Oracle acquires BEA; users lose middleware choice

by Dennis Byron on January 17, 2008

Oracle-BEA. Sun-MySQL. Now that the enthu­si­as­tic financial-analyst brief­ings and press con­fer­ences have been held (well at least the Sun-MySQL was enthu­si­as­tic) and now that we’ve had time to crunch the num­bers, let’s look at the details. Who wins and who loses in terms of com­peti­tors, part­ners, investors and users.

I ana­lyze the two deals in two sep­a­rate blog posts in order to pick up some sep­a­rate issues such as the depress­ing effect of the Sun-MySQL deal on open source soft­ware (OSS) pure­play val­u­a­tion and the depress­ing effect of the Oracle-BEA deal on WebLogic and Aqua­Logic rev­enues for the rest of 2008 (and maybe forever).

Oracle-BEA is dis­cussed below. But there is one major inter­con­nect­ing theme in the two deals: For users and investors the inde­pen­dent mid­dle­ware mar­ket as we know it is going away.

BEA pretty much launched the inde­pen­dent mid­dle­ware mar­ket as a sep­a­rate entity in 1995 by acquir­ing sev­eral Bell Labs/Novell TUXEDO-based dis­trib­u­tors, IMC and ITI. To be com­pletely accu­rate, investors still have a few choices. TIBCO (TIBX) is left because Reuters spit it out a few years ago. And Iona (IONA)–which pre­dated BEA–and Cape Clear are still stand­ing pre­sum­ably totally because of Irish hard­head­ed­ness. In addi­tion, IBM (IBM) and Ora­cle have suc­ceeded in chang­ing the def­i­n­i­tion of mid­dle­ware such that some might argue that Attunity (ATTU) and SAS Insti­tute make middleware.

When I say mid­dle­ware, I mean run-time code that’s really in the mid­dle of some­thing. And when I say inde­pen­dent I mean “pure play,” soft­ware sup­pli­ers not sell­ing other types of infor­ma­tion technology.

So what does this sea change mean for the users? In the 1990s, users did not buy from pure plays because they were pure but because they offered a choice other than–at the time–depending on the mid­dle­ware built into appli­ca­tion suites such as SAP R/3, built into their rela­tional data­base, or spit out as a by prod­uct of their mostly systems-supplier-centric devel­op­ment tools. The bot­tom line for the user com­mu­nity now, as I have been mod­el­ing for years, is that because of the end of the inde­pen­dent mid­dle­ware mar­ket each enter­prise has to choose whether to get in bed with an appli­ca­tion sup­pli­ers’ middleware—SAP’s NetWeaver or Oracle’s Fusion—or an infra­stru­ture sup­pli­ers’ mid­dle­ware. (The above para­graph sum­ma­rizes exten­sive por­tions of a Research 2.0 client deliv­er­able, so if you want to dig deeper, give us a call.)

Larry Elli­son explained in his conference-call script how Ora­cle fits in both camps because of his com­mit­ment to open stan­dards but that is not the same as open source. And there­fore the Ora­cle approach does not really rep­re­sent inde­pen­dent choice for the user.

Part­ners such as smaller ISVs that used BEA mid­dle­ware in their prod­ucts and even much larger com­pa­nies than BEA that use BEA mid­dle­ware in their ser­vices offer­ings face the same more lim­ited mid­dle­ware choice as users.

For investors, the inde­pen­dent mid­dle­ware mar­ket was a great lead­ing indi­ca­tor of the over­all IT mar­ket. This acqui­si­tion makes the chal­lenge eas­ier or harder depend­ing on your investment-research tech­nique. Do you try to under­stand the upstream/downstream aspects of the mar­ket when you are invest­ing in auto mak­ers? If so, your life just got harder because there is no more equiv­a­lent of Behr, Dana (DCNAQ), Del­phi (DPHIQ), Eaton (ETN), Honeywell/Bendix (HON), Magna (MGA), Modine (MOD), Tower (TWRAQ), ZF and so forth in the IT mar­ket. But if you just bet on an auto maker because you like the looks of the lat­est coupe, and an IT sup­plier because the user inter­face looks cool, you’re all set.

The most impor­tant short term invest­ment aspect of the deal is that it will prob­a­bly kill BEA license sales until the deal is signed, sealed and deliv­ered in about 6 months. I think the two com­pa­nies rec­og­nize this, account­ing for the dif­fer­ence between the $21 ask­ing price and the $19.375 final bid. And the two com­pa­nies will prob­a­bly make every effort to pre­vent it from hap­pen­ing (but I’m not sure of that given the cold­ness of the joint con­fer­ence call).

As for com­peti­tors, well, as I said above, there are none left– Den­nis Byron

Tags: , , , , ,

{ 1 comment }

Geva Perry January 18, 2008 at 6:06 PM

Dennis —

There is an alternative to the middleware mega-vendors — IBM and Oracle. For years now alternative open stacks have been emerging and coming of age. They include open source elements such as Spring, Tomcat and Hibernate. And they include commercial products such as our own GigaSpaces XAP.

We just published an Open Letter to BEA WebLogic Customers, which discusses this, on our company blog: http://www.gigaspacesblog.com

Comments on this entry are closed.

{ 3 trackbacks }

Previous post:

Next post: