Startups So Cheap You Might Not Notice Them

Maybe we're seeing less investment because entrepreneurs don't need much anymore.

Manufacturing on the cheap.

Photographer: PAUL J. RICHARDS/AFP/Getty Images

It costs a lot less to start an Internet company than it used to. This is venture capitalist Mark Andreessen, speaking in 2014:

 If you wanted to start a new Internet company 10 years ago, you probably needed to raise $20 million, just to get started. You would spend $5 million on Cisco routers and $5 million on Sun servers, and $5 million on Oracle software, and then you’d write Yahoo a $5 million check to get distribution. And then you could try all your new ideas.

Today it’s advanced to the point where entire companies can get up and running -- to provide global services, in some cases to millions or tens of millions of people -- for less than a million dollars. 

It costs a lot less to start certain kinds of manufacturing companies, too. Here’s Mark Hatch, chief executive officer of the do-it-yourself workshop chain TechShop, whom I talked to Thursday:

Pre-TechShop it was a million to a million-and-a-half dollars to start a hardware company. Now you can start a hardware company for $5,000.

You get your tools -- and often the skills to use them -- at TechShop, Hatch went on, raise whatever capital you need on Kickstarter or Indiegogo, and reach customers via Etsy.

There lots of exceptions to this cheap-startup rule. I spoke with Hatch during a break at O’Reilly Media’s Next:Economy conference. A few minutes later, Saul Griffith of Otherlab was on stage describing a company he co-founded: wind-turbine maker Makani, now owned by Google. “In 2006 we said it will take 10 years and $100 million,” he said, and that’s about what it will work out to be when the company goes into production next year.

There are also Internet companies, if that’s what you can call Uber or Airbnb, that have felt the need to raise billions of dollars to bankroll their efforts to conquer the world. Our economy isn’t being entirely taken over by bootstrapped virtual startups or Kickstarter-backed "makers."

Such low-overhead enterprises do exist, though, and there are probably getting to be enough of them to have an economic impact. So when somebody points out that alarmingly little money seems to be going into investment in general and tech investment in particular in the U.S. in recent years, as economist J.W. Mason did in a Roosevelt Institute report that I wrote about Wednesday, it’s worth pondering whether this is a question of demand as well as supply. That is, there may be less demand for capital because many once-capital-intensive aspects of starting and growing a business are now cheap or free. Back to Andreessen:

A lot of the Internet startups that we see coming through here raising money are startups that raised a half-million dollars, that still have most of it in the bank. It’s four kids and their laptops, and the entire company is run on the cloud, on Amazon Web Services, on, NetSuite and Gmail. These companies have effectively no capex. It’s literally their laptops and their ramen noodles, and that’s it. And they walk in the door, and sometimes they have five million users in 170 countries. And that’s just a phenomenon that’s never existed before.

Look at the economy-wide numbers, and we appear to be in an era of low investment, low productivity growth and declining entrepreneurial activity. Hang out in or near Silicon Valley (I’m in San Francisco right now), and we appear to be living in a time of epic excitement and upheaval and ferment. Part of the disconnect may have to do with the fact that we measure economic activity mainly by money spent and people employed. Technology-enabled startups that spend very little money and hire few people may go under the economic radar, at least for a while.

This column does not necessarily reflect the opinion of the editorial board or Bloomberg LP and its owners.

    To contact the author of this story:
    Justin Fox at

    To contact the editor responsible for this story:
    Jonathan Landman at

    Before it's here, it's on the Bloomberg Terminal.