IT执行官生存指南(1)

Exclusive Interview with Gartner Analyst Gene Phifer 1. To what extent is today's enterprise IT infrastructure adequate to support the business demands of tomorrow, and what should CIO's be sure to have in place? Most of the infrastructure in enterprises today is not capable of handling the business demands of tomorrow. Businesses are looking for agility, the ability to address new business opportunities, develop and launch new product offerings and ideas, and pursue new channels of interaction very rapidly. All too frequently, IT has been an impediment to business agility. IT systems and infrastructures have been very inflexible, and require months to years of evolution before allowing businesses to adopt new channels, new products and new strategies. Obviously we need to change that. We need to extend the way we have architectured infrastructures and applications, and the way we deploy technology in support of new business opportunities. That is going to require a new level of interoperability and new capabilities for integration and adoption of service oriented architectures in order to build and architect agile systems solutions. We will assemble and reassemble services as necessary to meet changing business objectives or business opportunities. In parallel, the technology investments that companies have already made must be preserved, while infrastructures must evolve to focus on service oriented capabilities. IT will utilize web services as one of the primary methods of interoperability, and will allow legacy applications to extend into the future in a way that offers flexibility. Hence, today's IT infrastructure will seriously change for tomorrow. 2. After the last two years of \and spending priorities? The last couple of years have not been pretty from either an IT perspective or a vendor perspective. There has not been much money spent on technology – not nearly as much as the vendors would have liked. However, we are seeing 2005 as a turnaround year in allowing enterprises to take advantage of investment spending, not just in maintenance, but in new capabilities, technologies and applications. We have to be smart in how we spend this new budget money because first of all, there is not much of it. Second, there is going to be a real hard view of how this new investment spending is going to derive long term benefit for the organization as well as deliver short term measurable value. So we have to look at specific types of application capabilities that are necessary to meet business objectives. From an infrastructure perspective, we also have to look at how we can evolve to deal with the organization's agility needs. How are we going to invest in our infrastructure to obtain that level of agility and to utilize service oriented approaches that lean heavily on web services? Also, how will this investment allow us to build composite applications so existing applications and new business logic can be reassembled in new and creative ways that will meet the demands placed on IT to deliver realizable, visible, short term business value? We are looking at investment spending, not spending that is going to be realized five years from now, but spending that will show returns within a 12 to 18 month timeframe. One set of applications meeting those criteria is what we call gap applications, or gap apps. Gap apps fall in between your traditional stovepipe suites: ERP, CRM, supply chain, or partner relationship management suites. They include all those three letter acronym suites we have deployed over the years, as well as those equivalent legacy home grown applications used to solve similar business problems. Enterprises have many business processes and functions that happen in between those suites, in what I call \gap.\an area where we see composite applications coming in, because in most cases gap apps are composite applications, a class of application that we call packaged composite applications. These can be horizontal in nature; project management and staffing are examples. They may also be vertical in nature; for example, retail support. Different vendors will deliver different sets of gap apps to target where they can best differentiate themselves, deliver the highest value and highest return, and make the most money in the process. 3. In the May 10th, 2004 issue of Business Week, there was a published cover story called \Strikes Again.\ Yes. During the Internet gold rush (1997-2000), a tremendous amount of money was thrown into a black hole with little-to-no possibility of return. But people did it anyway because that is what they thought they needed to do to stay competitive. E-business, specifically e-commerce, became huge, and a plethora of companies deployed e-business and e-commerce strategies, then figured out they didn't know what they were doing. The dotcom meltdown occurred in 2000, giving e-business a bit of a black eye. Consequently, it has been relegated to the back burner for the last few years. However, recently we have seen a resurgence of e-business. Companies that entered e-business during the gold rush now see their technology growing a little stale. They are looking for the next generation of e-business and e-commerce functionality, and the guys who got into it and stayed in it are looking for the next generation of these technologies. The companies that either looked at e-business and avoided it, or maybe got into it, played with it, and then got out of it, are revisiting that decision. We have now seen a lot of maturity evolve in the world of e-business. We have seen e-commerce websites grow up, merchant servers become extremely powerful, and certain e-business sites have become part of the vernacular, such as Amazon and eBay. E-business concepts are now common concepts. Another angle is the aspect of self-service. It is expensive to run storefronts, call centers and back office staff to handle interactions with customers, suppliers, and employees. From a retail perspective, the people cost alone is a tremendous expense. By shifting funds from high cost, people intensive channels to low cost online channels, we start to see significant productivity gains. At the same time, we see increased customer satisfaction. So it looks like e-business will no longer be a dirty word in 2004. The black eye has healed nicely and we will continue seeing investment spending in e-business for at least the next few years. 4. Do you see more line of business ownership, and demand for the development of custom applications that deal with unique business requirements? Also how should a CIO balance conflicting demands and limited resources? Applications should always be driven by the business. Unfortunately, the historical approach has been that many applications were driven by IT. Often, those applications were either unsuccessful or only partially so. A line of business application needs to be owned and driven by the business – period. This has been a significant visible trend in the last few years while IT has struggled to align itself with the business. In some cases, IT has been considered an enabler of these kinds of applications. In others, it has been perceived as an inhibitor. Traditional development methodologies and development tools used by IT can lead to a multi-year development project, which has historically been required to get large business applications and solutions out the door. Getting back to the agility theme, business needs to be agile and needs IT to be agile in order to deliver the right kind of technology solutions to support the business. In many cases, we are going to see business analysts actively engage in the development of new composite applications and processes. Business analysts have always had the ability to be a part of the process, to sit down with IT to generate business requirements and then hope the solution meets the need. We are talking about a different role here for business analysts. Through the advent of service oriented architectures and web services, plus the evolution of development tools, we will start seeing business analysts actually build new processes in technology and reengineer processes?without needing a developer do it for them. Developers will obviously still have a major role here. They will still build the core components and services of the line of business applications and systems. However, the assembly of many applications will be driven by the business analyst. They will be able to create new processes and new applications using visual assembly tools, as opposed to the traditional Java or C++ or C# development effort. This is a major change in the way we deal with the business, and I think a welcome change. In spite of this, some IT departments will see it as a threat. IT needs to participate heavily to manage this, to make sure that end users cannot program themselves into a corner, or bring the system down to its knees because they did something wrong. However, with these constraints in place, IT should provide as much autonomy as possible, allowing end users and business analysts to do as much as they possibly can on their own. Since it is an exceptionally different approach to building and deploying business processes and business applications, the changes to both the business and IT must be managed carefully. 5. With business process improvement emerging as a key issue for next year, what does this mean for IT? And does it require a change in IT skill set? When you look at IT skill sets, and how applications are built, designed, developed and deployed, you are going to see some interesting changes. I mentioned the evolution of the business analyst taking on more of the responsibility for building processes and applications. IT will focus its time and energy on building a set of transactions and services which can then be consumed into the core business processes. Developers will build processes using BPM technology, and these same processes will be provided to the business analysts so they can extend and reengineer them, as well as create new ones. In the future, IT skills like Java, C++ and C# will still be required, and developer skills and methodologies will still be required. However, one new thing that may be different for many people is service orientation. IT developers need to learn service oriented concepts, and ensure their methodologies include the design and development of service oriented applications, which Gartner calls SODA. From an acquisition perspective, service