IBM, Sun, Microsoft, BEA, Hewlett-Packard, Oracle and other vendors will be duking it out at the Butler Group Web...
Services Symposium in London this week. Or will they? Neither of the architectural enemies -- .NET and Java -- is likely to prevail over the other. All players will also have to agree to use the same standards for Web services, or systems-to-system communication simply won't happen.
Context: So how will vendors differentiate their strategies? 'Cooperate on standards and APIs, compete on implementation,' is the stock answer. But already the battle lines show a much more strategic deployment of forces and resources. IBM and Microsoft have gotten hitched again.
Butler Group chairman Martin Butler argues the Web services marketplace will be characterized by this and other cartels like it. It will be primarily a B2B phenomenon -- witness Microsoft's already defunct Hailstorm/MyPersona initiative. It will be characterized by offshore development, the commoditization of horizontal applications and the creation of component and assembly industries.
But why has IBM chosen to reestablish a strategic relationship on Web services with Microsoft and not with Sun, even though Java is underpinning IBM's own Web services initiatives in WebSphere?
Sun and Microsoft are just technology providers. Sun has servers -- but no services and no history of executing a software strategy. Microsoft on the other hand owns the on-ramps -- the desktop and in future the Xbox or other clients. IBM doesn't need servers; it has those. It has services, it has software, but it also needs to be able to leverage the on-ramps, primarily Microsoft Windows clients.
Financial impact: Technology vendors need to maintain and develop revenue and earnings growth. With companies already spending much of their profit on IT -- in some cases 100% -- that figure is only going to decline. So what can technology companies sell instead? They can sell the services that users currently spend fortunes to develop and maintain internally -- as Web services.
Out of a total $1,400 billion in IT expenditures, US corporations spent $1,022 billion on developing internal IT systems and processes, according to Butler Group figures. Sales of hardware, telecom, services, packages and programming products made up the rest. The $141billion US software industry has its products implemented mostly by third-party intermediaries. The opportunity is to turn these into Web services, deliver them directly over the Internet and suck up the middleman's money.
Butler Group forecasts 25% of this internal spending will be converted to external spending over the next 20 years. This is a $500 billion market opportunity for Web services. And the role of the CIO will shift toward the procurement of external services. If it sounds like a comeback for the ASP/utility model of yesteryear -- then it is, Butler argues.
So if software is destined to become a service, who is most likely to want to own the delivery? IBM. And it's no coincidence that integrators such as Accenture are playing an unprecedented part in the development of Web services standards.
Market assessment: That IBM will be the number one player in Web services looks like an odds--on bet today. But where is the rest of the field? Microsoft is in second spot, driving thought leadership and mind share in conjunction with IBM. Look at their strategic cooperation on Open Grid Services Architecture (OGSA), Web Services Interoperability (WSI) and WS--Security.
Sun has been left sniffing their exhaust, with little more than a me-too strategy at this point. We've heard much about rebranding, but little about the meat and potatoes of what it is actually doing or its differentiation.
And what of the rest of the pack? Oracle has been atypically quiet and has had little input into Web services processes. Apart from saying that it will make its applications available as Web services it's not said much at all.
Likewise Hewlett-Packard. It's begun giving away its application server, but that's mostly because it has little other hope of getting any significant market share from IBM, BEA or Sun. It affords a vision of utility computing but has said little specifically about Web services.
Conclusion: Stripping away the hype, Web services equals super-XML integration by another name; it is a natural extension of process automation. A SOAP bubble perhaps, as the title of the451's current Web services report suggests. But nevertheless it's the catalyst that at this point looks likely to drive a significant realignment of the IT industry.