During the phase of planning, you created your program of task, defined the relations of task and the dependences, and assigned all the essential resources. Hopefully you have set most or all
tasks to Auto-Scheduled mode.
Many organizations require that you capture a
baseline of your schedule at the end of the planning phase. You should
capture a baseline of the entire project after you are satisfied that
your schedule is a realistic structure for work activities. The
baseline enables you to later assess the performance of the project
compared to the plan as you execute the details of the project. Make sure you check your organization’s policies to determine when a baseline capture is required.
Looking for Logic Errors
Logic errors in a project occur when, for example,
you forget a task or summary task while setting up your project, or
schedule your tasks in the wrong order. It can be difficult to judge
your own schedule as the errors might be harder to notice because of
how familiar you are with it.
However, you can do some initial logic checks by answering the following questions:
Is all required work included in the schedule?
To answer this question, review your project scope statement and your
Work Breakdown Structure (WBS) to ensure that your schedule captures
the full scope of your project and all work has been included.
Does the organization of the schedule achieve the project’s Measure of Success?
When starting up any project, you define the Measure of Success, which
is your metric for what it really means for your project to have
successfully achieved everything you planned for it to achieve. The
tasks and milestones you have defined in your schedule must represent
the Measure of Success you have defined.
Does the schedule include all and only necessary deliverables and tasks that achieve the work in the WBS?
It is important to recognize all of the project deliverables that
support the Measure of Success you have defined. It is also important
to remember that a successful project schedule should not include
extraneous deliverables that might seem like they are needed, but are
really not part of your project scope. For example, you have a house
remodeling project in which you set out to replace the floors
throughout the entire house. The tasks to purchase the materials, to
hire the contractors, and to replace the flooring in each room of the
house are all part of your scope. But the natural human thinking may
lead you to consider that while you are replacing the floors, you might
as well replace the doors on the interior of the house, even though it
is not part of the original plan. Following this way of thinking, you
can end up remodeling your entire house, even though your initial
intent was to only replace the flooring. This is called scope creep
and can be a dangerous scenario for any project, because scope creep is
often responsible for projects finishing over budget and behind
schedule. Even though the example described here may be an exaggerated
one, situations similar to this occur and can lead to serious problems.
Are the tasks scheduled to occur in the correct order?
The order of tasks can be as essential as the tasks themselves. It is
important to remember that not only should tasks occur in chronological
order, but you should also allow for enough slack in tasks with the
most risk. For example, adopting a new technology can be a risky task
because you may encounter unforeseen problems, so scheduling those
tasks earlier can be beneficial and can help you avoid delays later in
the project. In addition, you must ensure that each task has all of its
prerequisites met. For example, you would not schedule movers to come
in and load the moving truck before you have packed your belongings.
Similar logic should apply to all of the tasks in your project.
A good practice for determining logic errors is to
have a trusted associate or your project team look at the project for
you. Make sure the reviewer understands that he or she is just looking
for your logic errors by reviewing your scope, tasks, task arrangement,
and planning logic. It is possible your associate
will want to change your schedule because he or she might have done
things differently in another project. Everyone has his or her own way
of doing things, and just like everything else in Microsoft Project,
there is more than one way to achieve the same result. In this case,
ask the people reviewing your schedule to concentrate on your logic
without focusing on trying to improve your project using his or her own
standards. However, be open to the criticism, too; that is one of the
best ways to learn.
Schedule Estimation Methods
Schedule estimation errors are the second type of
errors you need to check against when reviewing your project schedule.
Estimation errors are caused by mistakes in setting your task
durations, work, and budget. There are several ways to help you resolve
estimation errors.
In order to discover estimation errors, you need to ask yourself the following questions:
Are tasks, work, and duration estimates viable?
Make sure that your work and duration estimates are realistic for all
tasks. Many people underestimate, which leads to additional work, time,
and in turn budget increases. Similarly, overestimating the work and
duration is also undesirable because you might not be able to meet your
project requirements. Ideally, your estimates should attempt to reflect
the reality as closely as possible.
Are budget estimates viable?
Budget estimates are also an important aspect of project planning.
Ensuring that the budget you have allocated for your project is met and
not exceeded can be the deciding point between success and failure.
Ensure that the initial budget estimates you have are realistic and
meet your project requirements.
In order to better understand estimation errors, it
is important to review the three main estimation methods—analogous,
expert judgment, and parametric duration.
Analogous Estimation Method
Analogous duration estimates simply ask the
question, “What was the duration the last time we did this?” If you
have been following the project management best practice of recording
lessons learned at the end of each project, you have proof, which
increases in accuracy with every project you complete. You can use the
past performance as an estimation comparison if your previous project
is similar to your current scope.
The advantage to analogous estimates is that they
focus on system-level activities, such as integration, documentation,
configuration management, and others. They also require minimal project
detail and are usually faster and easier to develop; plus your
information is readily accessible in previous Microsoft Project
schedules.
The downside to analogous duration estimates is that
if you do not have any previous information, you have no detailed basis
for justifying your estimates. Also, because it focuses on the system
level, it can be hard to identify lower-level problems that can raise
costs.
Expert Judgment Estimate Method (Using PERT)
Expert
judgment is a great tool to use when hard data is scarce. Former
versions of Microsoft Project included a Program Evaluation and Review
Technique (PERT) analysis tool. PERT uses a network diagram analysis
technique to create three estimates: Optimistic, Expected, and
Pessimistic. These three estimates are then combined to estimate an
activity’s duration, using a weighted statistical average of the three
values. The statistical technique considers: 1 part Optimistic, 4 parts
Expected, and 1 part Pessimistic, all divided by 6 parts. This method
yields a weighted average duration for a given task.
Unfortunately, with Project 2010, PERT analysis is
supported only with a combination of custom fields and custom code or
with third-party products.
Parametric Duration Estimation Method
Finally, the parametric duration estimates are
simply mathematical models used when estimating task durations. If you
had a machine that produced a product at a rate of 10 units per hour
and you needed to make 200 units, you are looking at 20 hours. This
process produces exact estimates, but your estimate can change if you
add another machine, increase the productivity, or decrease the
duration of the task. Many different industries already have standard
calculations, devised from years of experience, that can be used for
this type of estimating.
