Red Hat's Billion Dollars And the Power of Free

Recently, there was some justified excitement that Red Hat had finally done it, and turned in annual sales of over $1 billion. A couple of years ago, I wrote a post here on Computerworld UK wondering why there were no companies based around open...


Recently, there was some justified excitement that Red Hat had finally done it, and turned in annual sales of over $1 billion. A couple of years ago, I wrote a post here on Computerworld UK wondering why there were no companies based around open source that had managed to achieve such billion-dollar turnovers, and suggested that the key reason was one put forward by Red Hat's CEO, Jim Whitehurst:

He said that he did think that Red Hat could get to $5 billion in due course, but that this entailed "replacing $50 billion of revenue" currently enjoyed by other computer companies. What he meant was that to attain that $5 billion of revenue Red Hat would have to displace software that currently costs $50 billion. Selling $50 billion-worth of software – even if it only costs $5 billion – is somewhat hard, which is why it will take a while to achieve.

Indeed, I went on:

I would go so far as to say that very few open source startups will ever get anywhere near to $1 billion. Not because they are incompetent, or because open source will "fail" in any sense. But because the economics of open source software – and therefore the business dynamics – are so different from those of traditional software that it simply won't be possible in most markets.

However, I did think Red Hat stood a chance of getting to the one billion level:

Red Hat stands a chance because it has (wisely) colonised the biggest sector, that of enterprise infrastructural products - "we are plumbers", as Whitehurst put it with brutal frankness.

But I remain sceptical that it will get to that $5 billion level any time soon, simply because it will be replacing $50 billion worth of conventional business, and that's a mammoth task.

Someone rather better qualified to comment on Red Hat's achievement is Michael Tiemann, who occupies a pivotal position in free software. As he writes in a Google+ post:

Back in 1989 was so certain that free software could be the basis of a $1B enterprise, I started one just to prove it (

Cygnus – arguably the first open source company – was later bought by Red Hat, and that move led him to the following insight:

In the 10+ years we have been selling Red Hat Enterprise Linux, I have found that for every $1 Red Hat sells, we have to displace $10 of proprietary junk that never really worked in the first place. Rationally, it sounds easy. Indeed, this 10:1 value proposition was precisely what enabled Cygnus Support to bootstrap from an initial $6000 in capital to $20M in revenues in less than 10 years. But practically, it's not easy, because in the mainstream old habits die hard.

In other words, Tiemann confirms that open source brings about the same 10:1 reduction in costs when people move away from traditional proprietary solutions as Whitehurst had noted. Tiemann then goes on to point out the huge potential for saving money if more of the $1.5 trillion the world spends on enterprise went to companies based on open source programs.

Unfortunately, that's probably not going to happen any time soon because of the huge institutional inertia that is present within most enterprises – and the fierce rearguard action from proprietary software companies trying to defend their fat profits. Nonetheless, what might be called the open source effect on revenues can be observed in another high-profile story that has garnered a certain amount of attention recently. Here's what Charles Arthur wrote in the Guardian:

Android generated less than $550m in revenues for Google between 2008 and the end of 2011, if figures provided by the search giant as part of a settlement offer with Oracle ahead of an expected patent and copyright infringement trial are an accurate guide.

The figures also suggest that Apple devices such as the iPhone, which use products such as its Maps as well as Google Search in its Safari browser, generated more than four times as much revenue for Google as its own handsets in the same period.

Android is open source (well, more or less), so that $550 million would represent $5.5 billion in revenue for a proprietary company, assuming the same 10:1 ratio applied here. That sounds reasonable, assuming the other figure quoted – around $2.2 billion for Google's iPhone revenue – is the right order of magnitude.

Given this disparity between Android's $550 million and the $2.2 billion revenue it derives from its arch-rival's iPhone, does that mean that Android is at best a damp squib, and at worst a major failure for Google? Hardly. It's important to remember why Red Hat has any income at all, despite the fact that its core product is freely available. The reason is that Red Hat makes money not directly from selling its flavour of GNU/Linux, but indirectly from the fact that (lots of) people use it. Because it is popular, it has a well-established software and business ecosystem, and Red Hat can make money from ancillary products and services in that ecosystem.

The same is true for Android. Google doesn't – and can't – make much money from Android directly, but the simple fact of its popularity places it in a very strong position to make money in other ways. The Android ecosystem is now vast, embracing not just smartphones – both top-end and el cheapo versions – but also tablets, laptops, and an-ever widening range of consumer and domestic devices.

At the centre of that web sits Google. Even if companies like Amazon start forking Android, they are still reinforcing the overall ecosystem – for example, in terms of growing the pool of software engineers that can program for it. That puts Google in a very strong position when it comes to protecting its core business, that of Web advertising, because it implicitly controls such a large number of how devices access the Web.

In many respects, this situation is very similar to that of open standards. As I noted the other day, it is in Apple's interest to offer RF licences to its new nano-SIM technology just to get it accepted as the industry standard. Similarly, Google and Red Hat don't try to make money directly from Android or GNU/Linux, because that would limit the uptake of those products. Instead, they give them away freely in order to establish them as the de facto standards in their respective domains. Once that happens, money can be made in other ways from products and services that build on that dominant position. Red Hat's recent milestone is proof of that, and there will doubtless be similar demonstrations of Android's power to make serious money for Google in due course.

Follow me @glynmoody on Twitter or, and on Google+