Last month I speculated that Oracle was preparing to abandon the traditional software license model for a subscription based scheme. I based that speculation on a couple of observations, the first being that Ellison has been devoting a lot of his public speaking to reinforcing a notion that Oracle already generates a significant subscription revenue stream. Second, the simple fact is that Oracle gets their ass kicked 70% of the time they encounter SAP in a competitive sales situation, and the acquisition spree they have been on hasn't changed that dynamic, in fact, over the last couple of quarters our win rate has actually increased. Lastly, Ellison is a smart man and he sees the trends pretty clearly with regard to SaaS and open source, both predominately based on a subscription model.
The idea that Oracle has subscription revenue today is somewhat of a shell game because what Ellison is doing is suggesting that maintenance revenue is subscription revenue. Technically it is, but it's a stretch to suggest it is of the same ideology. Every public company software executive that has maintenance revenue is faced with the same dilemma: maintenance rev is valued at 2-2.5x while anything called "subscription" revenue is valued at 3-5x by Wall St., so it's entirely logical that Ellison would try to reposition his maintenance stream as subscription revenue in the best interests of his shareholders (which is primarily himself). Heh, if your stock has been stuck sideways for 6 months despite all the attention you have generated you would be looking at any option that could improve your valuation without requiring you to actually execute your business better.
On top of the pragmatic valuation concerns I am quite sure Ellison is looking at the success of companies offering their products as services and concluding that removing barriers for customers to get on your products quicker is a good thing. This is where the open source components start to enter the picture because they have zero barriers to acquisition and have built up a significant user and developer community as a result. Acquiring the LAMP stack (minus MySQL) provides Ellison with a significant pipe to drive his Fusion middleware products through, and ideally draft success in applications as the ultimate goal. It is difficult to execute on a hybrid services/license model to the scale that Oracle would require, therefore it's even more logical to assume that Oracle is paving the way for this eventuality by moving to subscriptions on a defined path rather than a disruptive reaction to events that they would have set in motion with the acquisitions they are apparently making now. There is a lot more to their open source moves, IMO, than simply an economic model to be sure.
Yesterday there were a couple of analyst notes, one from First Albany and the other from Goldman, suggesting that Oracle was in fact preparing to abandon the license model, even if just under the guise of "subscription accounting". Just like you can't be a little bit pregnant, it's evident that you can't call it subsciption accounting without it actually being a subscription license option. Furthermore, any subscription revenue plan is dependent on growing deployments, and even though subscription licensing can even out the quarterly volatility that software companies fight, attention still goes to unit shipments as an indicator. As Sherlund points out, their valuation models are based on cash flow.
Today Oracle "clarified" the issue by saying they weren't going to move to a subscription licensing scheme, although it's really hard to swallow the notion that they just gave up on it after a few analysts came out critical of the move. My suspicion is that nothing changed in the plan other than the timing and way they communicate it.