The technology boom of the 1990s ended in a spectacular wave of failure, as huge swaths of the sector proved to be based on fundamentally unsound business models.
Today's business models look, on the whole, to be much stronger. And that's why I've never believed the dozens of people who've spent the past four years crying bubble. But Conor Sen of New River Investments has a very smart post making the point that the expansion cycle can fail for other reasons. In particular, he says, "there’s just not enough 'stuff' to go around."
To keep growing, the technology sector needs to keep adding skilled workers and office space, and both San Francisco and Silicon Valley are running out of both. That means that even companies with perfectly plausible business ideas are going to find costs rising faster than revenue, and the sector is going to need to stall out until some companies fail and free up more resources for the others.
Nobody wants to rent to startups
Jason Freedman from the real estate website 42Floors has a fascinating observation that could spell the end of the current cycle of expansion in the internet startup sector. It's not about an investment bubble or any problems with the basic product of any of today's startups. Instead, it's a funny observation about the commercial real estate market.
Office rents are skyrocketing and vacancy rates plummeting in San Francisco. But there's always more money.
The real problem, he writes, is that many of the city's landlords "are trying as hard as they can to get non-startup tenants right now. We at 42Floors recently experienced this firsthand when we were searching for our new office space."
And it's not about money. It's about risk. "Landlords look at their tenants as a portfolio of revenue streams and risks," Freedman says, and they want to hedge against a possible bust in the startup market by making sure that they don't have a portfolio that's too crowded with internet startups. There comes a point when a landlord is willing to take a significant hit on the rent to be leasing to some lawyers or accountants or architects or anything that won't vanish if a wave of venture capital skepticism suddenly sweeps the country.
It's hard to leave the technology hub
Of course, your company always could try to leave the ultra-hot San Francisco/Silicon Valley real estate market in favor of someplace more affordable. There's plenty of office space for rent in Cleveland or Detroit and no problem building more in Phoenix or San Antonio.
The problem is that almost any move you could make to mitigate your real estate problems is going to exacerbate your skilled labor problem.
The reason is that startups are, inherently, risky places to work. But if you have the kind of skills to get a good job at a startup, you can greatly mitigate that risk by living in a city that's full of similar startups that need people with similar skills. You mitigate that risk, in other words, by living in Silicon Valley or San Francisco.
Which means that if you're a company that needs startup workers, you're going to find that moving to Cleveland to avoid the Bay Area office crunch is going to leave you with a shortage of workers. Heck, even moving to Oakland is going to leave you with a shortage of workers since, at the end of the day, you'll be at a recruiting disadvantage vis-a-vis companies that can offer a more pleasant commute.
Nothing lasts forever
These supply constraints on the startup economy are, of course, surmountable, and given time they will be surmounted. More people are learning to code to take advantage of job opportunities there, and smart companies are making efforts to be more attractive workplaces for a more diverse population in order to be able to draw on a broader talent base.
But unless the Bay Area changes its attitude toward new building permits, the office space crunch will remain, and there will be a fundamental physical constraint on the number of people who can find housing in the technology hub.
Rezoning would be the easiest fix but, in some ways, the most politically unlikely one. Breaking the bottleneck, then, will eventually require some new technology hubs to emerge in places that are more eager to embrace new construction. But until that happens — and it's difficult to pull off — the "only so much stuff to go around" problem will be with us, and technology booms will be somewhat self-limiting.