Sun plunked down $1 billion this week to purchase MySQL, illustrating that open source companies can achieve impressive capital valuations, even if questions remain about their potential for achieving significant (read: as big as proprietary software companies) revenues. On the heels of the Yahoo! purchase of Zimbra ($350 million) and Citrix’s purchase of XenSource ($500 million), Sun’s buy shows that the numbers keep going up to get into the open source game.
As I mentioned, people ask how well open source companies can do financially. My own view is that open source companies can be extremely profitable, but, compared to the most successful proprietary companies, their growth rates will always appear stunted. License fees can grow dramatically (that’s why their charts are called hockey sticks!), whereas open source companies are subscription-based, which tend to grow more slowly, but have long-term strength. This is because a subscription, once purchased, tends to be a perennial — renewed year after year.
So one question is whether Sun paid the right amount for MySQL. Rumor is that MySQL did around $50 million last year (maybe that’s a bit low), and is growing pretty well, so might do around $100 million next year (maybe that’s a bit high). So Sun paid around 10X revenues, which is pretty steep, even for an annuity-like revenue stream. Of course, MySQL is growing, so, depending upon the growth rate, the multiple might rapidly shrink.
Sun made kind of a big deal about how having a company the size of Sun backing MySQL would make big customers more comfortable using MySQL. I have to say I’m not that convinced about that argument. People usually buy open source subscriptions only after they’re committed to the product, and the commitment usually occurs bottom-up; in other words, a technical person uses open source for some task or project, and a year or two later the company realizes they really, really depend on the open source product, so they ought to buy a subscription to get industrial-strength support. The universe of technical people who wouldn’t have used MySQL in the past, but would now because Sun owns it, seems pretty small to me. This isn’t to criticise Sun, but to disagree with their analysis of the drivers for adoption of open source.
On the other hand, I think Sun has made a smart strategic purchase. Sun does a ton of open source work, but doesn’t get as much credit as they deserve. In part this is because they’re not part of the Linux crowd, but it’s more due to the fact that, despite their estimable work, their own open source products tend to be also-rans. It’s hard to get credit if you’re not a winner, even if you work hard.
In MySQL, though, Sun has bought a clear category leader, which will let them put a much more impressive logo on their products slide. Sun’s bought (open source) street cred.
Moreover, Sun is the major technology company most committed to using an open source strategy: drive low-cost adoption and then harvest revenues post-adoption. It’s explicitly said that getting its open source products into people’s hands gives them a chance to have a conversation and begin selling other (read: Sparc and high-end X86 boxes that resource-hungry databases run on) products. So, even if the financial analysis for the acquisition doesn’t pencil out, investing $1 billion to get a sure-fire product to use as an adoption driver can be thought of as a very sophisticated loss-leader.
On balance, I think this is a very good decision on Sun’s part, especially given that many of the other big download open source companies are already spoken for.
One big question remains — why didn’t Red Hat buy MySQL? It seems like a database would be a natural fit to round out the OS/App Server offering it already has. In some ways, MySQL would be an even better fit for Red Hat. One can speculate that Red Hat is a conservative company, less willing to pay what it might see as an inflated price for MySQL. And, of course, it has less need of a loss-leader to enable other selling activities.