Tuesday, 16 October 2007

BEA and Oracle: 0.5 + 0.7 = 0.8?

The problem with takeover bids is that once made, things can never be the same again for ‘target’. At this point, I guess we cannot be absolutely sure that the proposed acquisition of BEA by Oracle will go ahead, or whether other potential buyers will enter the game as James is encouraging with his mischievous egging on of SAP.

Whatever the outcome, the shame is that BEA is now going to find it hard to shake off the cloud of uncertainty. I say ‘shame’ because despite my disagreement with some of the na├»ve views expressed around packaged applications, I quite liked what I heard in terms of core strategy at BEA World a couple of weeks ago when the ‘Genesis’ story was presented. BEA seemed to have a good solid view of what its customers needed in terms of an SOA based middleware ‘fabric’ that is generally agnostic of specific technologies and applications. Even though Genesis was presented as a work in progress, things seemed to be heading in the right direction.

Meanwhile, as Neil points out, Oracle has had some gaps in its own middleware portfolio being pulled together under the ‘Fusion’ banner. We also know from recent Freeform Dynamics research (which we’ll be publishing soon) that, contrary to the way Oracle often spins the numbers, Fusion middleware adoption is pretty much exclusively aligned to Oracle application incumbency, i.e. there is very little penetration into organisations that do not use Oracle EBS, PeopleSoft, JD Edwards, etc. In this respect, there is little difference in the position of Oracle versus its main application rival, SAP, whose NetWeaver offering is similarly aligned to application incumbency.

So what happens when we put all this together?

Well, according to my own admittedly very subjective metric, I reckon I would put BEA at 0.7 on a scale of 0 to 1, where 1 would indicate an ideal set of open enabling ‘middleware’ solutions to form the linchpin of a future-proof corporate IT infrastructure. I don’t think BEA would argue too much with this – the guys there articulated some ambitious plans but acknowledged that there was still much work to be done.

Turning to Oracle, I think Neil is right when he highlights the solution gaps, but would also call out the challenges Oracle has been having in being taken seriously as an ‘independent’ vendor in this space, given the application alignment we have seen – hence I would put Oracle at 0.5 on our notional scale.

While bringing the BEA into the mix would round out the Oracle offering, there is a corresponding a risk that it would also undermine BEA’s positioning as genuinely independent option, especially given Oracle’s almost rabid competitive stance against SAP. There is then, of course, the obvious redundancy between the two portfolios that will need to be resolved in one way or another. Oracle seems to have got away with the ‘Apps Unlimited’ strategy based on maintaining multiple packaged application code lines, but ‘Middleware Unlimited’ would be stretching the concept beyond the realms of credibility – as well as stretching Oracle’s ability to manage an ever more fragmented R&D effort.

So, the acquisition arithmetic is probably something like 0.5 + 0.7 = 0.8. Don’t take this too literally, I am just trying to make the point while there might be some net overall goodness generated if the acquisition proceeds, the end result is not going to be the answer to everyone’s prayers.

Meanwhile, the most obvious beneficiary in all this is IBM, who can just sit there smiling as the one remaining genuinely independent middleware gorilla, unless, of course, you include Microsoft in the picture, but that’s another story.