Estimating overall project duration for larger projects is time consuming: when following the bottom-up approach of the project planning process, we have to estimate the duration of each work package and then integrate all data into a project schedule, usually, in form of a Gantt chart which would only be available well into planning phase.
On the other hand, important stakeholders – like a customer or major sub-contractors - ask us already in definition phase or early planning phase for our estimated duration. Since the bottom-up approach is not yet finished we start guess work about the project duration. Our stakeholders take it seriously; they acquire our "guesstimation results" as basis of their further preparation and planning. Later in planning phase, we come back to them with our "real result", usually longer durations, they realize that they have to change their plans and might even become angry with us – very understandable.
How can we resolve this dilemma? - Let us look at an example.
We slip into the role of a project manager, Adam, who works for construction company YoNIWeBI (You Name It We Build It). In YoNIWeBI we have a long history of managing all kinds of construction projects, one of our divisions, AB, providing solutions of planning and constructing large apartment buildings. Adam joined division AB 10 years ago and managed apartment building projects ever since.
Division AB very much supports the idea of lessons learned workshops at the end of each project, so that they have a large number of project records available, including information about duration of all kinds of work packages that are similar in scope.
Just two weeks ago, Adam took over a new project which is still in definition phase. So far, it is clear that this new project is about another apartment building with 20 floors and six 3-bedroom apartments on each floor. Besides a meeting with major stakeholders, the customer asks Adam for a quick estimate of the overall project duration, so that they can kick off discussions with their financing partners. They would like to know the probability of finishing the project within a certain period of time.
So, Adam turns to AB's internal data base and finds the following, based upon records of projects of similar scope and volume:
Probability to complete definition, planning and design phase within
Probability to complete implementation and closure phase within
The question is: what is the probability for an overall duration of the project of 18 months or less?
Adam remembers some math from high school and comes up with the idea of applying the method of probability trees. Here is his solution:
There are 6 different ways we can complete the whole project in 18 months or less, with respective probabilities:
By adding up the probabilities we obtain a probability of 48% for overall project duration of 18 months or less. Exactly this is Adam’s answer for the customer.
In case you would like to use practical and useful packages of tools, templates and checklists, here you can get them. They save you a lot of time, are easy to use and easy to change:
All four PM Phases in one Set