Professional management works best in an industry that doesn’t change. Every year AI leaves fewer industries like that.
I listened to David Senra’s conversation with Marc Andreessen. It’s long and covers a lot, but the part I keep coming back to starts with a book from the 1940s.
Two ways to run a company
Marc often recommends The Machiavellians by James Burnham. Burnham described two models of capitalism.
In the first, the founder runs the company and his name is on the door. Henry Ford in the 1920s is the classic example, and Marc’s modern one is Elon Musk.
The second is managerialism, the belief that management is a universal skill and a trained manager can run anything.
That belief built the Harvard and Stanford business schools and the professional CEO who takes over from the founder. In the 1970s it produced conglomerates, where a good manager was expected to run thirty different businesses at once.
Burnham saw this as inevitable. The world had become large, and large systems needed people trained to run them.
For about fifty years Silicon Valley mostly agreed. Once a startup grew, investors brought in an experienced CEO for “adult supervision”. Senra mentions Nolan Bushnell, who founded Atari in his late twenties, as a rare founder who wasn’t replaced right away.
Soup is soup
Marc’s answer is that managerialism works while nothing changes. As long as “soup is soup is soup”, in his words, a trained manager can run the company for a long time.
When the industry changes, the manager gets lost.
His example is a graduate of a top business school working at a rocket launch company. The economics of the whole industry assumed that a rocket flies once. Then SpaceX started landing its boosters and flying them again.
What are your management skills worth at that point?
Where a16z came from
The same argument produced Andreessen Horowitz.
In the early 2000s Marc and Ben Horowitz were angel investors, at a time when, by his count, there were about eight serious angels in the Valley.
Founders and venture funds kept pulling them in to settle conflicts. The funds usually believed the founder couldn’t run the company and should be replaced by a professional manager as soon as possible.
Marc says they were spending about eight hours a day on these disputes.
At some point they saw how absurd it was. They had written a $100,000 check and were refereeing for someone who had written a $10 million one. It was simpler to write the $10 million checks themselves.
a16z launched in 2009. Even then, the idea that a founder should run their own company was controversial.
Founders have to learn too
Marc’s thesis is that it’s easier to teach a founder to manage than to teach a manager to build something new.
He is also honest about the weak side. On day one, founders don’t have that skill. An engineer who spent years in a lab has never run a large organization, and some learn while others don’t.
Mark Zuckerberg had never had a job before Facebook. He learned to manage in public.
Why the debate matters less to me now
We have built Customertimes as a founder-run company for many years, so I’m not neutral here. Even so, founders versus MBAs interests me less than it used to.
With AI, the ground under a business moves every few months. What a team of ten did last year, a team of three does now. A process designed in spring is out of date by autumn.
There are fewer and fewer industries where soup stays soup.
In that world a diploma tells you very little. It won’t show whether someone copes when the rules change three times in one year.
The same test applies to founders. If you build a company for one market for a long time, it’s easy to miss the moment that market changes. A founder can turn into a manager of their own company without noticing.
Last week I wrote about Nik Storonsky’s view of the founder’s job at Revolut: find the next bottleneck, and make whatever worked for one team the standard for everyone. That describes someone who keeps rebuilding the company, and it fits a hired CEO as well as a founder.
What I look for now
When I think about who should lead a business or a new direction, I look past the CV to three things:
When they last shut down a process or a product they had built themselves.
How quickly they take a method a junior team found and make it everyone’s default.
How much the way their own team works has changed in the past year.
People who have already rebuilt a company once or twice tend to handle the next change better.
Founders and hired managers now face the same test, and it comes around every few months.
The full conversation:
My piece on Revolut: https://maxvotek.com/p/the-company-as-a-machine
love love love this, about founders becoming managers of their own companies without realizing it is particularly interesting becasue i found myself in that situation, honestly the real advantage of founder-led companies isnt the founders vision, but their willingness to keep challenging it