Download the authoritative guide: Cloud Computing 2018: Using the Cloud to Transform Your BusinessSpring is in the air, and that means it's time for my annual IBM column, in which I state that IBM needs to figure out what role it really wants to play in the enterprise software market. Is IBM a counter-revolutionary force, turning back the clock to the glorious days when all software was custom-built and "packaged software" was an oxymoron like open standards and application integration? Or is it, based on the recent agreement to SAP-ize IBM's flagship DB2 database, the best friend enterprise software ever had?
This year is starting out with IBM in full bipolar mode. IBM's retro vision of the software world has been fueled by a host of acquisitions and announcements over the past year, and by an ongoing marketing juggernaut that is giving Regis McKenna's quintessential marketing bible of the 1980s, "Who's Afraid of Big Blue," a second lease on life. Meanwhile, the partnership with SAP shows that the "good" IBM can work closely with its leading partners to the benefit of all.
The split was in full force during a recent analyst briefing on service-oriented architectures, in which Global Services described an IBM-centric world that further refined the notion that all innovation comes from a Global Services custom-developed project, based on WebSphere. Meanwhile, the packaged software from the SAPs and Oracles of the world is relegated to a commodity -- read non-strategic -- support role.
When asked what guarantees Global Services was willing to give that it was going to remain neutral in the selection of a platform -- bearing in mind that its big applications partners are also selling infrastructure to compete with WebSphere -- you could hear the waffling noises loud and clear. First they said that WebSphere was the preferred platform, but then they clarified that by saying they were technologically agnostic, and that technology wasn't really a major part of the initial SOA engagements Global Services was working on. Then they said...Well, it doesn't really matter. The basic problem is that one part of IBM wants to be friend number one to its enterprise software partners, while another part is increasingly packaged software enemy number one.
How long can this bi-polar disorder remain untreated before a major meltdown occurs? Well, part of the answer lies in understanding the value of the relationships that IBM GS has with its enterprise software partners. SAP is probably the best example: I've heard that IBM earns anywhere from $4 billion to $7 billion in revenues from its relationship with SAP. Meanwhile, IBM's total software revenues for 2004 were $15.1 billion.
Bear in mind that IBM's software portfolio includes its DB2, Rational, Tivoli and Lotus products, as well the WebSphere products that compete with infrastructure partners like SAP. It's hard to get a specific breakdown of IBM's total WebSphere revenues, but it's pretty safe to assume that SAP's value to IBM is at least as large if not larger than WebSphere's value to IBM.
So if you were a bean counter in Armonk, what would you say about dissing the gang from Walldorf?
I'd look at SAP's role in IBM's $46 billion in services revenues, and its $15 billion in software revenue, and I'd start asking if this golden goose is really worth trying to slaughter. Of course, coopetition is an old tradition, and IBM and SAP are both mature enough companies to understand that they can play on each other's varsity teams even as the junior varsity is mixing it up on the field.
But in the end, there needs to be a much more consistent IBM voice about the enterprise software space that clearly -- not schizophrenically -- delineates who are its friends and who are its foes. There are some clear enemies that raise no questions in IBM's strategic world view -- Oracle and IBM have little at stake in each other's markets, other than some relatively small potatoes in the PeopleSoft and JD Edwards customer base. But for the companies that matter a lot to IBM, like SAP, I think it's time to show a little more love and a little less fang. Because, at a minimum, the "bad" IBM custom software view of the enterprise software world isn't just bad for business, it's also bad for customers too.