One of the more interesting questions to arise out of Oracle's acquisition of Siebel Systems is what will become of the technological marriages between Siebel and IBM that fuel some of Siebel's existing offerings. Although salesforce.com CEO Marc Benioff was one of the first to point out the wrench that marriage throws into the works when it comes to Siebel's CRM OnDemand offering (which Siebel just updated), Siebel's reliance on IBM technologies extends into its analytics offerings as well. The Siebel-IBM union was consummated two years ago when the two companies announced (also, see CRN's Siebel, IBM may team on online midmarket CRM) that Siebel's CRM OnDemand would rely on IBM's Java 2 Enterprise Edition (J2EE)-based WebSphere and IBM's DB2 relational database management software. IBM's Global Services outfit was pegged to host the solution on its systems and IBM's version of Unix -- AIX -- was clearly identified as another part of the formula when news of the analytics collaboration arrived.
So, now that Oracle will essentially inherit an IBM partnership that involves at least two IBM solutions (WebSphere and DB2) that go directly head-to-head against two of Oracle's core offerings (its J2EE-based application server and database), inquiring minds want to know what will happen next. While Benioff was the first to highlight the potential conflict of interest, IDG News Service's Stacy Cowley may have been the first to give it the due diligence it required (see Siebel's Oracle union may end its IBM OnDemand alliance). According to Cowley, "Executives from both Siebel and Oracle say the hosted CRM (customer relationship management) service will go forward, but IBM is likely to be left out of Siebel's OnDemand future." Likely? That's the understatement of the year. With DB2, IBM is second to Oracle in database marketshare and is always looking to close the gap. Anybody who knows Oracle CEO Larry Ellison also knows that it'll be a cold day in hell before he lets a "trojan horse" like DB2 behind his firewall for very long (let alone WebSphere).
In the news story, Cowley quotes Siebel senior vice president Bruce Cleveland as saying:
Siebel CRM OnDemand was built on Siebel 7.5.2, and the fact is that that product line runs on any database and any application server......we created a partnership with IBM and we chose for that partnership, for obvious reasons, to run it on WebSphere and DB2. There is no technical linkage. We can make a different decision going forward if we choose.
If what Cleveland says is true -- that there's no technical linkage -- then it occurs to me that this could be one of the better portability tests for the J2EE promise that Sun says is critical to the survival of the J2EE ecosystem (and that's one reason that Sun will not buckle on either compliance certification or trademark licensing). If there's no technical linkage at all, then Oracle should be able to:
- Remove the code and data from the systems belonging to IBM Global Services and load them onto something more aligned with Oracle like Solaris or Linux (how about a new AMD Opteron-based Sun Galaxy server that Sun president and COO Jonathan Schwartz says doubles the performance and takes less than half the electricity at less than half the acquisition cost of a similarly equipped Dell system?)
- Drop in Oracle's J2EE-based application server in place of IBM's WebSphere
- Drop in Oracle's database in place of IBM's DB2....
with almost no pain whatsoever, particularly on the J2EE and database fronts. The other option for Oracle is to kill the Siebel/IBM offerings altogether, given that they'll be redundant with Oracle's existing solutions (or at least Larry Ellison's investment portfolio). But, according to Cowley, Siebel's Cleveland is clearly saying that this won't be the case and Ellison reiterated that Siebel's CRM OnDemand offering is something that Oracle intends to invest in, not shut down.
All I can say is that If I were Jonathan Schwartz, I'd be salivating at the opportunity to turn this into a case study and a justification of why Sun can't fully release Java to the open source community. On the flip side, if what Siebel's Cleveland says isn't true and there is a lot of proprietary technical linkage in the way of making such substitutions, then the case study would even further emphasize the J2EE standards and compliance story while exemplifying the risks of going off the ranch (venturing into the J2EE proprietary extensions territory where there's additional functionality, but significant pain involved in switching).