e Program Evaluation and Review Technique (PERT) is considered
a project management classic. Besides being one of the original sched-
uling techniques,
PERT was the rst signicant project-oriented risk
analysis tool. PERTs objectives included managing schedule risk by
establishing the shortest development schedule, monitoring project
progress, and funding or applying necessary resources to maintain the
schedule. Despite its age (relative to other project risk techniques),
PERT has worn well the test of time.
Technique Description
PERT is based on a set of mathematical equations known as Runge
Kutta. Best- and worst-case scenarios are established and weighted
against the most likely set of occurrences. Task PERT mean and stan-
dard deviations and the project PERT duration and standard devia-
tions are established, which allow the project manager to evaluate the
likelihood of achieving specic schedule targets based on the network
and PERT durations.
When Applicable
PERT is particularly applicable when historical schedule data are
limited. In many projects, information is insucient to ascertain pre-
cisely how long a given task might take; or sometimes team members
are reticent about sharing planned duration for activities they have
never performed. By allowing or encouraging each team member to
provide a best-case duration, a worst-case duration, and a most prob-
able duration for each activity, team members have the opportunity
PERT was originally developed during the Polaris submarine program in the late
risk ManageMent
to share information they might not otherwise have considered (in a
single data-point estimate). Consequently, PERT is normally applied
early in a project when uncertainty is high.
Inputs and Outputs
Inputs for PERT include the multiple duration data points for each
activity and the basic network of activities (see Chapter 22). Gathering
this information may require a signicant level of eort, but it is nor-
mally tracked with the work packages in the project management
software. Most project management software packages incorporate
PERT elds in their databases.
Outputs from PERT are mean durations for the projects critical
path, as well as normal distribution curves to establish the likeli-
hood of meeting various schedule targets. ese outputs are normally
more pessimistic than the duration derived from critical path method
(CPM) analysis because they take into account best and worst cases
(and worst-case scenarios tend to diverge further from the most likely
duration than do best-case scenarios). us, PERT duration reects
more risks inherent in the network and the project as a whole.
Major Steps in Applying the Technique
PERT is applied in two general phases, rst at the task level and again
at the project level.
At the task level, there are three steps that must be conducted for
each task:
Gather the task duration information. As mentioned earlier, this
will consist of establishing best-case, worst-case, and most
likely durations for each task in the network. is informa-
tion is normally extracted from individual team members per-
forming the task.
Calculate the PERT mean and standard deviation for each task. is
is frequently done by using computer tools, although it can
be calculated manually. For the PERT mean, the following
formula is applied:
Optimistic Most Likely Pessimistic +4
To establish the PERT standard deviation, some of the same
information is used:
Pessimistic Optimistic
Catalog the information. Storing the information for easy retrieval
is important because PERT data at the task level have limited
utility. at may be helpful for establishing the basic duration
of a task, but to apply the robust nature of the PERT process,
the entire network must be considered.
At the project level, there are three steps that must be conducted
after PERT information is available for each task:
Establish the PERT critical path. e project manager must cal-
culate the critical path based on PERT durations rather than
the conventional, most probable durations. Because PERT
durations frequently dier from their most likely counter-
parts, there is the distinct possibility that the PERT criti-
cal path will represent a dierent set of activities than the
conventional critical path. e duration of this path therefore
becomes the PERT mean for the project.
Establish the standard deviation for the PERT critical path. is
tends to be one of the most confusing steps in the process
because it involves calculating the square root of the sum of
the squares of the task-developed standard deviations. e
process (again, frequently performed by computers rather
than people) is not as onerous as it may sound. First, square
each of the individual task standard deviations. en add
those squares together. Finally, calculate the square root of
their sum. e formula looks like this:

Get Risk Management, 5th Edition now with O’Reilly online learning.

O’Reilly members experience live online training, plus books, videos, and digital content from 200+ publishers.