It looks like Oracle/PeopleSoft is a done deal. But why would Larry Ellison spend $US10.3 billion for overlapping technology? For me, the answer comes in two parts.
In the short term, Oracle wants PeopleSoft's customer-support business. Trolling for new licensees is expensive, but servicing current users is a high-margin endeavour.
If you talk to Josh Greenbaum, an analyst at Enterprise Applications Consultants, he'll tell you Oracle needed this acquisition badly. Almost all of the application vendors get one-third of their revenue from maintenance, one-third from service, and one-third from new licences, Greenbaum says, adding, "The real money is in the installed base, and you need that critical mass to keep that rolling."
So it's certainly not a problem for Oracle to promise to support PeopleSoft users for the next 10 years. In fact, it's very profitable. Although 10 years of support sounds like an eternity to some, many users I spoke to echoed the feelings of one executive, who asked not to be named: "The Oracle statement that it will support PeopleSoft for 10 years clearly indicates that it has no long-term interest in PeopleSoft."
Yet another user who asked not to be named recommended turning to a third party, rather than going with Oracle support. "It could save a lot of money."
I especially like this response from Atif Siddiqui, an application specialist at a company I can't name: "Given billions of dollars of investment in this area, I wouldn't be surprised to see some small mum-and-dad shops grow up offering support for 'unsupport' of PeopleSoft." In other words, migration services.
Siddiqui believes third-parties will also offer support services for unsupported PeopleSoft applications. "You can almost think of mini-replicas of PeopleSoft being formed in the next two or three years by PeopleSoft junkies."
Support is one thing, but what Oracle has not promised is to continue R&D for PeopleSoft products beyond the next release. So although all the PeopleSoft users I spoke with recommend a wait-and-see approach before transitioning, the truth is that, at some point, transition they must.
As Tom Hins in applications solutions at the University of Oklahoma put it, if you don't migrate to Oracle you'll have no choice but to stick with your current PeopleSoft version indefinitely. But he adds this bit of positive thinking from his own first-hand experience migrating in the opposite direction: "Having worked through Oracle Version 7 and designing systems with Designer/2000, I found the transition to PeopleSoft relatively easy."
Because of its huge installed base, SAP has a good technology story to tell with its NetWeaver platform. It can say to its customers, "you use so many of our applications you should be running on our platform". Once it transitions the current PeopleSoft installed base, Oracle, by doubling its number of licensees, will have an equivalent story to tell.
In Greenbaum's view of the world, all of the companies that sell technology platforms -- in other words, the middleware stack -- should be worried about Oracle's buyout of PeopleSoft.
Greenbaum says if he were IBM he would be worried; if he were BEA Systems he'd be scared to death. And my favourite quote: "If I were Microsoft, I would wonder what I was doing in the applications business."