Several months back I wrote a post about project management in which I shared words of wisdom (more like frustration) from Department of Veteran Affairs CIO Roger Baker, who told Federal News Radio "it turns out to be much harder to stop an ongoing project than to start a new one." The article I cited went on to describe a vetting process called the Program Management Accountability System (PMAS), which Baker employed to kill a dozen of the VA's IT projects crippled by poor software quality, cost overruns and other problems.
The post also offered a list of common project-management problems, among them not enough courage to kill projects with poor prospects for success. An item that isn't on the list, but should be, is lack of an exit strategy. Mustering the courage to kill a project won't do you much good if you don't have a strategy for doing so.
Gartner CIO Darko Hrelic addresses this topic in a recent Forbes interview. He said that while it's logical to have an exit strategy as well as an entry strategy for any project, exit strategies aren't common in IT. He said:
We think about that in anything else in life. When do you go to the top of the ski slope without any idea of where we're going to stop and how? Or in the car and "How are we going to stop this car?" We always think about both ends of it. Yet in IT, it's nonexistent. We don't say, "What if this doesn't work?"
In addition to having an exit strategy, Hrelic stresses the importance of ensuring IT projects satisfy a business need and making sure they don't end up like the Charlie-in-the-Box on the Island of Misfit Toys -- irrelevant and forgotten by users. Lack of application management is a lingering problem for many organizations. Applications are brought in to solve pressing issues, but then aren't phased out or combined with other applications after they've served their purpose.
The result, said Hrelic:
And a whole bunch of resources and money is held back on trying to keep the old stuff going. And so, as a result, you either fund IT more, which is typically not the option, or the capability you provide is not as great, which is a problem.