On September 1, 2024, Y Combinator founder Paul Graham’s latest article “Founder Mode” swept Silicon Valley, triggering a large number of company founders to forward it on a large scale.

For example, Musk liked and retweeted it, saying it was worth reading.

For example, Arpan Shah, a member of the founding team of Robinhood, forwarded his agreement, “One of the reasons why simple delegation doesn’t work as the company scales is that most of the actual work is done on the leaves of the tree. It is impossible to accurately play the ‘telephone game’ with the CEO through the process. All effective companies have some solutions. I often see some trusted people helping the CEO eliminate the noise in the organizational chart.”

The full text is as follows:

Last week, at a YC event, Airbnb co-founder Brian Chesky gave a speech that was unforgettable for everyone in the room. Afterward, most founders I spoke to said it was the best speech they had ever heard. Ron Conway (note: SV Angel founder and Airbnb investor) forgot to take notes for the first time in his life. I won't try to reproduce it here. Instead, I want to talk about a question it raises.

The theme of Brian’s talk was that the conventional wisdom about how to run a big company is wrong. As Airbnb grew, well-meaning people advised him that in order to scale, he had to run the company a certain way. Their advice could be optimistically summed up as “hire great people and give them space to do their jobs.” He followed that advice, and it was disastrous. So he had to come up with a better way himself, and he did so in part by studying how Steve Jobs ran Apple. So far, it seems to be working. Airbnb’s free cash flow margins are now the best in Silicon Valley.

The audience at the event included many of the most successful founders we’ve ever funded, and one after another they said they had been in the same boat, that they had been given the same business advice as their companies grew, and that advice had hurt them instead of helping them.

Why was everyone telling these founders the wrong things? This was a big mystery to me. After some thinking, I found the answer: they were being told how to run a company you didn’t start—how to run a company if you were just a professional manager. But this way was so much less efficient that the founders felt it was not right. There were things that founders could do that managers couldn’t do, and not doing these things made the founders feel wrong, and that was the truth.

There are actually two different ways to run a company: the founder model and the manager model. Until now, even in Silicon Valley, most people have implicitly assumed that scaling a startup means switching to the manager model. But from the frustration of founders who have tried the founder model and the success they have had trying to break free of it, we can infer that there is another model.

As far as I know, there are no books written specifically about founder patterns. Business schools are not aware of their existence. So far, all we have are experiments by individual founders who have been figuring it out on their own. But now that we know what we are looking for, we can go looking for it. I hope that in a few years, founder patterns will be as well understood as executive patterns. We can already guess at some of the differences.

The way managers are taught to run companies seems like modular design because you treat the subtrees of the org chart as black boxes. You tell your people what to do, and they figure out how to do it. But you don't get involved in the details of how they do it. That would be micromanaging them, and that's not good.

Hire great people and give them space to do their work. Sounds great in a nutshell, doesn’t it? But in practice, from what founder after founder reports, this often means: Hire professional scammers and let them bring the company down.

A theme I noticed both in Brian’s presentation and in my conversations with founders afterwards was that of being lied to. Founders felt lied to on both sides — by the people who told them they had to run their company like managers, and by the people who worked for them when they did. Usually when everyone around you disagrees with you, your default assumption should be that you’re wrong. But this was one of the rare exceptions. VCs who haven’t been founders themselves have no idea how founders should run their companies, and C-level executives, as a group, include the most skilled liars in the world. [1]

Whatever the founder model entails, it’s clear that it will break the rule that the CEO can only engage with the company through his or her direct reports. “Leapfrog” meetings will become the norm, not something so unusual that they even have a name. And once you abandon that restriction, a host of options open up.

For example, Steve Jobs used to hold annual retreats for the 100 most important employees at Apple, who weren’t the top 100 employees on the org chart. Can you imagine the willpower that would take to do that at a normal company? Yet, imagine how useful something like that could be. It could make a large company feel like a startup. If these retreats didn’t work, Steve probably wouldn’t have continued. But I’ve never heard of any other company doing this. So, is it a good idea or a bad idea? We still don’t know. That’s what we know about founder patterns.  [2]

Clearly, founders can no longer manage a 2,000-person company the same way they managed a 20-person company. Some level of delegation has to occur. Where the boundaries of autonomy are, and how clear they are, will likely vary from company to company. And that will vary over time, even within the same company, as managers earn trust. As a result, founder mode will be more complex than manager mode. But it will also be more effective. We already know this from examples of individual founders fumbling their way toward it.

In fact, another prediction I made about founder mode is that once we figure out what it is, we’ll find that many founders are already pretty much doing it—just in the way they do it, many people think they’re weird or worse.  [3]

It’s inspiring to see how little we still know about founder patterns. Look at what founders have accomplished, against a headwind of bad advice. Imagine what they would do if we told them how to run a company like Steve Jobs instead of John Sculley.

Notes:

[1] To put it more politely, experienced C-level executives are often very good at managing up. I don’t think anyone who knows anything about this world would dispute that.

[2] If this retreat practice becomes so common that even mature, politically driven companies begin to do it, we could quantify the aging of a company by the average depth of the invitees on their organizational chart.

[3] I have another, less optimistic prediction: once the idea of ​​founder mode is established, people will start to abuse it. Incompetent founders will use founder mode as an excuse. Or managers who are not founders will decide they should try to act like founders. To some extent, this may work, but if it doesn’t, the results will be messy; a modular approach can at least limit the damage a bad CEO can do.