Why you really don't want just one vendor running an open source project

There's a lot of money to be made by controlling open source, but far more can be made with community-driven open source.

When someone calls out Linux and Hadoop as two multi-vendor open source communities that have "made commercialization of the technology extremely competitive and difficult," it would be reasonable to wonder what planet they live on. After all, as Henrik Ingo challenged, "Surely those are the two biggest and most successful ecosystems???"

Joseph Jacks, who made the first statement, is active with the Cloud Native Computing Foundation. He's not a newbie to open source. In arguing for single-vendor open source "communities" and their allegedly superior economics, he has perhaps unwittingly argued for (one) winner-takes-all when far more money is available in (many) winners-take-much markets.

But first, here's what we're not talking about.

Open source economics 101

We're not talking about whether you can make money with open source. Red Hat, MongoDB, Cloudera, Hortonworks, and other companies have all gone public selling software or services around open source software, and make hundreds of millions—to billions—in the process. This is why it's bizarre to me to read interviews like this one with Juniper CTO Bikash Koley, and read statements like "open source is not free." This is a discussion for 1998, not 2018.

No, Jacks isn't dredging up the "Can you make money in open source?" debate. Instead, he's trying to make the case that the so-called Open Core model—wherein a company offers an open source project and then builds and sells proprietary software around it—is superior to any other when it comes to making gobs of money with open source. As Jacks wrote that "open core is the dominant business model used by the most successful commercial OSS companies."

Jacks backs this with a spreadsheet that purports to show Open Core vendors piling up $110 billion in value, with pure-play open source vendors netting a mere $30 billion.

SEE: Linux distribution comparison chart (Tech Pro Research)

There are a few problems with Jacks' calculations, the most egregious of which is how he tallies up "value." In Jacks' estimation, Red Hat's value is $24 billion, as that's its current market valuation. Such a mechanism, unfortunately, only measures what investors think the company is worth, not what customers closest to the software may believe.

I think the better option would be to calculate revenue, which gets us closer to a company's real value. Red Hat, for example, is on a $3.2 billion run rate right now. Jacks grossly overestimates most of the software vendors in his $110 billion estimate. All of them are at $100 million or more in revenue, yet virtually none of them have filed to go public (which strongly suggests they're doing nowhere near $100 million).

Even if they are, this isn't really the point.

Getting rich together

Jacks wants us to believe that the best model is the one that makes a single company filthy rich. To Ingo's point, however, a far better model is one that makes many companies rich (filthy or no). And not simply software vendors.

Take Linux. Red Hat got to $2 billion selling value around Linux (support, a hardened distribution, etc.). But while Red Hat made that money, a host of others, including Oracle, IBM, and many more made tens of billions selling hardware, software, and services around Linux. If we use Jacks' calculus of company valuations to measure total value, Linux has driven multiples of that $110 billion valuation he gives to the Open Core crowd.

SEE: GitHub: A cheat sheet (TechRepublic)

For a more recent example, look at Kubernetes. No single vendor controls it, though Google and Red Hat contribute most to it, and are in pole position to profit thereby. With companies like Chik-fil-A and Toyota pushing Kubernetes into the fabric of their infrastructure, too, we can expect to see a range of companies chasing the "tens of billions of revenue" that Red Hat CEO Jim Whitehurst explained is available to them (and, presumably, to others).

It's not winner-takes-all, or even winner-takes-most. Such multi-vendor open source communities can generate many winners driving huge value for themselves and for their customers.

Is there money in tightly controlled, single-vendor Open Core businesses? Sure. But it pales in value and influence compared to true, community-driven open source projects.

Also see

Image: iStockphoto/wattanachon

About Matt Asay

Matt Asay is a veteran technology columnist who has written for CNET, ReadWrite, and other tech media. Asay has also held a variety of executive roles with leading mobile and big data software companies.

Editor's Picks

Free Newsletters, In your Inbox