Sachin Monga on Substack Recommendations, Writers on the Internet, and Building with Principles
Guest: Sachin Monga — Head of Product at Substack; previously founded Cocoon (acquired by Substack); spent seven years at Facebook on video, camera, developer platform, and ads growth.
Host: Lenny Rachitsky
Source: Lenny’s Podcast.
Overview
Sachin Monga traces the Substack recommendations feature from a single insight about organic cross-discovery to the moment it drove 70% of Lenny’s subscriber growth and millions of subscriptions across the network. The episode pairs that product story with a broader argument: Substack is a principled company, and starting from “who should be in control?” is what made the feature work. Along the way, Sachin distinguishes the PM’s job at a startup (sequencing trade-offs) from the PM’s job at Facebook (permanent trade-offs), and describes how to build product alongside a strong, product-minded founder.
Key ideas
- Recommendations origin: writers pick, no algorithm. The team noticed organic cross-discovery happening through guest posts and comments. The most natural implementation — Substack inserts “you might like” suggestions — would have taken control away from writers. Instead they asked: what if writers just say who they recommend? The feature went viral through goodwill emails to recommenders, driving millions of new subscriptions and making 1-in-3 new subscriptions network-driven.
- Writers in control as design principle. Every product decision at Substack bottoms out in the same question: does this give writers (or readers) more or less control over their experience? Recommendations is the clearest example — writer-curated lists rather than an algorithmic feed preserves the principle while driving growth.
- Build with Writers / Product Lab. A standing invite-only group of ~100 writers participates in feature development before rollout. The recommendations feature went through this process; what shipped day one was materially different from the initial proposal as a result.
- Facilitator, not decision-maker, with a strong founder. Sachin’s first job at Substack was to ensure Chris could see what all the teams were doing, and teams could model Chris’s thinking. Weekly start/end check-ins created enough shared context that trust formed quickly. The PM’s role in a product-minded founder environment is facilitation and catch-up, not agenda-setting.
- Startup sequencing vs Facebook permanent trade-offs. At Facebook: doing A might mean B is worse forever (a new tab that cannibalises existing ones). At Substack: most decisions are time-sequenced — do this first to unlock the next thing. A startup PM’s job is to understand the order of operations, not to manage irreconcilable conflicts.
The recommendations feature
Mechanic: When a reader subscribes to any Substack, they see a curated screen of newsletters the author recommends — writers opt in and pick their own lists. The subscribe flow is the current primary surface; the social graph being built (of writer-to-writer goodwill and influence) is the durable asset.
Why the principle-driven approach worked: The alternative — algorithm-driven discovery inserted into emails or posts — would have put content Lenny didn’t choose in front of Lenny’s readers. Substack’s control principle ruled that out. Writer-curated lists are slower to scale but aligned to the model.
Viral mechanic: When Writer A recommends Writer B, Writer B receives an email listing every reader being sent their way. This generates immediate goodwill and often a reciprocal recommendation — a loop built on generosity rather than engagement engineering.
Scale (at time of recording): Millions of new subscriptions driven by the feature; tens of thousands of writers receiving recommendation-driven subscribers; 1-in-3 new subscriptions and 1-in-10 paid subscriptions coming from the Substack network.
Facebook vs startup PM trade-offs
| Dimension | Facebook (large, mature) | Substack (small, formative) |
|---|---|---|
| Nature of trade-off | Permanent: A hurts B forever | Sequential: do A now means B comes later |
| Main variable | What is possible at all? | What order should things happen in? |
| PM’s central skill | Navigating permanent conflicts | Understanding order of operations |
| Team structure stability | Frequent reorgs (every 3–6 months) | Stable because teams orient around customer problems, not surfaces |
The implication: process expertise from a large company transfers less than expected to an early-stage startup — not because the skills are wrong, but because the fundamental game is different.
Working with a product-minded founder
Key patterns from Sachin’s early months at Substack:
- Start as facilitator: ensure Chris knows what all teams are doing; ensure teams can model Chris’s thinking.
- Weekly start/end check-ins: one hour Monday (what are we focused on, what are we worried about), one hour Friday (check in on progress).
- Catch-up on vision depth: Chris has been thinking about the problem 5× longer; closing that gap is part of the PM’s job.
Sachin’s framing: “If Chris could have a really good sense of what all the teams are doing and if the teams knew where he was coming from and could start to get better at modeling him and his vision, that would be a win.”
The golden era of writing
Sachin closes with a claim about the moment Substack represents: the economic model for writing on the internet has been poor for the entirety of the internet’s history. Advertising requires attention accumulation at scale. Substack’s subscription model changes the equation — a writer with 1,000 true fans paying $10/month makes $120K/year, without needing millions of pageviews. The network effects now compounding inside Substack make that audience increasingly reachable.
Reference: Kevin Kelly’s You Are Not Late and 1,000 True Fans — both cited as canonical framing for Substack’s thesis.