Microsoft Project, in its various releases (i.e., Online desktop, 2019, 2016, etc.), can become an accurate estimating tool for top-down (Analogous) and bottom-up (Parametric) estimating. This is accomplished by inserting custom columns and simple formulas. However, some tricks are necessary to get around user impediments.
With accurate estimates, a Risk Reserve, an Estimating Uncertainty Buffer, and appropriate updating, a leveled schedule can provide a realistic and feasible current forecast of the delivery date for products and services, as well as a realistic basis for impact analysis of issues and change-requests.
Unfortunately, the large majority of project managers estimate tasks with durations and update tasks by entering %-complete as reported by team members. Perhaps the reasons are that the column, “Durations,” appears when you first launch Microsoft Project, and furthermore, that resources think in terms of what they have accomplished on their deliverable, instead of in terms of a current estimate of work and their average hours on tasks per workday before successor task(s) can start.
I would be remiss if I did not cover a few definitions before continuing.
- The formula “Work = Units * Duration” applies when these three terms are set, and then one of the terms is changed during planning or updating. Work is the productive workhours applied to producing a task’s deliverable, such as a design, prototype, document, etc.
- Duration is the difference in workdays between a tasks’ planned Start and Finish dates.
- Units is the number of resources or fraction of a resource assigned to a task. An example is a resource assigned an average of four workhours per eight hour workday appears as Bob[50%]. “Units” when assigned by the project manager are called “Assignment Units.” When “Units” are calculated by Microsoft Project, the result are referred to as “Peak” units.
To manipulate Work, Units, and Duration on the same screen, view the Gantt Chart with the Work column inserted. On the View ribbon, check Detail. On its dropdown, select “Task Usage” with needed columns added (i.e., Assignment Units, Peak units, task Type, and Effort Driven).
Before making a change to Work, Units, or Duration, identify which of the other two terms you want to be affected. Thus, the third term will be fixed (Fixed Units, Fixed Work, etc.) During planning, “Assignment Units” needs to be fixed or changed to avoid Microsoft Project calculating “Peak” units. For further clarification, see MPUG content on Task Types, so you’re ready for my upcoming webinar.
Using a Schedule Template for Estimating
Now, let’s circle back to the topic at hand. For accurate forecasting, it is better to begin planning using a schedule template, initially calibrated to workhour data derived from a past medium sized project. Copy the template to make a project schedule and scale the tasks according to the size of the current project.
Appropriate updating keeps all progress appearing in the past and all remaining effort appearing in the future. Constantly alerted to product delivery date slippage, you can plan and implement schedule recovery while time allows. Actual work data provides a means to improve estimating data in the schedule template.
It’s not hard to do. Here’s what a piece of what a parametric estimating schedule looks like:
Fig. 1 – Detail tasks factored by the number of average requirements, for example, produce median and high work estimates.
Here’s what an analogous estimating schedule looks like:
Fig. 2 – Summary tasks from the parametric schedule are scaled by analogy, enabling negotiation of project cost, resources, and product delivery date.
Parametric estimating multiplies average effort per item by the number of items a detail task is expected to produce. For example, five work hours per part, times three parts equals 15 workhours for the task. Analogous estimating multiplies aggregate estimating data by a ratio. That is, the summary task effort of 40 workhours * 1.25 = 50 workhours.
These views include estimating uncertainty, which can be set consistently throughout a schedule and is used to justify a project’s Estimating Uncertainty Buffer. That buffer pulls together from tasks the difference between their high and median estimates. Ahead of such buffer, insert a Risk Reserve using a lag that you reduce manually to compensate for delays from known risks that actually occur. Unknown risks are treated as issues as they arise.
I’ll show the operation of these features and the techniques during my on-demand webinar. They will also be included in the associated schedule template available for download. You can copy these features to your schedule or schedule template using File > Organizer.