Control Over Attention: How Substack Builds Product for Writers and Readers

Monga joined Substack about a year before the conversation, through the acquisition of Cocoon, a small photo-sharing app for close friends and family he had started about three years earlier with Alex Cornell. Before that, both founders had spent years at Facebook working on helping people share with friends and family. Monga himself joined in 2011 and worked on growth, platform, ads and the sharing team.

The problem they kept hitting, he says, was advertising as the business model. It forces a product to accumulate time and attention and turn it into sellable eyeballs, even when it is easy to imagine a better experience. Cocoon was an attempt to explore what the alternative looked like for one narrow use case.

Substack was their example of the same principle at work, with direct paid subscriptions between readers and writers. After getting to know the founders, they decided that a blogging tool and a photo app had consistent underlying motivations. The whole Cocoon team joined, and Monga now leads Substack's product and design teams.

When Monga arrived, Substack had no product managers, a few designers and around fifteen engineers. He now describes the company as at the end of a one-time transition to being product-driven. There are four PMs besides him, and three full-stack product teams that each have a PM, an engineering manager, data or design people and engineers.

The three teams serve writers, readers and growth, and a separate systems team without a PM keeps the lights on and helps the platform scale. Monga is surprised the structure has lasted, since Facebook reorganized every few months. He thinks it holds because teams are built around customers and timeless problems, not around surfaces like an app, a dashboard or podcasting.

Lenny adds that Substack has a platform advantage in which supply drives demand, since writers bring their own audiences, which explains why the reader team is newer. Monga agrees that Substack has only recently begun a concerted focus on readers.

Lenny asks how a head of product works with a founder with strong product opinions. Monga says he treated his role early on as facilitator rather than decision maker. Substack had outgrown a single weekly meeting where the CEO, Chris, could decide the next two weeks, so teams no longer knew what Chris had in mind and he did not know what they were doing.

For the first months his only job was to close that gap, helping Chris see what teams were doing and helping teams model his vision. Trust then built through repetition, and the two now meet for an hour at the start of each week to pick the big problems and again at the end to check in. Monga also tries to catch up on the five years Chris has spent thinking about the problem, noting that the company has recently begun doing things from Chris's 2017 Y Combinator demo.

On the hardest part of the job, he says any process you figure out becomes obsolete once the company grows, perhaps 2x. He tells his teams not to expect a perfect planning or communication process, just to get better every week, month and year. He adds that this suits a particular personality that is comfortable being humbled and not knowing what it is doing, and that it is a different job from product management at a stable company. Lenny agrees and says the advice is to do the best thing you can think of now and expect it to change in three to six months.

Monga says less of his Facebook experience transfers than he expected. In the core Facebook app, which tries to solve many problems for many people in one small rectangle, a PM's job is largely managing trade-offs. Adding a Watch tab at the bottom might mean no Marketplace tab, so prioritization is not about doing things in order.

Some basics still hold, such as favoring high-impact, low-effort work. But at Substack the main variable is time, and sequencing matters because the company is becoming something new. It began as a single-player tool for writers, and describing it today as only a newsletter tool would be reductive.

Lenny summarizes this as there being more one-way doors at a large company. Monga says it is more that the second-order effects are different, and that doing something now at Substack may unlock the ability to do something later.

Monga says the founders, Chris, Hamish and Jay, start from principle. They have an opinion about how the internet should work, with people far more in control of their destiny than in the last 10 to 15 years, when everyone spent time in a few ad-powered public squares. For Substack, that means writers deliver their best work on their own terms and earn directly from subscribers, and readers have real agency over what they see.

In practice, when there are several ways to build something, the team picks the one that gives writers or readers more control. Lenny describes Recommendations as letting a writer pick about ten newsletters to recommend after a reader subscribes, with no algorithm involved. He says about 70% of his growth now comes from the feature and his growth chart became a hockey stick the day it launched.

Monga explains the origin. Readers were already discovering newsletters through the lens of a writer, via guest posts or by clicking on commenters' profiles. The obvious approach would be a Facebook-style People You May Know unit that Substack picks for you. But inserting that at the bottom of a post would take control away from the writer, so the team went back and asked what if writers simply said who they recommend.

Chris was skeptical, not of cross-pollination but of whether so many steps would work: writers have to opt in, pick good people, and the recommendations have to be surfaced well. They tried it and it took off faster than imagined. Because recommended writers get an email saying who is sending them readers, it created a goodwill viral loop.

