Has your project schedule shifted? Projects are often judged by how closely they adhere to their baseline schedule. Unfortunately, many projects often take much longer to complete than initially projected. One of the reasons that projects aren't completed on time is that risk is not adequately built into baseline schedules during the planning phase. Typically, teams do not adequately assess the risks to a project and present overly optimistic timelines to their management.

Timelines are often built by having team members estimate how long their tasks will take. The team members usually give the project manager a duration based on the average duration for that type of task or a duration that they are 80% confident they will achieve. They typically don't factor in the amount of time that a task could take due to risks being realized during the project. The following example will demonstrate what can happen when a team estimates the 80% probable task time for several tasks and then assembles them into an integrated timeline:

Figure 1:
Example Timeline:

Task A(5 days) -> Task B(5 days) -> Task C(5 days) -> Task D(5 days)

The total time for the project in Figure 1 should be 20 days if each of the four above tasks finishes on time. Each task was projected to be complete in 5 days 80% of the time. Let's assume that the other 20% of the time, each task will take 7 days instead of the 5 days that was initially projected, due to a risk being realized. Extend^TM is a simulation package that can be used to demonstrate how variability can impact the time it takes to complete a series of tasks. A computer simulation of the project in Extend^TM demonstrates the possible outcomes when this 4 task project is run 1000 times.

The output from the model shows that the average time to complete the project is 22.9 days and the worst case scenario is 28 days. When risk is modeled into this simple timeline, it shows that the project is likely to take several days longer to complete than planned. This could translate to several months when risk is factored into a longer or more complex project.

How can risk be incorporated into project plans? One step that should be taken in project planning is to use three-point estimates to determine task durations instead of a single point estimate. Having team members provide an optimistic, most-likely, and pessimistic duration for each task will give the project manager a much better idea of how much timeline risk is involved with each task. This information can be incorporated into the project timeline by fitting distributions to tasks with uncertainty using a software package that is capable of a Monte Carlo simulation. One software package that is capable of doing this in Microsoft Project^TM is @risk^TM. An example of this has been created using @risk^TM, assuming the following:

For each of the four tasks in Figure 2 below, assume that the optimistic duration is 4 days, the most-likely duration is 5 days, and the pessimistic duration is 7 days.

Figure 2:
Timeline with Uncertainty

Task A(4-7 days) -> Task B(4-7 days) -> Task C(4-7 days) -> Task D(4-7 days)

After the time estimates for the steps is gathered, it can be incorporated into an actual project timeline by fitting probability distributions to tasks with uncertainty using @risk software. After this step has been completed, these tasks can be linked together and a simulation can be run using @risk^TM software. See Table 1 below for results.

Table 1:
@Risk^TM Model Results

Summary Statistics

Statistic	Value	%tile	Value
Minimum	17.780	5%	19.300
Maximum	24.940	10%	19.740
Mean	21.364	15%	20.090
Std Dev	1.246647244	20%	20.360
Variance	1.554129352	25%	20.540
Skewness	0.036933085	30%	20.710
Kurtosis	2.999679111	35%	20.900
Median	21.330	40%	21.070
Mode	21.120	45%	21.200
Left X	19.300	50%	21.330
Left P	5%	55%	21.480
Right X	23.500	60%	21.650
Right P	95%	65%	21.850
Diff X	4.200000763	70%	22.020
Diff P	90%	75%	22.180
# Errors	0	80%	22.390
Filter Min	85%	85%	22.650
Filter Max	90%	90%	22.890
# Filtered	0	95%	23.500

In this example, the 80% probable date is 22.39 days, which is over 10% longer than the 20 days that would be expected without accounting for uncertainty.

This approach was applied successfully to a drug development project. The project team started with the timeline shown in Table 2 below for completing a clinical study and the subsequent clinical study report:

Table 2:
Standard Timeline

Task	Duration	Start Date	Finish Date
Enrollment	600 days	1/28/2009	5/17/2011
Treatment	80 weeks	5/17/2011	11/27/2012
Clinical Study Report	158 days	11/28/2012	7/5/2013

The team put in their best estimates for when the three tasks would be completed and found that the project would finish on 7/5/2013. The team felt that there could be some variability in the time it would take to complete the enrollment and clinical study report tasks and decided to provide three-point estimates for these tasks. The three-point estimates are shown in table 3 below:

Table 3:
New Estimates for Task Durations

Task	Optimistic Duration	Most Likely Duration	Pessimistic Duration
Enrollment	400 days	600 days	900 days
Clinical Study Report	148 days	158 days	198 days

Note that the team believed that the durations they originally reported were the most likely, but they acknowledged that there were risks to completing these tasks on time. After entering this information into @risk^TM and running the simulation, the team found that the original completion date of 7/5/2013 was only 35% likely. The true 80% date for completion of the study report was 1/8/2014, which was six months later than the original projection. This prompted the team to consider the risks involved in the project, reduce the risks where it was possible, and commit to a more realistic date for completion of the clinical study report.

The common practice is that a project team would commit to a timeline without factoring in this uncertainty. However, using the approach and the tools described above, the project manager can factor in risk appropriately and evaluate what the true 80% probable time is to complete a project. This will allow the project manager to communicate more accurate project schedules to upper management. Ultimately, this can lead to more projects being completed on time and a much higher level of satisfaction for all of the project stakeholders.