Even after the explosive growth of the Internet and computer markets, software is still a tiny portion of the American economy. It’s hard to know the exact percentage, but call it 1%. If this seems low, remember that we know that programmers are only about 1% of the workforce, and most companies employing programmers sell something other than software.
To start a company in the other 99% of the economy, one needs large amounts of capital. Any sort of mass production, or real estate development, or hardware engineering, or original lab research, requires millions (sometimes much more). And most smart, ambitious people don’t have millions, which is one reason why they’ve flooded into Silicon Valley over the last few decades, instead of building Manhattan apartments or starting car companies. The capital requirements are way lower, so it’s easier to go into software, and lots of people do.
But the flip side of that is those who’ve started successful software companies are in a very special position. Out of everyone with their level of intelligence and determination, they’re in the small fraction with access to large amounts of capital. And that small fraction has the other 99% of the economy all to itself. Instead of competing with Larry and Sergey, you could start a company in any of a zillion different fields, and compete with MBAs who never passed calculus.
Empirically, this happens very rarely. (Elon Musk is notable as one of the few exceptions.) Why? Are people, after they get rich, content to just sit around and do nothing? It seems unlikely, because people who founded software companies often start other software companies. From an article about this phenomenon – “Can you name any non-serial entrepreneurs? I can’t think of any of the top of my head… all entrepreneurs seem to be serial offenders. Max Levchin, Evan Williams, Justin Khan, Steve Jobs, Jack Dorsey, Mark Pincus… (and a dozen other names).”
So, why no second hardware startups? The most likely reason is probably the obvious one: after spending a lot of time in software, it’s hard to go learn a whole new field, especially if you’re already rich and don’t have to work to pay bills. But this means that, for anyone in that position, there are huge piles of money sitting around for anyone who wants to get to work claiming them.
(Note: By “hardware”, I don’t just mean silicon chips, but anything that’s physical and capital intensive, like how most small businesses are.)
The Rationalist Conspiracy 
Perhaps time is a major factor, as well. Software can be designed, built and deployed in months, while the relative complexity for these processes for anything physical is pretty high. I think many entrepreneurs, who typically self-characterize as risk-takers, are actually still quite risk averse – hence the flocking around Silicon Valley where the biggest risks are informational and temporal.
I really believe that the retail space is the next arena to be leveled up to ‘2.0’. American strip malls are such an anachronistic and inefficient use of today’s resources. They are okay at attracting foot traffic, but have stagnated completely when it comes to customer service and overall user experience (outliers, notwithstanding).
I moved up to the bay area at the beginning of the year to pursue entrepreneurship, and I’m very much committed to the notion of stepping up America’s retail game. Very seldom do I feel excited to go shopping at a strip mall. But, as you mentioned, hardware is a lot more capital intensive, which is hard to come by for a fresh college graduate. I can’t help but still feel that it’s still easier now, though, to acquire capital if you are determined. Kickstarter, Show HN, reddit, youtube are all resources that I hope to use to attract investors. With Source Filmmaker just released, it’s also easier to make a 2-minute attractive video that helps to illustrate one’s ideas. I think that’s the route I’ll be taking.
Nonetheless, I’m in it for the long haul, and hope to make my own dent in the universe!
P.S. I believe the spelling to Justin’s last name is ‘Kan’, instead of ‘Khan’
P.P.S I’m not Justin Kan, just a similar name
I thought “Retail 2.0” already existed and was called Wal-Mart.
“Instead of competing with Larry and Sergey, you could start a company in any of a zillion different fields, and compete with MBAs who never passed calculus.”
beyond this comment being hilarious, the reason ppl compete in the software world is that it scales very well. in many bricks & mortar (B&M) businesses, income does not scale very smoothly and is highly dependent on lines of credit, equipment financing, etc. i can speak to this from the perspective of having achieved B&M success, only to realize that your income tops out in ways that software does not.
i spent years of my life learning math and physics and i don’t feel it puts me at a significant advantage over smooth-talking-douchebag-in-a-suit ppl from other B&M businesses. B&M businesses often rise and succeed by merit of things besides how smart the principals or employees are, e.g. how good someone looks in suit, a firm handshake, knowing the right ppl. competing with these sorts of people is depressing in a special way that continues to drive ambitious youngsters to work on software.
If you build something physical, you might get sued.
I would have to agree with Jake. I did calculus at university and I don’t feel like this gives a significant advantage over the suits. I think second hardware setups fall down because of demand and a focus on software.
I think there are two important, and inter-related reasons:
1. The search space of bankable projects in non-software fields of endeavor has been covered much more thoroughly, such that growth is less driven by innovative deliverables, and more driven by innovations in rent-seeking;
2. Products or services involving physical presence have greater (and less predictable, and less linear) marginal costs, which gives encumbent players in non-software markets more of an advantage over startups.
The market outside of software sends a lot of signals that less needs doing; most of what’s needed is already being done in a way that makes sense, by firms that have done a lot of it, at razor-thin margins, for a very long time.
It’s no coincidence that investors fleeing the collapse of the tech bubble fled into real estate: little else seemed to be growing. Real demand, even for houses, wasn’t growing much, so now there’s an oversupply of almost every physical thing in the US. Even demand for gasoline has declined for the past four years.
According to this spreadsheet, application software is a highly profitable industry:
http://biz.yahoo.com/p/sum_qpmd.html
Makes sense when you think about it, too. A randomly chosen pair of software companies is highly unlikely to be competing with one another, and once you’ve written the software it’s easy to make copies of it.
It seems to me that the basic issue here is whether one believes that the higher average intelligence of start-up founders would enable them to outcompete in other business fields. I’m pretty sure it would, if and only if they were also willing to acknowledge and develop the skills that currently cause those fields to be dominated by dopes.