The Unsurprising Surprise of Open Source M&A

Share via Twitter Share via Facebook Share via Linkedin Share via Reddit

There must be something going around, something in the water. That was how one commenter put it to me this week. How else to explain the rash of open source acquisitions? From MySQL to Covalent to Trolltech, the appetite for open source vendors appears to be insatiable – even when the acquiring firm is itself an open source play, as is the case in the Covalent deal. Speaking for many, Matt Asay’s headline discussing the Trolltech acquisition included the question: “Will any open-source companies be left to change the world?”

Tongue firmly in cheek, presumably, as Matt knows the realities of this market as well as anyone and better than most. What has interested me over the past few weeks has been the surprise – even shock, in some quarters – over the popularity of open source vendors. Not the valuations, mind you, which have provoked the expected differences of opinion over. But the mere popularity.

Which has inspired a few comments in the discussions I’ve had on the subject with vendors, individuals, and media alike.

The Trend is Not New

Though it has been treated as a recent trend in some quarters, the reality is that open source acquisition has been in vogue for a while; it is far from a recent phenomenon. Perhaps you’ll recall that Oracle was rumored to be after JBoss, Sleepycat, and Zend – at the same time. True, only one of those rumors was accurate, but Red Hat accounted for another one of the players and today there’s but one left standing.

Nor have commercial entities been reluctant to invest in open source firms at financial scale (MySQL/Xensource) or in areas that compete with existing products (LogicBlaze, Gluecode).

Open source as a business model may yet be unproven relative to its closed source competitors, but it clearly does little to inhibit the potential for acquisition. Quite the contrary, in many cases.

Why might that be?

Trading Revenue for Penetration

In his piece “Hiring is Obsolete,” Paul Graham makes the case that the practice of established firms acquiring startups is simply an effective form of outsourcing risk.

Big companies also lose because they usually only build one of each thing. When you only have one Web browser, you can’t do anything really risky with it. If ten different startups design ten different Web browsers and you take the best, you’ll probably get something better.

An assertion that few would dispute, I think, at least on a general basis. But which has nothing intrinsically to do with open source, of course.

But if you’ll allow the above, it’s actually easy to see where open source fits in. Open source, after all, is the ultimate embodiment of risk – according to some – in its embrace of a model that eschews traditional licensing and customer conversion mechanisms. Open source, in other words, trades up front licensing for the potential of ubiquity.

Think of it this way: in this day and age, there are essentially two paths towards ubiquity available to startups – open source or Software as a Service. Exceptions like Skype notwithstanding, those are the options if you’re pursuing a volume opportunity. For some, the prospect of scaling a SaaS application is terriying, in resource requirements, knowledge required or both. Which makes open source an attractive risk.

If you’re a well heeled closed source shop, that risk can appear unacceptable. But outsourcing that risk to smaller firms, then acquiring the one that achieves ubiquity? Unsurprisingly, that is slightly more popular.

But are closed or mixed source firms acceptable stewards for open source codebases?

Growing Open Source Businesses

Opinions on the subject will differ, of course. But the fact is that larger firms, for all of their conservativism, typically can bring resources to bear that startups couldn’t dream of.

In the same Graham piece mentioned above, he claims that “big companies are good at extracting the value from existing products, but bad at creating new ones.” Again, as a comprehensively authoritative statement it leaves something to be desired, but I think even the skeptics would acknowledge the kernel of truth it contains. Consider Microsoft Office, one of the most successful software products in history, which was itself built in part through acquisitions.

To date, there are few examples of closed source firms growing open source firms massively beyond their roots. But then there are no billion dollar open source pure plays, period.

Certainly larger entities have resources and institutional knowledge and experience that smaller open source vendors lack. The former was, in fact, one of the justifications given by Marten for MySQL’s chosen exit path.

What it Means

That we’re going to see more open source vendors acquired. Many more, and some in the not too distant future.


  1. Open source hybrid companies, those take try to leverage a community and exploit it for commercial gain have not succeeded in general. With the exception, that is, of RedHat, MYSQL, and a few others. The better approach is a new breed of commercial vendors that embrace open source (and provide open source APIs and SDKs to the open source community), but don’t open their code. Salesforce.com (www.Salesforce.com) has been successful doing this with their app exchange portal. Another early stage company that does this well is http://www.PacketTrap.com. Let the open source community build extensions to proprietary commercial solutions.

  2. I should really pen something down about the transition of the proprietary vendors to the record industry business model: pick up things people like, make ’em popular, profit.

    Punk’s not dead! Only commercialized. 🙂

  3. There is no problem for the population, nor for open-source.

    Software firms try to take over the best what was developed as open, to diminuish development costs.

    But the population has the option, to continue to use the open version at the time of the privatization, and everybody can fork it. So it’s a much bigger risk for them to spent plenty money for takeover a project, but later nobody buy it because people continue a slightly older version or a fork. This liberty is warranting the GPL

Leave a Reply

Your email address will not be published.