What I learned quitting my job to found a tech startup

I’ve been earning to give as a software developer for the past several years, and it started to become clear that I could make more money in a different job. But I was torn between a finance career which put my math skills to use and founding a company where I might achieve the vocational equivalent of winning the lottery.

I eventually decided to pursue entrepreneurship because I thought it would better build career capital, i.e. it would prepare me for a wider variety of future careers. After four months of running a company that idea still doesn’t seem completely idiotic, but it doesn’t seem completely true either.

I’ve encountered several people who are in similar positions, so I’d like to give an overview of my motivations (particularly the ones which haven’t been discussed here before), how I went about my career change, and of course how I should’ve gone about my career change. Optimizing for one narrow career path is a bad idea, so I hope this post is useful to everyone, not just potential entrepreneurs.

Motivations and The Founder’s Dilemma

The founder’s dilemma contrasts the two basic motivations of entrepreneurs: to make a lot of money and to be in a position of power. Noam Wasserman calls these outcomes “Rich” and “King”. The “Rich” goals have been elaborated on before, but I would like to give a brief summary of some reasons why King outcomes are interesting for EAs.

First off, it might be unclear why “Rich versus King” is a dilemma much less “the” dilemma – why can’t I get rich and remain in control? Don’t CEOs make exorbitant amounts of money?

Those people are exceptions – if you want to build a great company you’re going to need to hire great people. Those great people aren’t going to be your lackeys. The only way to keep control for yourself is by working with subpar people, and subpar people don’t build great companies.

Fun fact: the more successful you are as a startup CEO, the faster you will be kicked out of the job. Why? Any jackass can run a failing startup, but it takes top talent to run a big company. And you are not top talent.¹

So why was I motivated by King outcomes? Exploration value.

My direct employment career path was a progression from “writing code” to “telling people to write code”. As a founder I deal with legal, accounting, customer development, business partnerships… It turns out that I actually enjoy some of these things and would probably pursue them even if I decide startups aren’t a good idea. I never would’ve found that out at my previous job.

Additionally, I meet more interesting and influential people in a given week of startup networking than I would in a year at a big company. I would give a high probability to my next job being through one of these people.

Networking tip: do not mention the founder’s dilemma to other startup CEOs. “You just closed that big round of funding? So I guess you’ll probably be fired in the next few years!”

By taking the startup option, I got to try my hand at a whole range of tasks which I wouldn’t have otherwise. But as my company grows, I need to delegate to others and prioritize my own work – if I was focused on growing career capital, I would hire people who have skills which I already have, freeing me to learn more. But since I wish to focus more on the financial aspects of entrepreneurship, I am choosing to work with people who have complementary skills instead, which means I’m experiencing a narrower array of tasks.²

How I went about it

I toyed with the idea of starting a company for a good deal of time before actually becoming serious because, well, I’m not insane. Fortunately, there are a lot of changes you can make to your career which are robust improvements not just for starting a company but for other goals you may have.

The first change I made was to double down on my efforts to move into management. This is great not just for prestige and salary reasons, but it will teach you project management skills and the ability to communicate your vision to others. There are few career paths where this isn’t valuable.³

The second change I made was to focus more on customer interactions. This was actually a move down the career ladder at my employer, and in retrospect it wasn’t very useful. I think being able to talk to people, understand their needs and build solutions which genuinely help them is a tremendous skill. But the “customer-facing” jobs seemed to center more on handling complaints about decisions which were out of my control.

As I finalized my decision to leave, I used my remaining vacation time to spend one day per week researching startups (my “Weekend Wednesdays”). Terrified and excited, I turned in my resignation and became unemployed self-employed.⁴

How I should’ve gone about it

This may be hard to believe, but my approach wasn’t perfect. Here are some major things I should’ve done:

1. Meet more people. I think it’s hilarious and depressing that the highest technology companies live and die based on a personal networking process which hasn’t changed since the Middle Ages. A startup meets all its customers, business partners and investors through personal introductions. I should’ve focused on growing my network as much as, if not more than, growing my skills. (Social capital is a strong predictor of entrepreneurial success.)

