How Facebook and Yahoo are spawning billion-dollar startups
There's an increasing business case to be made for giving software away for free.
Take Facebook, for example.
As Facebook got bigger and bigger, it started running into some serious growing pains — it was generating more data than its data centers could handle. This was a problem, given that people expect Facebook to always be available and snappy.
Normally, this is where a company seeks help from a vendor or a consultant. The problem was that there had never been a web company with Facebook's problems before, so Facebook was forced to come up with its own solutions.
A piece of software called Cassandra was originally designed in 2008 as the engine that underpinned storing and searching through all the messages in your Facebook inbox, which is a tall order considering how many millions of messages the company has to deal with every day.
"The aim was to design a solution that not only solved the Inbox Search problem but also provided a system as a storage infrastructure for many problems of the same nature. Hence was born Cassandra," wrote Cassandra inventor Avinash Lakshman in the Facebook post announcing the project.
Cassandra worked great. But Facebook's business isn't in making data software. It's a social network and advertising company, and its resources, while vast, are limited. Internally, Facebook could only take Cassandra so far.
But by releasing Cassandra to open source, via the Apache Software Foundation (a non-profit that manages open source software projects), developers all over the world could see what Facebook cooked up and put it to work in their own companies.
Apple and Wikimedia use Cassandra in their own web software, these days, as do plenty of smaller companies.
The developers who use Cassandra make improvements to it, applying what they've learned to make it work better for their own specific usages. And very often, those improvements get submitted all the way back up to the Cassandra project.
In other words, Facebook reaps the benefits of a better Cassandra, without having to invest the money in making it better internally. And since search isn't a core part of Facebook's business, either, there's little risk that they're giving away a critical company secret.
"Open source is not a threat to anyone's business plan anymore, it's now a way to go even further," said GitHub CEO Chris Wanstrath on stage at the Bloomberg Technology Conference this week.
Wanstrath also noted that when Microsoft open sourced its .NET programming framework late last year, it took about a day for an Apple Mac version to appear — not something Microsoft would have prioritized, but it served their goal of getting .NET to more places just fine.
Plenty of other technology companies, including Google, LinkedIn, and now Apple with the Swift programming language, regularly release open source code to the world. Even Apache Hadoop, the buzzed-about big data software framework, is an open source project that traces its roots back to Yahoo.
The proliferation of open source projects is making it easier for startups to get off the ground.
Cloudera, a big data startup that's raised $1 billion at a rumored valuation of around $5 billion, built its software on the open source Apache Hadoop. DataStax, a startup that built a commercial product on top of Cassandra, took a $106 funding round late last year as it prepares for IPO.
"Open source is driving this massive acceleration of the number of companies that exist today," said LinkedIn SVP of Engineering Kevin Scott on stage at the Bloomberg Technology Conference this week.
A lot of the people who originated open source projects often leave to turn their idea into a sustainable business. Mesosphere, a high-profile data center management startup helmed by the inventor of Apache Mesos,raised a $36 million Series B round last December.
In 2014, venture capitalists made 37 investments in open source-based startups, according to Redpoint VC Tomasz Tunguz in a recent blog, and he sees no signs of slowing.
The fact that these startups are based on open source means that their developer customers are already familiar with the concept and technology, and just want someone else to do it for them and do it well, Tunguz says.
"Companies of all kinds are starting to use software as competitive strategic advantage, and the pace of innovation in infrastructure and software continues to increase," writes Tunguz.