Does a fully predictive life cycle ever make sense for a project?

 

One of the misconceptions I like to clear up with learners in the project management fundamentals courses I teach is that there is no such thing as a "waterfall" or "agile" project. Stakeholders might choose to use a predictive or adaptive life cycle or specific methods associated with either of these approaches for delivering the scope of their project, but using these terms as an adjective furthers the erroneous perception that there are only two options available to us. In reality, there are an unlimited number of ways of delivering a project when you consider the wide variety of method, tool, role and life cycle choices available.

But let's consider the purely predictive life cycle. A basic assumption of this life cycle is that when we have completed one stage of delivery there won't be an opportunity to return to that stage again.

But projects are unique endeavors possessing uncertainty. Even operational processes which are in control will experience common and special cause variation.

And as the complexity of a project increases, the level of uncertainty does as well. How many times have you been on any moderately complex project where nothing substantially changed over its lifetime? Faced with change, if we don't provide the opportunity to iterate back, the project is unlikely to meet all of its success criteria. Any project manager that stubbornly refuses to alter their plans to address the new reality they are facing won't be in that role for long.

Even Dr. Winston Royce had provided this caution in his 1970 paper "Managing the Development of Large Software Systems" about a pure waterfall approach without iterations: "I believe in this concept, but the implementation described above is risky and invites failure." As readers of the PMBOK Guide know, the processes in the PMBOK framework are iterative in nature. Finally, the term "waterfall" itself is inaccurate as going over one of those in real life is usually a one way trip!

So the question is not whether or not we will incorporate change into our plans, but rather about the level of effort which we should expend on planning up front. With projects where uncertainty is low, team member experience is high, and we are able to control many sources of variation, we can develop high confidence plans for the entire life of the project whereas with others, a rolling wave approach to planning is wiser with our time horizon for detailed planning being defined by how foggy the road is in front of our headlights.