The IPO plan of Swedish software company MySQL is eagerly awaited by fund managers, who want more benchmarks to gauge open source's value
For all the success of open-source software—developers the world over flock to the code available freely over the Internet—its purveyors able to thrive as public companies are few. Linux operating system seller Red Hat (RHAT) has generated billions in value for investors, but its shares have slipped 3% in the past year amid new competition. Novell (NOVL), which supports a version of Linux, has been criticized for striking a cooperation deal with Microsoft (MSFT) seen by many as a threat to the spread of Linux.
That small community of open-source stocks may soon be widening. MySQL, a fast-growing maker of database software used by some of the Internet's most recognized brands, is preparing to file for an initial public offering, perhaps as soon as late 2007. The offering could value the company at between $600 million and $1 billion, according to sources, and inject some pep into a tech IPO market that's seen only a handful of successful offerings in the past year. Credit Suisse (CS) is a top contender to lead the underwriting of the transaction, BusinessWeek has learned.
An S-1 filing by the Swedish software company, which grew more than 50% in 2006, to about $50 million in sales, and broke even for the first time, also could give investors a new yardstick to measure the value of open-source software, which lets users modify its code to suit their needs. "Red Hat really is a bit of a lone wolf out there in terms of public open-source companies," says Jim Zemlin, executive director of the Linux Foundation, a trade group. "Fund managers are clamoring for other benchmarks to measure open-source software companies."
Outside investment and tech circles, though, most people haven't heard of MySQL, which closed an $18.5 million round of funding in 2006 to raise its total venture backing to $39 million from investors, including Benchmark Capital (see BusinessWeek.com, 2/13/06, "MySQL Answers Back With $18.5 Million"). Yet MySQL's fast, inexpensive software is used by such Internet heavyweights as Google (GOOG) and Yahoo! (YHOO), and it's making inroads into more traditional companies. Going public could give MySQL more credibility with brick-and-mortar shops, and furnish it with currency for acquisitions. "The open-source model, as Red Hat has proven, can be extremely profitable," says Kevin Harvey, a general partner at Benchmark Capital, and MySQL's chairman. "It's not a story of profits at first; it's a story of profits you'll generate as you grow."
And MySQL would like to see considerably more growth, even before selling stock to the public. It would do that in part by upping the percentage of paying customers. Of the roughly 11 million copies of MySQL in use, the company only gets paid for about 1 in 1,000, underlining the risks of the business model governing much of open-source software: Give it away for free over the Internet and then charge large commercial users for technical support. "There are many users who will just never, never pay," says MySQL Chief Executive Marten Mickos. "It's not like we can just go in there wholesale and change that mindset." Making matters tougher, Microsoft's affordably priced SQL Server database competes with MySQL. And database software vendor Oracle (ORCL) has made inroads into the small and midsize business market, helping it gain share.
MySQL, which a few years ago rejected a takeover offer from Oracle, is undeterred by the challenges. "We're working toward an IPO," says Mickos, a sturdy, sandy-haired Swedish Finn who migrated to Silicon Valley in 2003, two years after he became CEO following a series of executive posts in Europe. Mickos, who conducts business fluently in English, German, Swedish, and Finnish, has known MySQL's founders, including Monty Widenius, since 1981, when they studied graduate physics together in Helsinki.
MySQL has already started courting investors. It held a 2007 pre-IPO road show in New York and Boston to talk about the company's brand recognition, sales, and the appetite in the public market for an open-source software company, Mickos says. "We've gotten good feedback from some of the biggest public investors," he says.
There's reason for the warm reception. Growth at MySQL has taken off the last few years as some of the Web's hottest companies have adopted its technology. Google's ad-serving software runs on MySQL's database, as does its YouTube video site. "MySQL is a terrific database," says Chris DiBona, Google's open-source programs manager. Yahoo's Flickr photo-sharing site runs on MySQL, and the company uses the software for its finance and games sites.
Other Web companies including Wikipedia, Facebook, Craigslist, and Linden Lab's Second Life are adherents as well. "The technology was crucial to us being able to deliver so much so quickly," says Scott Dietzen, president and chief technology officer at Zimbra, whose open-source e-mail software ships with MySQL inside.
Compared with database software from Oracle, Microsoft, and IBM (IBM), MySQL's product dispenses with many features aimed at running financial software and other business applications in favor of a stripped-down approach that serves up Web pages at blazing speeds. "We grew up with the Web companies," says Mickos. "People say, 'MySQL, will you ever grow up to be an IBM?' And we say, 'No, that's the old world.'"
The company has carved a niche among Web companies that use its technology in conjunction with Linux and other open-source software to run their sites on the cheap. "That's the aspiration of these companies—they want to grow big, but they don't want to spend a lot of money to do it," says Zack Urlocker, MySQL's executive vice-president of products. MySQL aims to sell its software for 90% less than its competitors, he says.
A public MySQL could be a good buy as it fills the underserved market for an affordable database aimed at companies exploiting new, more interactive aspects of the Web. And it's gaining traction at other companies, too. MySQL helps power Nokia's (NOK) cellular network and Gap's (GPS) checkout systems. NBC and The New York Times Co. (NYT) are also customers. Cisco Systems (CSCO), Symantec (SYMC), and other tech vendors distribute MySQL with their products.
But can MySQL keep up the growth without adding hefty sales and marketing costs—and getting squeezed by competitors? The company employs just 30 field sales staff out of a head count of 360 and strives to close deals more quickly than rivals. Most employees work from home. "Managing the cost of sales and marketing in an open-source company is the key to profitability," says Mickos, sitting in a small, spartan office adjacent to a sea of cubes in the company's Silicon Valley digs. "We're not just innovating in software, we're innovating in sales."
Rivals aren't taking the threat lying down. In 2005, Oracle bought a Finnish software company called Innobase, whose technology is used by MySQL. MySQL is building its own version of the software in a project code-named "Falcon," but for now must pay licensing fees to its bigger rival. And IBM and Sun Microsystems (SUNW) are backing an open-source database called Derby, which competes with MySQL.
How investors will value MySQL is another open question. Some reckon Red Hat, whose $4.55 billion market value is about 11 times its fiscal 2007 revenue, is a good starting point. Like MySQL, Red Hat also has millions of users and a low-cost sales model, says MySQL Chairman Harvey. Jereme Le Blanc, vice-president at investment banker Boston Corporate Finance, says MySQL's 50%-plus growth for its size could catapult its market value well above $500 million—if the company can convince users to keep paying for subscriptions. "That's pretty robust growth for a company that's already at $50 million," he says. MySQL's ability to keep selling service contracts for software that's also available free will affect its valuation, though. "A lot of people think the open-source market for database software is very lucrative," Le Blanc says. "It's too early at this point to gauge whether the model works."