Life Begins at Requirements (Continued)

Plan for the Macro Lifecycle
Needs and requirements form the start of the application development lifecycle, but there is a higher-level lifecycle, too. This macro level exists at the corporate level and uses applications and infrastructure as the building blocks. While many companies consider their infrastructure and legacy systems to be core business assets, few of them actively plan how new systems will join that select band.

Some applications start life as a quick hack in Excel VBA, only to become business-critical. Often departments hack together their own VBA solutions because they can't get the resources they need from the harried central IT service. Alternatively, they do it deliberately to stay under the radar of the IT folks, who would probably have a fit because there are no backups, no integration, no security, or any of the other concerns of enterprise developers.

At the other end of the scale, projects that start as massive, high-investment megaprojects fade into obscurity as the business moves in a different direction. This is clearly a huge drain on corporate resources and can be avoided by better integration between business and IT planning.

Let us call this macro-level lifecycle the application roadmap. Such a business-driven plan provides a much better perspective of what a company is most likely to need. The idea is to manage systems before they become critical legacy problems and to avoid installing white elephants. This kind of course-grain planning has a scale of years and does not address how systems are built. It looks at how corporate systems need to evolve in line with management expectations. It needs to track what vendors are likely to do in the future, predict market trends, and even track the competition.

An example might show migration of systems off an old mainframe onto newer platform, at the same time as planning for retiring NT 4.0 machines in favor of Windows Server 2003. At an infrastructure level, the roadmap might show a move to Web services over the next five years.

Many managers are uncomfortable with this kind of planning because it is largely speculative and definitely will change. Developing this kind of roadmap is quite different from planning the conventional lifecycle and is beyond the scope of this article. It is, however, important from the prospect of putting the development lifecycle in a business context.

About the Author
Richard has been steadily moving away from code into the heads of business people. He is now CEO of Prosumer Solutions, which is changing the economics of Web app development through Business RAD. He holds a Ph.D. in computer science from the University of Edinburgh. You can contact him at richard.marshall@prosumersolutions.com.