2. Remove the uncertainty. My biggest fear was not knowing what it even meant to start a company. Do you just buy a domain name and then wait for the money to start rolling in? Startups are so vogue right now that every major city has co-working spaces and incubators up the wazoo – they’ll lay the baby steps out for you and give you the social proof your system one craves.

3. Be a specialist. Startups don’t succeed because they are better than average in a lot of dimensions, they succeed because they are vastly better than average in one tiny dimension. My greatest asset so far has been my knowledge of this one topic so esoteric that I personally know most of the people who understand it. (I haven’t found much research on this, but industry experience does seem important.)

In the startup

I spent the first month not really knowing what it meant to run a company. (I still don’t know, but I can now pretend well enough to write blog posts on the subject.) I didn’t have many ideas for what I wanted to do, and the ideas I had turned out to suck. To keep some illusion of progress I wrote a small app which I thought was trivial – it turns out that other people didn’t think it was trivial (see #3 above about specialized knowledge) and that has now turned into the focus of the company.

I’m hesitant to describe what it’s like running a company, because my description will be “hard” and then we will have even more of a bias towards overconfident people starting companies. Additionally, the things which seem to indicate that a person will be a good entrepreneur verge on the tautological. (“Oh, you’re a hard-working self-starter? Then we don’t want you here at ACME Corp.”) One interesting predictor though is a “break the rules” attitude:

“People who both engaged in illicit activities as teenagers and scored highly on learning aptitude tests… tend receive much larger increases in earnings when they become incorporated self-employed business owners than people without that combination of traits.”

• Smart and Illicit: Who Becomes an Entrepreneur and Does it Pay?

One thing which inspired me is that a rising tide lifts all boats. In less poetic terms: it’s better to be a crappy company in a growth industry than an awesome company in a contracting one. I happen to have a lot of knowledge about an industry that most analysts predict to grow (healthcare), so tying my compensation to the company’s performance made sense.

Am I getting career capital?

80,000 hours’ current recommendation about entrepreneurship says this:

Starting a tech company will force you to rapidly learn highly useful entrepreneurial and technical skills, meet lots of people, and it looks impressive (even if your business eventually fails, at least in failure-tolerant cultures such as the US and UK). Overall, we think it’s one of the best options for career capital.

Breaking that down:

• Starting a tech company will force you to rapidly learn highly useful entrepreneurial [skills]
  • Anecdotally this seems completely true, but statistically it seems false. I think there must be some fraction of people who do rapidly learn in this environment, but I have no idea how to tell if you are one of them. The fraction of your knowledge which was self-taught might be one good proxy.⁵
• [You will also learn] technical skills
  • Most of my technical time is spent doing grunt work that would be outsourced to lower skilled employees in a big company. If you are a non-technical person this may be a step up for you, but you will definitely learn way more as a software developer at a big company.
• meet lots of people
  • 100% true. Not only will you meet lots of people, but you will get meta-skills like understanding how to ask for introductions, and you will become familiar with a ton of potential future employers.
• and it looks impressive (even if your business eventually fails, at least in failure-tolerant cultures such as the US and UK)
  • People I talked to seem to be impressed, but entrepreneurs who return to direct employment make pretty much the same as those who never left, which implies that employers don’t find it much different than direct employment.
• Overall, we think it’s one of the best options for career capital.
  • It’s definitely possible to run your business so that this is true. Any part of business that I want to get better at I can do myself, and hire other people to do the rest. But it’s critical to realize that this is not great for your financials.

My basic point is that there is no such thing as a free lunch. The more time you spend learning technical skills, the less time you will spend learning entrepreneurial skills, and both of those will take away from your potential earnings.

My experience so far is that entrepreneurship does have career capital advantages over employment as a software developer, but those advantages aren’t as extreme as 80,000 hours implies.

(Note: 80,000 hours is currently in the process of redoing their recommendations, and their new recommendations will reflect a lot of what I’ve said here.)