FOSSLC is a non-profit organization that specializes in technology and know-how to record conferences with excellent quality. Click on the icons below to view great videos from communities we are actively involved with:


Stagnant MySQL Billings at Sun

12 Feb

Sun's MySQL business unit has been in the headlines recently, both for the wrong reasons (loss of important personnel) and for what appear to be the right reasons (big revenue gains). Unfortunately the headlines about massive billing increases, such as Matt Asay's blog post, suggesting billings of $81 million in a single quarter for MySQL, are wildly, howlingly wrong.

There is no doubt MySQL is fantastically popular, and there are excellent reasons why it would be. But it is making nothing like as much money for Sun as some say.

This is important to the wider open source sector because revenue stands as a proxy for a lot of the measures of health and stability of a software vendor. Anyone choosing an open source SQL database system today could well be influenced to select MySQL over a potentially more suitable product on the entirely mistaken basis that MySQL is more profitable than any other, or that its billings are soaring.

To be fair, this misapprehension probably wasn't created by Sun deliberately; the correct information is readily available. If we
refer to page 6 of Sun Microsystems Q209 Quarterly Results Release, we see the true data, though it is not presented at all clearly.

As you read this document, note that Sun's fiscal year starts in July, so they're already reporting their Q209. Also, recall that Sun
announced the completion of the acquisition of MySQL AB on February 26 2008 (Sun's Q308), somewhat less than a year ago.

Page 6 is the Billings Disclosure by Category. Under Software Billings there is a line item called MySQL/Infrastructure. At face value we are invited to assume this line item is all about MySQL. Whatever "infrastructure" is, it's apparently too unimportant to merit any explanation at the bottom of the page, unlike some of the other line items. However, closer inspection makes it clear that infrastructure and not MySQL is the real cash-cow, by a factor of more than three!

To see why, notice the sum of billings in this line for Sun's Q307 to Q208 (that is, the last four quarters in which MySQL contributed no billings) is $201M. The sum of billings for Q308 to Q209 is $264M. So assuming the billings for "infrastrucure" didn't take a massive tumble, the net contribution to billings by MySQL is roughly $63M during the last four quarters--maybe less, or about 23% of the value of this line item.

So over an entire year, MySQL made only 75% of what some have claimed it did in a single quarter. What is more, billings growth seems to have been pretty much flat. There is no really authoritative information about MySQL AB's billings prior to the acquisition, but some estimates are between $50M and $60M a year. Taking the most generous number, going from $60M/year to $63M/year is really no growth at all.

So when you see headlines like "MySQL's sales [are] catching up with Red Hat's", remember Red Hat's last quarter alone was twice MySQL's entire year, and MySQL billings growth is apparently stagnant.


Number fudging

I agree with the premise of your post, and would also like to point out that Sun posted these numbers in their presentation slides and not in their 10-K or 10-Q forms which means that they have not been audited. The billings data that we have cannot be relied upon and the mysterious "infrastructure" item does not have to be explained at all. Personally I wouldn't bet my life on the numbers but what I would say is that Sun has definitely fudged them to justify the nearly $1b price tag of MySQL.