The Standish Group and others publish scary statistics showing very high failure rates for IT projects – with these at considerable variance with my 50+ years of IT experience. The problem seems to be the definition of “failure” and the definition of “project” with both being far too broad by the scary statistic folks. For a real project, that is one with a careful definition, an engaged sponsor, and a plan and a budget … along with “success” being defined as “the customer liked what he/she got” … my experience suggests a success rate of 60% or above. For some CIOs, I’d guesstimate a success rate of 85% or higher.
If you define as a “project” something without a plan, budget,or sponsor, you will have a Self-Fulfilling-Prophesy of frequent failure. So, my advise to those who read those scary statistics is … to avoid following in those footsteps, don’t step into them at the get-go. Instead, work your tail off at the start … supper careful definition and scope that everyone understands and agrees to, an engaged and highly visible sponsor, a realistically achievable plan and budget, a user+IT team comprised of the best darn folks you can get your hands on. Given these, you will succeed; without these you will fail … and just add to those scary statistics.
Yes, I’m cheating. I’ve defined a project with perimeters that insure success. Is that wrong or unfair? Aren’t we all supposed to succeed when we start something? Including as “projects” things that are doomed from the get-go just seems dumb …