MySQL's Planned IPO: OSS Buzz Isn't Enough

The open source software [OSS] blogosphere has taken its eye off attacking Microsoft (MSFT) for a few days to focus on a recent BusinessWeek article that predicted a 2007 initial public offering for MySQL. The significance according to OSS partisans is the way it indicates that OSS is taking over information technology to the exclusion of all other software development/business models and user choices. The significance to investors is that Red Hat (RHT), as the only significant pure play OSS provider that is already public, will act as a guidepost to establish MySQL’s valuations. OSS will probably suffer on both counts.

I leave discussions of the community aspect of this to my OSS product/market research on ebizq.net. But as for the investment research implications, don’t rush into secondary and tertiary rounds of venture funding just yet for every OSS "plan" that crosses your desk. A good business case is still more important than a buzzword such as OSS. And a good business case rests on the functionality that will be delivered, not the buzzword. Even the buzz won’t help assuming investors keep 1998-1999 in mind.

In my opinion, the investment opportunities—and OSS’s future—lies in the application functionality side of the market, not the infrastructure side. I think the OSS operating system/middleware/database train has left the station The BusinessWeek article (as picked up in my link above by Linux Insider) reports that MySQL is only getting one in a thousand downloads to turn into paying business.

The Business Week article does not provide its source for its data but it appears in a sentence preceded by a quote from MySQL’s chairman and followed by a quote from MySQL’s founder. That’s pretty actionable to me. Panelists at a UBS technology conference I attended in 2006 said the same thing for OSS in general. I saw a similarly low ratio of conversions with JBoss when I was doing middleware research for IDC in 2005 and before, a statistic borne out in JBoss revenue numbers revealed in subsequent Red Hat SEC filings after Red Hat acquired JBoss.

There is nothing inherently different about the OSS model in terms of going public. But middleware/database opportunities will suffer because the leading public information technology [IT] companies have already taken over the OSS movement at the low end of the stack. There are very few infrastructure market sectors where an OSS startup can build a revenue stream without running right into IBM (IBM), HP (HPQ), Oracle (ORCL), and so forth. This group, even arguably Microsoft (but that’s another post), have whole-heartedly embraced OSS and embedded Linux, Apache, and so forth into their products and services.

Red Hat’s timing was excellent. This was true from a product/market perspective. What Red Hat wanted to do was essentially get users to migrate from a moribund Unix market made up of that IT Top 12 group of suppliers, not counting Microsoft, who were backing off their support of Unix. And there it was true from an investment perspective. But JBoss wasn’t able to achieve the same result either with the product or investors (other than Red Hat management).

WebSphere and BEA (BEAS) WebLogic were too well established to let JBoss do to the middleware market what Red Hat had done to the operating system market. And in both cases it was the functionality—operating software in one case, application serving in another—that users cared about, not the OSS development or business model.

Using the same criteria, MySQL is probably five years too late. IBM, Microsoft and Oracle will block any kind of growth from the mainstream. They will even do it with an OSS message.

If there will ever be a separate OSS investment story, it will be in the applications space, but even then that will be more a software as a service (SaaS) play than an OSS play.