Marketing Memos

I wonder what antitrust agency lawyers will make of Oracle’s acquisition of Sun. Not that it matters much as Larry Ellison has slain that dragon before. Rumor has it the heads of three lawyers are mounted on the wall of Ellison’s office (knew I liked Larry for a reason).

(next graphic borrowed from JoinVision.com – click to enlarge)

The database market is tiny when measured by the number of serious competitors. Oracle is the king of the enterprise database server beasts while IBM’s DB2 or Microsoft’s SQL Server holds a tenuous second places (depending on whose measure you use, the common analyst measures being revenues, installations, recent developer involvement, moon phases or the pattern of chicken bones tossed into a fire pit). Microsoft’s SQL Server has its adherents, but that is largely out of religion for Windows not over-arching trends. Also-rans like Ingress, Sybase and others have receded into obscurity. For enterprise database technology, this is a seriously contracted market and one that would cause any government lawyer to skip their triple martini lunches.

Now Larry is buying the bottom of the market.

In Sun he also gets MySQL, the Open Source competitor for which Sun overpaid. MySQL not only has absconded with nearly the entire low end of the market (a sector where Oracle never made money nor marketed their wares with any real intent), but it is eating most of the middle ground as well. When I talk to CTOs and ask them about non-mission critical applications, they contentedly use MySQL and forgo the expense of establishing another Oracle instance.

That has to annoy Ellison no end (knew I liked MySQL for a reason).

In buying Sun, Oracle is significantly consolidating the entire database market, and demonstrably the enterprise database market. Granted MySQL’s Open Source nature means Oracle doesn’t really control it at all. Larry can’t kill it off on a whim. Yet Oracle can influence its growth, support and overall market viability. Oracle can affect MySQL’s future.

More to the point, the government simply does not like one corporation owning a wide or deep part of any market, and now Oracle has both. It has width – covering one half of the enterprise database market (if you consider Microsoft’s SQL Server an enterprise solution) and it has depth now owning – for practical purposes – the bottom and middle via MySQL. Using rough estimates of penetration, we can visualize what most of us instinctively know about this merger – that the number of alternative providers in the market is suddenly smaller.

Ignoring for a moment MySQL and the steam venting from Steve Ballmer’s ears this morning, the other parts of Sun are equally interesting, both from a market and an antitrust standpoint. Like MySQL, Java is Open Source. But Oracle has a vested interest in Java given the architecture of Fusion and derivatives. With Oracle dominating the enterprise application market and competitors depending heavily on Java for their applications, there is an inherent competitive advantage for Oracle to control the fate of Java. Again, being Open Source, Larry can instigate only limited intrigue, but there is enough leverage to make trouble for his foes.

Despite the potential for using SPARC’s massive multi-threading chip to create a monster database engine, I cannot seriously believe Oracle wants to get into the server hardware business. X86 architectures are the new norm and if today’s VMWare introduction of vSphere 4 is an indication, the cloud is the new mainframe, making SPARC’s minor advantage irrelevant.

Finally, there is StorageTek, another outfit Sun bled too much cash for (was Jonathan Schwartz secretly on a buying spree for Larry?). All data needs backup, but storage is a highly competitive arena and Oracle prefers fat software margins over skinny and shrinking hardware margins. Spinning off StorageTek is highly likely though who the buyer might be is unclear. Perhaps Oracle will do with StorageTek what EMC did with VMWare and create a beast with two backs.

Regardless, we can expect the Fed’s to get deeply involved and for Ellison to crush them. Owning control over MySQL and Java are juicy enough for him to play hardball (as if he knows any other way). All this time I worried about the technology business being commoditized. I should have been worried about Oracle swallowing it whole.

Bookmark to:

Silicon Strategies client DeviceAnywhere has the good sense not only to employ us, but to take the pulse of their own industry on a regular basis. Knowing that trends change over time helps companies know how best to serve their customers – to anticipate their needs

DeviceAnywhere surveys examine what technologies mobile applications developers design for, how DeviceAnywhere services are used and most interestingly what kinds of mobile applications are being written. This last bit says everything about how the mobile application market has shifted and what consumers really want.

They want the Internet in their pocket.

The Internet is a success because it is Darwinian in nature. Every mutant content provider self-formed out of the digital muck, rapidly mated in orgiastic enthusiasm, and new species of content and application are perpetually being delivered. Go forth and iterate.

The reason a Darwinian Internet creates is popular is that every possible combinations of content and amusement is tried, with the failures dying in obscurity and the winners consuming ever more bandwidth. Even seemingly trivial Internet applications – and by this I mean Twitter – can evolve into ragingly popular services.

Humans are now Internet addicted. Waiting to get home or to the office to hear music, read news or toke on Tweets creates withdraw symptoms. DeviceAnywhere’s survey shows that the majority of developers are focused on mobile Internet – more so than games, downloadable content or mobile banking. Mobile smut was not specifically mentioned, but I’m sure access to such content while riding the bus to work is a significant part of the new mobile content movement.

Price reductions in mobile data bandwidth paired with increasing demand for Internet content and the simplicity of maintenance/delivery of web content/applications is conspiring to fundamentally change the mobile application market. Mobile internet is becoming as ubiquitous as mobile handsets. Over time developers have and will continue to drift toward creating mobile experiences based on Internet technologies, mainly web.

For handset makers, the breeds an imperative: creating stable, strongly standards-compliant micro PCs. The closer a handset can come to mimicking everything about the desktop Internet experience, the more content that handset can deliver and the more mobile application developers will drift toward that solution set.

Commoditization of the means of delivery.

This brings to the fore an interesting question: What happens to major parts of the mobile industry when everyone and his dog focus on mobile Internet application? Good things in general happen from the customer perspective, though certain vendors will be left eating air. First, the differentiation of the handset operating system becomes minimized. People will not pick handsets based on the relative irrelevancy of built-in tools but the ability to deliver the profusion of online applications. Sure, certain combinations have staying power. For example I’ll always begrudgingly use a Windows handset because Outlook runs my life and strict compatibility is essential. But for the super majority of buyers, Internet delivery (web, Java, 2.0, etc.) will be the growing decisive factor. After that, the ease and joy of using the device will drive handset sales, which is where Apple currently hold and advantage.

It also means a latent boon for web application tool vendors that can find mobile Internet leverage points for their offerings. Toolsets that allow for the creation of mobile web/Java apps will gain favor with developers who follow the trend. A mobile application needs to gather handset info, contribute to analytics and leverage handset features (like GPS) to make it more useful than the next mobile development toolset.

Does this mean native applications like BREW, Windows/exe and Android/Linux installables are going away? Slowly, yes. There are applications people need and want to run when not connected to the net. But given the integration of Java into the Internet experience and the ability to install over the air a Java application, handset support of a well architected native Java virtual engine will be more important. The divide becomes what applications are installed at the factory and which are so important the relative fragility of wireless data connections mandate manual software installation (my handset’s GPS software being an example). Everything else goes web and other Internet standards staples.

The more things change the more they commoditize.

Bookmark to: