Deal With
Estimation, Contingencies, And Uncertainty
Here is a series of
guidelines for developing the schedule and plan:
-
If you cannot estimate a task, what do you do? Break up the
task into smaller parts. Isolate the part you cannot estimate. Ask why you
cannot estimate it. This will surface one or more issues. This is positive
because it gets more issues out on the table earlier. Having identified the
issue behind the task, you can now deal with it and then develop an estimate.
-
Only develop detailed tasks for 3–4 months in the future.
Beyond this, stick to the template level. This will save you more time. For
template level tasks later you can temporarily break up tasks to do estimation.
-
As the plan is updated, the team members will define
additional detailed tasks so that there are detailed tasks for 3–4 months in the
future at all times (see Fig.
5.5).
Figure 5.5: Rolling
Level of Detail in the Project Plan
How do you deal with contingencies? A commonly used approach is to
pad the time for the work. However, if everyone does this, the project will not
finish for years. What is the underlying purpose of contingency planning? To
make management and the team aware of potential problems and impacts. We have
found that a better approach is to define contingency tasks. These are added at
the bottom of the project plan. They are not linked to any tasks in the plan. To
show the impact of the contingency, you merely link the appropriate contingency
tasks into the plan. Panic ensues as managers see the slippage—very good!
What is a major contingency for international projects? The team
members must be drawn off the project for other work. Here is another guideline.
Have each team member create their own plan that indicates what
other work they will doing. You should review this with them and discuss the
possible demands. This is much more precise that just yakking about their other
work in general terms. Later in the project, they can update this and indicate
to you what is going on. This will prevent many unpleasant surprises later.
Moreover, if a team member has a number of real conflicts, you can consider if
this person should be replaced on the team. This is a much better idea at the
start of the project than in the middle in a panic situation.
Now suppose you have developed the plan and have a list of
issues. How do you test if these are both complete? First, go through the list
of tasks and ask if any of these have risk. If they do, then do you have the
associated issues identified? If not, then you can add more issues and link
these to the tasks. Second, go down the list of issues. Ask if these apply to
the international project. If one does, then go down the list of tasks to see if
the appropriate tasks are present. If not, you can add tasks. The end result is
that you have verified both the issues and the tasks. Another step to take is to
review the milestones and label the ones that involve major issues. These issues
are related to the tasks that produce or lead to the end product.