthens this core value
benchmarks: {"pattern":"three components seems to recur ('rule of threes') in both examples given"}
rationale: Phrase alone doesn't get actioned by data scientists; metrics alone are meaningless to the broader team — you need both plus a repeated story to embed it in culture.
In their words
you need to define both from a theoretical or like a conceptual level as well as a metrics level, like what it means for somebody to come back to your product every single day.
Invest in winners, let strugglers fail on their own
Source
Lenny's Podcast: Product | Career | Growth — 90 minutes of unfiltered product advice from Snap and Discord’s product chief | Peter Sellis (September 20, 2026) · at 0:00 ▸ · at 46:25 ▸
When this applies
when allocating scarce management attention across a team with uneven performers
The takeaway
Concentrate scarce coaching/attention on the people already compounding value, rather than on the people struggling, unless you have an actual mechanism to develop them.
Why it works
Giving high performers more autonomy and responsibility creates a self-reinforcing confidence loop that increases their output, while propping up strugglers consumes disproportionate time for marginal gains and doesn't address the underlying skill gap.
One level up
Resource allocation should follow where marginal attention produces the most leverage, not where it feels most emotionally urgent.
Where it stops holding
This stops working when you have no mechanism to detect a struggling top performer quietly failing, or when 'average' performers could be coached up cheaply and are simply being neglected.
Detail
rule: Give your best performers increasingly more rope/responsibility; let weaker performers fail rather than invest heavily in improving them.
threshold: Push until you start to see bad decisions from a strong performer — that's the signal to intervene.
rationale: He always defends his team publicly and privately never criticizes, so people with consistently good decisions are loaded with more decisions until failure, rather than spreading attention evenly.
Long-term greedy: optimize ads for advertiser ROI today to compound revenue tomorrow
Source
Lenny's Podcast: Product | Career | Growth — 90 minutes of unfiltered product advice from Snap and Discord’s product chief | Peter Sellis (September 20, 2026) · at 1:09:02 ▸
When this applies
when a revenue-driven function faces pressure to optimize a metric short-term in a way that's causally linked to the long-term health of that same metric
The takeaway
Give teams an explicit, memorable framing of the causal chain between a short-term sacrifice and the long-term metric it compounds into, so the framing can function as organizational 'cover' when short-term pressure (leadership, investors, layoffs) pushes toward the easy lever.
Why it works
Future ad revenue growth is causally downstream of the ROI advertisers get today (via budget expansion and net revenue retention), but the easiest lever to boost ROI short-term (cutting price) destroys the very compounding effect you need, so an explicit 'long-term greedy' framing gives teams ideological cover to resist the easy short-term lever under pressure.
One level up
Sustainable advantage often requires accepting a short-term cost for a causally-linked long-term gain, and this requires explicit ideological language to survive organizational pressure, not just good intentions.
Where it stops holding
This stops working when the causal link between short-term sacrifice and long-term gain is weak or unproven, turning 'long-term greedy' into an excuse for simply missing numbers.
Detail
rule: Frame ads strategy as 'long-term greedy': prioritize advertiser ROI today even at the cost of short-term revenue, because ROI today causally drives budget growth and retention tomorrow.
threshold: Especially apply this discipline during the hardest moments — public company number misses, potential layoffs — when the temptation to sacrifice long-term ROI for short-term revenue is highest.
rationale: In modern auctions price is an output, not an input, so lowering price (the easy ROI lever) isn't the sustainable path; sustainable growth comes from delivering real value that compounds via retention and expanding budgets.
In their words
long-term revenue tomorrow for advertising is essentially causally related to the ROI you drive for advertisers today
Snap's camera-first, young-audience design structurally limited ad monetization
Source
Lenny's Podcast: Product | Career | Growth — 90 minutes of unfiltered product advice from Snap and Discord’s product chief | Peter Sellis (September 20, 2026) · at 20:50 ▸ · at 24:06 ▸
When this applies
when a product's core usage flow and audience demographic are structurally mismatched with the monetization model you're trying to bolt on
The takeaway
Before committing to a monetization strategy, check whether the product's core usage pattern and audience structurally support that strategy's fundamental requirements — don't assume execution can overcome structural mismatch.
