Oracle's potential math problem: Worrisome licensing trends ahead

Oracle's potential math problem: Worrisome licensing trends ahead

Summary: Oracle's technology licensing revenue adjusting for currency and acquisitions may be showing a few caution flags, but the software giant can weather the worries because of its maintenance and support cash cow.

SHARE:

Oracle's technology licensing revenue adjusting for currency and acquisitions may be showing a few caution flags, but the software giant can weather the worries because of its maintenance and support cash cow.

The big question: How long can the maintenance cash cow last?

Earlier this week, Cowen & Co. analyst Peter Goldmacher published an interesting analysis of Oracle's organic licensing growth. Simply put, Goldmacher's math shows that Oracle's database and middleware business as well as applications are under pressure.

In a research report, Goldmacher said:

Our math indicates that adjusting for currency and M&A, Oracle's Technology license sales grew -3.5% in FY09 and -4.5% in FY10, and the Apps license sales grew -9.5% in FY09 and +5.3% in FY10.

Why the slowdown? Goldmacher has highlights a few interesting cross currents. To wit:

  • There are low-cost database alternatives to Oracle's core database. Goldmacher said that until just a few years ago low-cost databases didn't exist. In other words, enterprise databases were needed to meet every workload. Today, there are open source databases and platforms like Amazon Web Services. Add it up and IT buyers aren't going to go with Oracle for non-mission critical workloads.
  • Meanwhile, the database demand is focused on unstructured data. "There is huge data growth as the digital age kicks into overdrive, but the vast majority of that data is unstructured; blogs, machine data, search indexes, media feeds, video, etc.," said Goldmacher. "The opportunity is to capture and analyze these data sets for the purpose of forecasting. This cannot be done well in a traditional relational DBMS...where Oracle is the undisputed leader."
  • On the apps side of Oracle's house, Goldmacher is also skeptical about future growth prospects. Why? High-end ERP is a saturated market. Meanwhile, other parts of the application equation---customer relationship, human capital and supply chain management software---are being dominated by on-demand software players.

Toss in a potential market pause ahead of Oracle's Fusion launch and there could be some problems ahead for the company. Goldmacher estimates that organic application revenue growth will be about 5.8 percent in fiscal 2011. The historical average is 23 percent and the last two year average has been a decline of 5.2 percent. Overall, technology licensing growth is expected to be 11.8 percent for fiscal 2011 a point better than the historical average, according to Goldmacher's estimates.

If anything, Goldmacher's analysis shows just how dependent Oracle is on maintenance revenue. In fact, Oracle's license revenue growth can fall further and maintenance fees can still support the company.

Goldmacher notes:

Given the recent organic decline in license revenue, one may wonder if software maintenance revenue contraction is far behind. This is highly unlikely as long as Oracle can generate $4B in license revenues which will generate roughly $1B in new maintenance fees to offset our estimate for 8% annual churn on $14B in maintenance. At present, Oracle generates nearly $8B in license sales. This assumes that Oracle can maintain its current maintenance pricing and there's no change in maintenance churn rates.

In other words, new licensing fees can result in maintenance gains until they hit a threshold that Oracle is not close to hitting. Everything works out nicely for Oracle as long as it can derive enough licensing revenue to offset maintenance churn and raise rates.

Goldmacher notes:

The dynamics exhibited by maintenance revenue streams suggests that there is a point at which Oracle's license revenues will no longer be able to support maintenance revenue growth due to pressure from churn. We are a very long ways away from that point.

Related:

Topics: Enterprise Software, Data Centers, Data Management, Hardware, Oracle, Software, Storage

Kick off your day with ZDNet's daily email newsletter. It's the freshest tech news and opinion, served hot. Get it.

Talkback

10 comments
Log in or register to join the discussion
  • RE: Oracle's potential math problem: Worrisome licensing trends ahead

    dxvx
    sahdkjashdkj
  • RE: Oracle's potential math problem: Worrisome licensing trends ahead

    they should have worked with open source instead of trying to bury it. Maintenance contracts on open source are legal and the revenue from the software sale was obviously not as important. Now it will eventually displace them.
    Socratesfoot
  • RE: Oracle's potential math problem: Worrisome licensing trends ahead

    Great post. This is exactly where we see the market heading and why our flagship product is a purpose-built database for unstructured information: www.marklogic.com
    kchest
    • Care to explain the theoretical background of your product?

      @kchest

      I would be interested in hearing an explanation.
      jorwell
    • Oracle cover your type of product too

      @kchest

      As Larry Ellison observed the women's clothing industry is far less fashion driven than the computer industry.

      Oracle have always been ready to supply a product to meet every fad.

      Some people have even been fool enough to buy these products though usually they give up using them quite quickly.

      If people started believing the earth was flat Oracle would be only too willing to sell them a flat earth geographical data system.
      jorwell
  • RE: Oracle's potential math problem: Worrisome licensing trends ahead

    I think a lot of these old school enterprise technology vendors are entering a slow decline. The next generation of IT professionals and leaders will have been weaned on totally different faster, cheaper, better technology architectures and interfaces and will seek to rip and replace it all over time. End users are already finding workarounds and innovating faster than IT departments are able to keep up. When vendors start celebrating lawsuits, you can almost hear the bell tolling in the distance.
    ladylaff
  • RE: Oracle's potential math problem: Worrisome licensing trends ahead

    Very well put. I would also add that in addition to the growth and new focus on unstructured data, companies are realizing that "immediately historical data" (like logs, call data records, stock ticks etc) which never require modification as soon as they are created don't need to be stored in RDBMS' like Oracle as the OLTP functionality is complete overkill. Similar reasoning goes for data that eventually becomes read only/static in nature which statistics show represent up to 90% of production databases, and are clogging up/slowing down those systems costing $$$$$ more than they need to. Our company RainStor (www.rainstor.com) is focused on retaining and allowing on demand access to historical structured data at economics 10x less than Oracle for example, further pressuring the rationale for renewing expensive maintenance contracts.
    Ramon Chen
  • The relational database isn't going away

    All that "unstructured" data has to be structured before you can start doing useful things like logical inference on it. <br><br>How do you structure it? With the relational model.<br><br><br>We will still be reading "the RDBMS is dead" stories in 2020.
    jorwell
  • Funny that all the replies to this are from vendors

    and not a single one from a customer demanding an alternative to the RDBMS?<br><br>I think "relational is dead" started in about 1994 with OODBMS, progressed to XML in 2000 and now we are beset with "no-SQL" DBMSs that are a throw back to hash table methods used on 1970s vintage mini-computer systems that weren't powerful enough to run ISAM.<br><br>It should be pretty clear by now that "relational is dead" is pure marketing hype and little or nothing has come out of all these proposed alternatives.<br><br>The relational model is based on sound mathematical principles and will endure long after these supposed alternatives are forgotten and replaced by the next short-lived trend.<br><br>However someone could definitely build a better RDBMS that Oracle, MS or IBM, no question about that.
    jorwell
  • RE: Oracle's potential math problem: Worrisome licensing trends ahead

    Great! !! thanks for sharing this information to us!
    <a href="http://www.yuregininsesi.com">sesli sohbet</a> <a href="http://www.yuregininsesi.com">sesli chat</a>
    birumut