The Founder Mode Tradeoff – By Kent Beck

Though I think this article could be better, Kent is off in the right direction. Paul is unusually ambiguous about "founder mode" in his essay. Implicitly, he was saying that founders have a super power by not just letting good people do good work, that they (the founders) should involve themselves in decisions/discussions/work that professional managers should/could not.

Many of us understand what he's going after, but he really didn't articulate it well. Kent basically makes that point in this essay by attempting to break it down and provide a more concrete foundation for the concept.

My issue with both articles is that they are ignoring the curse of knowledge that these outlier founders have, and from that they are ruining their own paradigm... If you ask a top performing athlete how to be a top performing athlete its likely not going to be very useful advice if its directed at a novice. However in a conversation between those top performers there will be the nuance and the context to what they are trying to express that can make it true.

I'm happy that Kent wrote this article, but I really think PG would benefit from revisiting his article to provide what he actually means to the rest of us.

The founder mode binary is way too simplistic.

There are examples of successful companies all over the founder mode spectrum.

What level of "founder mode" is optimal depends on a lot of factors. For example:

- size of the company - available people (at every level of the company) - type and number of products (is it possible to scale the work over different departments?) - qualities/personalities of the founders (trust me when I tell you, being micromanaged by a non-technical founder on a technical product is not effective) - type of business (for example; in aviation, you're going to need a lot of certification management)

And you can execute poorly all over the spectrum, even when making the right choice on the level of "founder mode".

As I see it, certain founders can function well in founder mode due to their knowledge, personality and luck in term of business positioning. Most can't. Most founders also won't make $100b companies. The latter type of founder trying to work in founder mode will make things much much worse and often result in going from a $1b exit to $0b exit.

Gee whiz. I guess this is just confirmation that postings are simply fodder for each person's own conceptual model of the world. My reading of Paul's post was something entirely different. What I took away (as opposed to whether it was in the article or not ), what I took away was the idea that founder mode is about the rejection of orthodoxy and the triumph of what I call "original thinking".

We humans can be the next best thing to lemmings. We do what we see people around us do whether and what we learn in class and from others who we think know something. Is that because we are lemming like or is it because it allows us to deal with the overwhelming chaos of the world?

My best illustration of this: I had two friend who went to Nepal for two years. This happened a long time ago. They came back to the US and went to a market and nearly had nervous break downs. In Nepal, they had soap. One kind. In the US they had 200 kinds of dish soap alone. They had no idea which soap to choose, and that extended to almost every product they needed.

Our follow instinct - if you see someone else pick up a kind of dish soap, get that - helps us deal with this overwhelming complexity and does well for us. Except when it doesn't.

What happens when you are faced with something new? In lemming mode we tend to go wrong when something is new. Really new. Like networking was new. Or little mp3 players (you know - those things that became iPods then iPhones and all the little iPhone killers like the zune.)

My take away then was about something completely different, although probably unrelated to either Paul's article or this response.

If you are to be a founder, one of the biggest challenges will be to build an organization that does not descend into orthodoxy.

[side note: I think founder mode is a misnomer because you can be an original thinker/doer without being a founder] [side note: There are really two situations where you want to be in original thinker mode: when there is something new or when the environment changes dramatically. We are in a time where the latter is true]

> I think Paul is essentially right but that the way he expresses his idea is shallow, ineffective, & prone to cause unneeded suffering.

I agree with this so much. I'm a huge fan of PG and his essays have been instrumental for me since I was in college. But that Founder Mode essay was so shallow. It's not a dichotomy. You don't choose between "black box management" and tyrannical micromanager.

You hire smart people. You give them some room, but not black box. You trust but verify. Sometimes you have to be involved in the details, sometimes you don't. It is extremely situational. But that's just management. You have to decide based on the situation how involved you need to be.

So much suffering will result because of misinterpretation of that essay. Founders everywhere thinking "I knew it! I should never trust anyone. I should do everything myself". Yea, good luck with that, Steve Jobs.

Note: I'm a founder, several times over. I've also worked at a FAANG and at mid-stage startups for other founders. Obviously I'm no PG, but I think I've seen multiple perspectives.

> Founder CEOs have 3 (at least) advantages that give them a unique ability to create value (i.e. increase the survival chances). [...] Incentives. Professional managers are also playing an iterative game, looking to their performance in this gig to set up their next gig. The Founder CEO is all in on this hand.

I want everyone in an early startup thinking like this.

Which also means it should be treated as the team's company -- not only that of the founders -- and there should be life-changing money and career growth for everyone if this is successful. (Not a pittance in ISOs, and not pushed back to scratch, practicing Leetcode hazing for the chance to work in the next someone-else's company.)

Appreciate the added depth around the addition consequences of founder mode.

I worry that 'founder mode' will be taken as 'be an a-hole' v2 or something as v1 was to mimic the success of Steve Jobs.

Being an a-hole was an aspect of Jobs but not the source of his success. Being an a-hole is not itself a gateway or required attribute for success as was the popular takeaway there.

Micromanaging, buzzing around, and bypassing structure should not be the takeaway from 'founder mode' or you will have bad leaders excuse those tendencies and chalk it up to being in 'founder mode' after graduating from 'a-hole'.

I just told my wife about "founder mode," and she asked me 100% seriously if I was talking about an article from The Onion.

For the win.

