A CIO scorecard on utility computing

At the CeBIT America conference held in New York at the end of May, I chaired a conference track on "Leveraging the New Services Economy," which afforded me a valuable opportunity to listen to a number of CIOs present their views regarding utility computing. Many of the CIOs participating in the event reported that they are already generating real benefits from this new IT architecture and are rapidly expanding their utility computing initiatives.

The presenters also testified to several of the emerging trends that I've called out in previous articles: they are fed up with the cost, complexity and hassle of their legacy IT operations; they view the migration toward a utility computing model as synonymous with a move toward an outsourcing model; they recognize that the enabling technologies associated with utility computing are still in an embryonic state, and thus implementing utility computing is an incremental process; they believe that a phased deployment is preferable because it mitigates risk and allows for better measurement of results; and finally, they recognize that utility computing not only changes how the IT operation works, but also requires that enterprises revamp their business processes to fully leverage this new IT architecture.

The CIOs who participated in my conference panel and others voiced frustration over the impossible chore of managing a diverse array of hardware and software products and custom applications. The day-to-day integration, maintenance, and management of legacy environments have precluded IT organizations from optimizing the performance of their existing technology, delivering reliable business functionality, and supporting corporate objectives. At the same time, IT budgets have tightened, staff resources have been cut, and business demands for new IT capabilities have escalated. These countervailing pressures are what spurred many CIOs who were a part of the CeBIT conference to move to a utility computing model.

In addition to a shared catalyst, these CIOs also reported taking the same initial steps toward utility computing. First, they are consolidating and standardizing their IT infrastructure, end-user computing platforms and primary business applications, which is helping them regain control of their IT operations and generate significant cost-savings. Next, they are implementing automated service provisioning, performance measurement and reporting systems, which are essential for realizing the on-demand features of utility computing and for metering actual IT utilization rates for direct or indirect chargeback. With these tools in place, the internal IT organizations, along with their external solution suppliers, can establish meaningful service level agreements (SLAs) and service pricing structures.

While their migration paths share certain common elements, the CIOs at
CeBIT also showed how enterprises can leverage various forms of outsourcing
to achieve their utility computing objectives. The CIOs of Fluor Corporation and the County of San Diego are both in the midst of seven-year mega-outsourcing agreements to make utility computing a reality. Fluor is working primarily with IBM, while the County of San Diego selected a consortium of major vendors led by CSC to meet its needs. Ingersoll-Rand has leveraged the application hosting and management service capabilities of Corio as the centerpiece of its utility computing initiative. The CIOs of New York City and its police department are using a variety of hardware, software and telecom suppliers to rapidly upgrade their overall IT operations and slowly shift their day-to-day maintenance and management responsibility to their suppliers.

Despite these success stories, some CIOs speaking at CeBIT remained skeptical of their technology vendors' ability to create true on-demand solutions or even to fulfill their outsourcing promises. I certainly don't begrudge them their skepticism, but I hold that the significant progress and tangible benefits reported by many of the CIOs at CeBIT represent additional proof-points of the tremendous potential and real progress of utility computing.