Paul Graham, Dropbox and The Single Founder Exception

First, he had a second "founder" before the program started, so it wasn't single-founder.

Second, luckily the second guy turned out to be "perfect". This could easily have backfired. It is far more likely that adding someone as a founder who you haven't worked with before will torpedo the company. IIRC, YC made this point previously.

So actually it was bad advice, like hitting on 17 in Blackjack, or not getting the right pot odds to call in Hold'em, and then calling anyway.

If you win, it's just dumb luck. If you try running the experiment a hundred times, adding an extra unknown person not only to a promising single founder but also trying to make promising two-person groups into three-person groups, you're going to decrease almost everyone's chances.

Third, there is pronounced over-romanticization of founders. It is extremely common for the founders to NOT be the most important cogs in their own companies. It is very common for companies to be started by people who have money but not skills, and of course then the people who DO everything are non-founders. It is also common for founders with skills to have employees with much better skills at the same thing. And even the decision-making is frequently done by advisors who guide the young and inexperienced founders on every key issue regarding having a company.

Tom was just the PR cover for MySpace. And who do you think really runs Facebook? Or Apple in its early days? Woz never ran engineering at Apple, and Jobs didn't run the company until his second coming. The gee-whiz kids in the dorms and garages are just the figureheads to market to their demographic. Just like Hot Topic markets to people who want to make a show out of how they refuse to conform by buying non-conformist uniforms in malls.

And dedication, sure. But some companies get lucky in less than a year, like Auctomatic. PG had to work for THREE years before he could retire. Drew has been going for four.

But dedication isn't as rare as it seems. Just look at the factory workers who work harder jobs for longer hours year after year making their Macbooks.

Overemphasis on founders hurts companies. One of the reasons it's so hard to hire is that founders think it's "generous" to offer a fraction of a percent equity to someone they desperately need to do the thing the company was founded to make in the first place.

IIRC, RethinkDB is/was in this boat. If you don't start with the needed expertise, and you find that you need to hire it or else you don't have a product, why would someone come work for you for 0.5% and average salary when by definition he could just start his own company doing the same thing?

So there are a zillion startups each with a handful of people at most and they all complain about the scarcity of talent, while they are offering the same or less as big companies that stick you with less responsibility, fewer hours, and no worries about burn rate.

Being a founder isn't special. What's special is solving hard problems and making the decisions that lead to a successful product or useful technology. That is orthogonal to being a founder.

No, really. Founders aren't special.

SBA.gov:

> An estimated 627,200 new employer firms began operations in 2008, and 595,600 firms closed that year. This amounts to an annual turnover of about 10 percent for entry and 10 percent for exit. Nonemployer firms have turnover rates three times as high as those of employer firms, mostly because of easier entry and exit conditions.

If you want a successful company, come off your high horse and spread out the equity. I've seen promising companies bite the dust or spin their wheels for years because the founder(s) couldn't bear to part with their equity, and end up holding close to a hundred percent of nothing.

Basically the YC model stems from PG recreating his experience with ViaWeb. He started a company with his best friend and another expert programmer, so that's the model that seems like the "right" model to them.

Others have very different experiences working with other people (e.g. the much-maligned group projects in college) so they have a different mental model of what works and what doesn't, at least for themselves.

There are extremely valid arguments both ways which largely boil down to individual variation, but here's a key point that gets overlooked: Perhaps if you haven't done it as a single founder, you don't know what it takes, and so you don't know what to look for.

He has done it in a group. So if you're good at picking twos and threes, or feel you are, you can fill up your whole complement with those.

And then it becomes self-reinforcing. You get more and more confident with the one than the other which is avoided. Some pitchers rely on their fastball and don't develop their slider.

So eventually you're looking at, let's say for sake of illustration, 90% confidence in picking groups, but hardly any confidence still in picking solo founders.

So when you're looking at the last slot, do you want to add another group you are 90% sure of, or say no to someone you are excited about and go with some solo founder you are dubious about? And what about the second-to-last slot? And all the way back... Surely your FIRST choice isn't going to be a solo founder!

Even though YC claims not to have a fixed limit, obviously there is a practical limit. And when they say they fund everyone they find promising, you have to take it with a grain of salt, because in their model, being a single founder already isn't promising.