Let's give founder mode the mathematician's treatment, generalizing it to something more well understood and then reexamining the specifics. To my ears, the discussion around founder mode can be heard as a discussion around a qualified principal-agent problem, where the principal is the founder and the agent is anybody and everybody else the founder has to bring on in order to further the founder's vision and desires for the business.

>The principal–agent problem refers to the conflict in interests and priorities that arises when one person or entity (the "agent") takes actions on behalf of another person or entity (the "principal"). The problem worsens when there is a greater discrepancy of interests and information between the principal and agent, as well as when the principal lacks the means to punish the agent. The deviation from the principal's interest by the agent is called "agency costs". - https://en.wikipedia.org/wiki/Principal%E2%80%93agent_proble...

This isn't a conflict that can be escaped, only managed, something inherent within any large organization, notwithstanding the specifics of the startups in discussion here. Hiring employees is hard and costs money that could otherwise be spent on the business or dividends back to founders and investors, and if founders could avoid having to hire others, they would. But, founders having a limited amount of time and specialization, must hire others lest the business die on the vine as opportunities pass them by for lack of available people/time to perform the necessary labor to take advantage of them.

The second and third sentence in that definition from Wikipedia is why I am writing this comment. Within the terms introduced above, I think PG is advocating that founder's should work harder to reduce agency costs by getting better at punishing their agents.

Punish is a strong term here to use for sure! However, if you ask the agents, i.e. the employees of the founder, much of the behavior advocated by proponents of founder mode is punishment. Micromanagement of projects, subverting the established lines of communication, second guessing the details of the agents' work and opinions, all of these things are often seen by agents/employees as undermining their agency and feelings of security within the organization.

Ask a product manager whether they actually want the founder to read and comment on all of the technical specs and reports the PM and their team produce and I imagine most of them would grimace. They would much prefer if the founder stuck to being given the "right" information by the PM and the founder presenting their feedback in the right settings/meetings. Going outside of that would restrict the agency of the product manager and might very much feel like a punishment. I've most often seen founders/higher ups get involved in a team's work directly and it was often as the result of the team fucking up and the "principals" feeling like they had to step in, do some light to heavy punishment through their involvement, and set things right.

Punish is still a harsh word to use here, but there is a fundamental conflict of interest that can't really be escaped. I'll make the assertion that most employees do not directly share the founder's interest in making the founder wealthy. Employees have their own interests, whether it be working on something cool, furthering their own careers, getting their own slice of wealth, or just having a steady job while they live life outside of work. The founder wants tools to win the conflicts that occur between the founder's desire to increase their wealth and the employees sometime desire to do an action that, in the founder's opinion, harms the chance of that happening.

To me, in my eyes, founder mode has been a implicit discussion of this conflict at the heart of the principal agent problem, with both sides talking past each other about the necessity and efficacy of common punishment techniques. Principals keep saying they need more ways to correct deviations from their visions and agents replying by saying that often times those corrections do more harm than good. Neither of them are wrong per se, but it does feel like a misunderstanding between the two groups that'll just cause more harm than good in the long run. Founders giving themselves leeway to indulge in punishing their employees for perceived deviations from the founder's interest will likely end in nothing worthwhile, while agents/startup employees resisting outright or implicitly any punishment/correction from the founder will likely cause the employee to get even more punishment.

The ideal successful founder experiencing success goes through a rapid metamorphosis from exceptional individual contributor, to team lead of a small team, to team manager of a large team, to a manager of managers of many teams, to department head of many such managers, to chief executive officer of many hundreds or thousands. At each stage, the founder needs to learn how to correct and prevent their employees from deviating from the founder's desired outcomes.

A lot of the replies from employees has been from within the context of their experience at the hands of founders who did not manage this transition and learning process well. The CTO that had to approve every pull request of the hundred person technical team, the CEO who had to interview every incoming employee and slammed the recruiting process to a halt, the founder who showed up to the team meeting and threw the product road map away and started over to the dismay of the engineering teams. Each time, the founder trying to keep the organization headed in the right direction, i.e. making the founder and the investors wealthy, but not really knowing how to do this well and fucking up in the process.

MBA's have a set toolbox for how to punish at scale in very large organizations that has been refined over centuries. Founders have tried applying this to their startups and found them to be lacking, too heavy handed and lacking in refinement for the smaller scales they are often operating on. Eventually, the founder hopes that they might lead an organization successful and large enough that they might need to use such tools, but for now, they are stuck making it up on the fly while also trying to do literally everything else. Part of the discussion of founder mode to me has been around founders having a discussion about how to actually effectively make these corrections at various stages of growth of a business.

I'll wrap this up by restating PG's point: the best founders know how to punish in ways that don't scale. The principal-agent problem will always be there, cannot get around it. My hope is that the discussion around founder mode is less about founder's giving themselves license to do whatever they feel like whenever they feel like it, and more around the development of a toolbox for founders to punish employees for deviating from the founder's goals that precludes what could broadly be considered abusive or ineffective.

FWIW the whole "Founder Mode" storyline just sounds like the further extension of Elon Musk's radius of influence into the heart of SV outright.

Of course there are upsides to having leadership that actually cares and is not afraid to get their hands dirty when they spot a problem wherever it may be.

That's not new but I can fully get behind that. Outside of shareholder value, companies and their products actually could care much more, always.

I'd wager Steve Jobs fell just more onto a spectrum there rather than this more discrete proposal.

My guess is, this will likely follow the twitter->x trend we've been observing since 2022 – and as either investor or worker it will be even more about picking the character of a company.