IBM Global Services to fuel Web services push

IBM is in the throes of readying an all-out initiative designed to help users accelerate the implementation and management of Web services with IBM Global Services (IGS) unit serving as the focal point. The company says this results from the confluence of several industry trends that it believes will finally propel Web services into the corporate mainstream.

Through the upcoming initiative, expected to be launched this fall, IBM hopes to put the word "services" back into Web services, making the point that there is more involved for a successful strategy than just industry protocols.

The initiative will see IGS working in concert with all of IBM's major hardware and software divisions and with the company's network of consultants, which will result in a range of industry-specific Web services solutions. The initiative is seen as part and parcel of strategy of going to market "industry by industry," and that will also be able to leverage the expertise of the recently acquired PwC Consulting.

IBM's continued hard focus on services, in the context of its Web services push, appears justified. According to market researcher IDC the professional Web services market is expected to reach $11 billion by 2007, up from the current US$206 million. IBM officials believe this leap will occur because corporate IT shops will begin much larger cross-departmental integration efforts instead of doing one-off integration projects for ERP or CRM.

Michael Liebow, the executive in charge of Web services for IBM Global Services sat own with Ed Scannell to outline some of the emerging industry trends driving the company's Web services plans as well as the importance of Systems-Oriented Architectures (SOAs) and how both will contribute to IBM's greater On Demand initiative.

Q:What is behind Global Services' renewed push for Web services?

Liebow: With the PricewaterhouseCoopers acquisition, Web services maturing, and IBM's software unit, Microsoft, and other vendors having done a great job with the industry bodies laying out infrastructure standards, many corporate users have crossed the line of just toying with Web services to actually implementing them. What IGS is at its core anyway is an applied technology company. It makes sense given our business model to make a strong move here now. What is giving us further impetus is the recent announcement of Websphere Business Integration (WBI) Server that gives users greater support for implementing (Web services).

Q:Howready are most corporate IT shops right now to start implementing a meaningful Web services strategy?

Liebow: Generally the skills that need tweaking are in the area of SOA XML schemas, UDDI, service bus designs, and things along those lines.

Q: What will this renewed push mean in terms of working more closely with IBM's product groups?

Liebow: Interesting question because what Web services allow you to do is create more assets. And as you saw with the WBI announcement, they are putting more and more features that are specific to an industry into that product. So those assets will be field-level created. We will work more closely with the software group in order to take field-created assets and turn them back into products. So there will be this cycle going on that creates that value and continues to build on that.

Q:How does this effort fit in with the company's ongoing On Demand strategy?

Liebow: When you think about enabling an On Demand environment Web services is critical to that. And if you think about the underlying platform and architecture there are two aspects to consider. One, which is at a high level, is this notion of a component-based company. In order to be components based, you must have a services orientation. And the services orientation is based on the architecture, and that architecture must extend beyond just the IT layer, applications layers, and into the business processes and business services.

Q:How important a piece is SOA in all this?

Liebow: It is very important. When you think of an SOA you think about things from the governance layer on down to the business services to the processes underlying that, to the apps supporting it, to the IT layer. You must create alignment across that whole stack. If you think about how PCs came into the enterprise, they came in through the back door. They were not covered by any central IT budget. You took it out of your typewriter budget. What happens later is you realize what the true cost of managing all of them is. So think about Web services like that. You can design Web services to do one little thing and cover it through an ad hoc budget. But if you don't have an SOA in place, you'll wake up some morning and find all these little connections running around which will seriously degrade the performance of your systems because they are poorly designed and not working together.

Q: So how hip are most corporate users to SOAs?

Liebow: In our experience it varies. It's all over the place. Maybe the financial sector is the furthest along relative to understanding them. Typically there are two approaches. There are going to be companies who understand the overall architecture and the implications of it. And then there will be companies that are just trying to survive quarter to quarter and they need to fix a connection and they need it fixed fast. They are not going to care if it is called Web services, they just need value. The reality is people have not tied Web services into an SOA, but once you do, once you have a services orientation for your company, that is when you unlock the Holy Grail.

Q: Will there be any productized offerings from these upcoming Web services that address things like business process integration?

Liebow: When we are talking about Web services on one hand you are talking about the connections, and on the other hand you are talking about the development and creation of actual business services based on the layer of industry semantics. Those things will be created with an industry partner, typically the dominant player in that industry. Some of things we now have under way.

Q: Do you envision delivering a CD, for instance, which would contain a collection of best practices for a specific vertical market?

Liebow: If you are looking for something to be shrink-wrapped -- business in a box -- there will be a set of tools and practices. There might even be licensable practices in order to extend to small and midsize business, but that day has not come yet. But for heavy duty stuff for enterprises? These are things that still require customization.

Q: Do you plan to augment the development of WebSphere as part of this effort?

Liebow: The infrastructure standards we are talking about are either in WebSphere today or are planned to come out in future versions. That layer is critical. As we get into field level engagements and we find that say the data parser needs to be improved because of what we learned with a customer, that feedback will be recycled back into the product. So that relationship is critical.

Q: Some reports from analysts earlier this year state that Web services are not meeting their forecasted potential. One of the reasons is they cite is they are not as easy to implement.

Liebow: First, people have to start thinking of Web services as a services play. I think people are under the misconception that it is easy, that they can do it themselves, and it just isn't that complicated and that service providers are going to be out of business. But we still believe that the services opportunity for Web services will be some multiple (in terms of revenues) higher than the software opportunity. In the past a typical organization could do half a dozen projects given their existing technologies. But with Web services they will be able to do many more than that. The complexity will not be in individual projects, but instead in the sheer volume of them. Web services are not materializing like some had hoped, but the real dollar value to us is huge to make all that actually work.