Why it works
Ads monetization depends on (1) a measurable, trackable, high-value audience, and (2) a consumption/feed-based attention flow where ad placement is natural; a young audience with regulatory tracking limits and a creation-oriented (not consumption-oriented) camera-first flow and a messaging-app core violate both conditions simultaneously.
One level up
Business model choice must be judged against the physics of the product's usage pattern, not just market size or growth.
Where it stops holding
This stops working when the structural mismatch can be worked around by a genuinely novel ad format not yet invented, or when alternate revenue models (subscriptions) are prioritized early enough to reduce pressure on ads.
Detail
situation: Snap built a billion-MAU product with the #1 camera, messaging, maps, and astrology app among teens, but investors expected a Facebook-scale ads business.
decision: Snap pursued rapid ads scaling despite a camera-first creation flow and a young, hard-to-track, hard-to-target audience, including attempts at camera monetization that Sellis calls one of his biggest failures.
outcome: Snap built a multi-billion dollar ads business faster than Twitter, Reddit, or Pinterest reached similar scale, but fell short of 'next Facebook' expectations because the structural constraints (young users, creation-not-consumption flow, messaging-app core) capped ad value per user.
lesson: A product with great engagement doesn't automatically support the monetization model investors expect if its structure fights against that model's requirements.
In their words
it's core, it's a messaging app. And like, there aren't really great ad products for messaging apps right now.
'There's always money in the banana stand' — Snap and Discord both grew fastest by perfecting the existing core, not chasing new markets
Source
Lenny's Podcast: Product | Career | Growth — 90 minutes of unfiltered product advice from Snap and Discord’s product chief | Peter Sellis (September 20, 2026) · at 50:52 ▸ · at 54:09 ▸
When this applies
when a network-effect or habit-driven product has stalled and the team is tempted to chase a flashy adjacent opportunity (e.g., a trendy use case growing on top of the product) instead of doubling down on the existing core users
The takeaway
When growth stalls, first exhaust the opportunity to deepen engagement among your most intent-aligned existing users by removing friction and improving core performance, rather than chasing adjacent trendy use cases that only look big from the outside.
Why it works
Network-effect products grow implicitly for people already using/wanting the product; incrementally improving performance/relevance for existing heavy-intent users moves their frequency (e.g., 10/30 days to 13/30 days) and compounds far more reliably than acquiring new low-intent users who may not retain, because the former group already has demonstrated intent and the product-market fit lever is friction removal, not demand creation.
One level up
The most reliable growth lever is usually hiding in plain sight inside your existing core, not in an exciting new frontier.
Where it stops holding
This stops working when the core user base is already maximally saturated/retained and truly has no more frequency headroom, making expansion to new cohorts the only remaining lever (see the companion framework on when to bet outside the core).
Detail
situation: Snap's DAU flattened after the disastrous 2018 redesign; separately, Discord's growth during 2024 was expected by outsiders to come from trendy adjacent use cases like Midjourney/crypto, even though gamers already had '100% penetration' of Discord.
decision: Snap's team laser-focused on Android performance for existing users instead of new acquisition; Discord's team focused entirely on making the product better for friends playing games together before/during/after, rather than chasing crypto/AI-adjacent growth.
outcome: Snap entered a multi-year 'growth renaissance'; Discord drove what Sellis calls possibly its fastest growth since the pandemic, by moving existing gamers from using Discord 10/30 days to 12-13/30 days — and got an added halo effect because gamers are an unusually vocal, detail-oriented user base that rewarded visible responsiveness.
lesson: Even at '100% penetration,' there is still real growth headroom in deepening frequency/retention among your core, and this is often more reliable than chasing adjacent trendy cohorts.
In their words
There's always money in the banana stand, right? Return to your core users, make sure it's perfect for them.