Why Oracle must kiss goodbye to its database past
Moneybags Larry hopes cloud will shower him in cash
Open ... and Shut A few years back Apple opted to drop the "Computer" from its corporate name, and instead became "Apple Inc." Last week, Oracle made a similar move, quietly stopping its decades old practice of reporting database revenues. Oracle chief executive Larry Ellison was quick to point out on Oracle's earnings call that for its database sales force "it's business as usual." But is it? By opting to stop breaking out its database numbers, Oracle is telling the market that it's no longer a database company.
That's the spin Cloudera CEO Mike Olson puts on the change, arguing in a Facebook post (which he allowed me to share):
Oracle wants investors to pay attention to something else now, and is changing what it talks about to make that happen….[I]f you don't report database revenue, you're no longer a database company. And I bet Oracle leadership would agree with that statement.
Not that Olson thinks the move is insignificant. In announcing the shift, he wrote: "Oracle Corporation no longer breaks out database revenues in its earnings reports. Next up: Frogs. Locusts." Steven J. Vaughan-Nichols, a long-time Oracle observer, for his part, "find[s] it troubling," arguing that "at the end of the day, [databases are] what defines them" and that analogous changes would be "IBM not reporting on mainframes, MSFT on Windows, etc."
But maybe, just maybe, that's what Oracle wants, to underline Olson's point.
Despite initially pooh-poohing the cloud, Oracle has doubled-down on cloud computing. It has also acquired a mass of companies selling everything from applications to hardware. In the midst of that "cloud" of software services and on-premises software that Oracle now peddles, the database has become just component among many in Oracle's arsenal, as Sean Kerner points out:
@mjasay don't forget it's challenging for Oracle now to define standalone db revs since they have engineered system & cloud integrated stuff— Sean Kerner (@TechJournalist) September 24, 2012
And yet, Oracle president Mark Hurd insists that it continues to have a dedicated database team that does nothing but sell its database offering. Surely that team is being compensated for those sales, so figuring out total database sales must remain reasonably straightforward.
If Oracle wanted to do so. But, apparently, it doesn't.
The charitable, and perhaps correct, view is Olson's. But others are less sanguine about Oracle's rational and, indeed, about its business, generally. Cowens & Co analyst Peter Goldmacher had this to say:
Management claims of investing in Sales and Marketing ring hollow considering Sales and Marketing spend is down year-over-year in absolute terms. Revenues are in structural decline as every segment of Oracle's business is under attack from a new generation of technology providers that offer equivalent or better technology at a lower cost. But as long as management can continue to cut costs faster than revenues slow, earnings will grow, until they won't. No part of the bull case for Oracle... addresses growth, competitiveness or market share gains.
Painful, but possibly true. Oracle, after all, is seeing its applications business under siege from more nimble competitors like Workday, while its core database business is under threat from a long-term shift to non-relational database and Big Data competitors like Cloudera (currently an Oracle partner) and 10gen (MongoDB).
Given the headwinds facing Oracle, coupled with the difficulty of changing direction for a company that employs 120,000 people, it's perhaps not surprising that Oracle is trying to shield scrutiny from the inevitable pain it's going to feel in individual product lines while it moves to becoming a true cloud computing company. El Reg's own Timothy Prickett Morgan emphasizes that Oracle's change in product revenue reporting means "we have no idea how these business units are doing individually anymore."
Which is almost certainly the point. As Ellison noted on the Oracle earnings call:
A lot of companies have a difficult time adapting to the next generation of technology. And those with the most are also the ones with the most to lose. You have to be willing to change the way you were doing business in the past and adapt to the new opportunities and exploit the new opportunities.
Reading between the lines, Ellison seems to be acknowledging just how difficult this transition to the cloud has been (Oracle has been working on its Fusion technology for seven years), and will continue to be. Buying some breathing room by focusing on the big picture (cloud) while minimizing short-term pain in individual products (such as databases or applications) is a smart move.
And a painful one. As ZDNet's Larry Dignan highlights, Oracle has spent years dabbling in hardware, only to have the hardware business consistently underperform and distract the company from getting fully into the cloud. Apparently Oracle has learned its lesson. What remains unclear, however, is whether shifting to the cloud will prove as successful as decades of database dominance have been. In the cloud Oracle is a neophyte, albeit one with gargantuan resources behind it.
No matter how Oracle financially engineers the painful shift to the cloud, at some point it's going to have to show financial results. So far, they don't look pretty. With all the competition it's facing, they may never look pretty. But full credit to Oracle for engineering a transition to the cloud, even if it had to abandon its past glory as a database giant to do so. ®
Matt Asay is senior vice president of business development at Nodeable, offering systems management for managing and analysing cloud-based data. He was formerly SVP of biz dev at HTML5 start-up Strobe and chief operating officer of Ubuntu commercial operation Canonical. With more than a decade spent in open source, Asay served as Alfresco's general manager for the Americas and vice president of business development, and he helped put Novell on its open source track. Asay is an emeritus board member of the Open Source Initiative (OSI). His column, Open...and Shut, appears three times a week on The Register.
Re: Relational DBs
Most NoSQL databases have data models strongly tied to the applications using them. Relational DB have not. When data are much more importand than the applications using them - as most valuable data are - you want them to be accessible by multiple applications in very many different ways. And Oracle will be still here when most NoSQL databases will be forgotten. As usual Asay has no clue about real computer needs, and keeps on living in his own small world.
Re: And now for something completely different
Agree that Oracle is trying to move to "engineered systems", but their customers, generally, do not want engineered systems. People forget that the reason Oracle became Oracle (and Cisco became Cisco for that matter) is because they argued, persuasively, that it was a better approach to use best of breed vendors for various components than one large vendor, called IBM, for an integrated system. They argued that the flexibility and cost benefits outweighed the integration benefits of the mainframe.... Now that these providers have taken over their niches, Cisco in networking and Oracle in DB, they are arguing exactly the opposite of their original argument. Turns out, as they have recently discovered, that having one provider do everything requires less integration... kind of like IBM has been saying for 30 years.
Pic from twitter says it all
Matt Asay knows nothing about a subject and announces to his fellow twats that he's going to write about it anyway and if they have any opinions he'd love to roll them into his bog-post. Someone send him a wank sock.
El Reg is turning into the regurgitation of the idiots: Matt Asay, Florian Müller, Tim Worstall, et al. Pity Mr Orlowski isn't pointing out it's because real journalists want to be paid real money which they can't do because we all have ad-blockers. Think it's going to SPB only for me in future with the odd dip into anything not related to climate that Mr Page may have written.
Relational DBs are alive, and will probably outlast newcomers.
YA, maybe these nosql DBs are fasta, etc, but at the end, relational, sql databases are what stays. That doesn't mean that Oracle will perdure.. there are many other nice databases.. and all less pricey .
Oracle reporting just database earnings....
...is like the devil reporting just horrific floggings. Sure it's one service they provide, but not necessarily a large part of the total package of misery they inflict anymore.