Monga says Recommendations have produced millions of new subscriptions across tens of thousands of writers. He adds that more than one in three new subscriptions on Substack and about one in ten paid subscriptions now come from the Substack network. Lenny says the readers are lower intent but not meaningfully so, since his open rates have fallen only a little.

Most subscriptions currently come from one flow, in which someone who has just subscribed to a writer sees recommendations. Monga wants to treat recommendations less as a step in a flow and more as a social graph of influence and goodwill. Long-time readers who trust you could see who you recommend, which should drive higher-intent subscriptions too.

Lenny adds that he rotates which newsletters he recommends, helping each reach around a thousand subscribers, in order to share the wealth. Monga jokes that he is the Robin Hood of Substack.

Monga describes a principle Substack is writing into its playbook: build with writers, build with readers. For a change that touches how Substack fundamentally works, the team does not roll it out to everyone with a dashboard notice. It calls about ten writers who might be interested, mocks up the idea and runs a small pilot first.

The company has also set up the Product Lab, an invite-only group of about a hundred writers who want to be on the leading edge of what Substack is becoming. It lets the team test features like Recommendations and get quick feedback. He says that what they ship on day one often differs a lot from what they had in mind before the process.

On press criticism, Monga says the key is separating signal from noise. He estimates about 90% of the chatter about Substack is a distraction for a team that should focus on executing, and says his skin was thickened at Facebook.

For writers, Monga says we are just starting a golden era. The economic model for great writing online has been poor throughout the internet's history. Early Substack successes like Matt Taibbi and Bill Bishop were established writers who had been undervalued. More recently, people who may not think of themselves as writers are making a living, even a fortune, with a relatively small number of people paying for high-quality work.

Over the next one to two years, he wants to make it simpler to get started. Substack will probably never have the biggest audience, he says, but it should be where your most valuable audience gets your best work. People with large followings elsewhere are using it as a home base where they own their audience's email addresses and can export them at any time. The company is also investing in video, audio and community features, and he points to Lenny's 30 meetups in a month as an example.

On the reader side, he describes a chance to build a place where the best culture is made and readers have extreme control over what they see. It would not replace scrolling feeds but would be somewhere you might go first. Substack launched an iPhone app about six months earlier and was about to launch an Android app.

Monga's first advice is to just start: gather subscribers, put a link somewhere, write a couple of things, or try video or audio if you are not much of a writer. He warns that it is a tenuous place to be if your whole following is locked into one platform where you cannot reach or monetize them reliably. Lenny says this is exactly what he did, with no plan, and that his newsletter has its name because it was the default when he signed up.

Lenny adds that starting is easy and continuing is hard. Monga says Substack wants to help with that without cheapening the experience, and cites guest posts and a pilot community feature he describes as the pub at the back of your Substack, where the writer sets the tone and rules. Automating discovery or posting would work in the short run, he says, but would kill the promise of putting writers and readers in charge.

On common mistakes, Monga says writers worry too much about how their audience will see them: whether three emails a week is too many, whether asking for payment is crazy, whether they can take a vacation. He argues that paying subscribers are forgiving and want to support you. A subscription is a costly signal, he says, which is like granting you write access to their brain.

Lenny adds three points. It is not too late. About a thousand paid subscribers brought him around $100K, in line with Kevin Kelly's 1,000 true fans. And it took him nine months of weekly free writing before he turned on paid. Monga points to Kelly's post "You Are Not Late" and says the internet is still early.

In the lightning round, Monga recommends books on architecture and urban planning, because people have spent thousands of years working out how to build spaces for interaction and only a blink of an eye on the internet. His favorite is Christopher Alexander's *The Timeless Way of Building*, which he buys in bulk to give away.

He summarizes the book's premise as follows. After decades of mass-produced suburban housing, Alexander argued that houses feel bad because for the first time the builders are not the people living in them. Monga sees a parallel in the internet, where the people building our spaces work under a complicated incentive structure that leads to poor user experiences, which is the problem he works on at Substack.

He also recommends Substacks outside the tech world, such as Darryl Cooper's podcast and Colin Meloy's behind-the-scenes posts, and credits the product manager Dayne Rathbone and designer Gabriel as leaders on Recommendations. For hiring, he says an early-stage startup needs people with grit, endurance and drive who can run through walls.