Estimating Activity Durations

---

Ready for a loaded question? 'Now how long will all of this take?' Project managers hear this one all the time, right? And maybe right after that: 'How much will all of this cost?' We'll talk about cost estimates in Chapter 7. For now, let's talk about time.

The answer to the question 'How long will it take?' depends on the accuracy of the estimates, the consistency of the work, and other variables within the project. The best a project manager can do is create honest estimates based on the information he's been provided. Until the schedule is finalized, no one will know the duration of the project.

The tasks are first identified, their duration is estimated, and then the sequencing of the activities takes place. These activities are required to complete the project schedule and the estimated project duration. These three activities are iterated as more information comes available. If the proposed schedule is acceptable, the project can move forward. If the proposed schedule takes too long, the scheduler can use a few strategies to compress the project. We'll discuss the art of sequencing in a few moments.

Activity duration estimates, like the activity list and the WBS, don't come from the project manager-they come from the people completing the work. Activity duration estimates may undergo progressive elaboration. In this section, we'll examine the approach to completing activity duration estimates, the basis of estimates, and allow for activity list updates.

Considering the Activity Duration Estimates Inputs

The importance of accurate estimates is paramount. The activity estimates will be used to create the project schedule, and predict when the project should end. Inaccurate estimates could cost the performing organization thousands of dollars in fines, lost opportunities, loss of customers, or worse. To create accurate estimates, the project manager and the project team will rely on several inputs:

• Activity lists You know this right? Activity lists are the work elements necessary to create the deliverables.

• Constraints An identification of the project constraints is needed since they may influence the estimates. A deadline is an example of a constraint.

• Assumptions An identification of the assumptions is needed since work estimates may be influenced by the assumptions. For example, the team may be operating under an assumption that the project must be completed within one calendar year.

• Resource requirements Activity durations may change based on the number of resources assigned to the activity. For example, Task A may take eight hours with one person assigned to the work, but Task A may be completed in four hours with two team members assigned. Some activities, such as installing a computer operating system, will take the same amount of time regardless of how many resources are assigned. Project managers must also take care not to overload resources in an effort to complete a task; too many resources can be counterproductive.

• Effort vs. duration Effort is the amount of labor that is applied to a task. Duration is how long the task is expected to take with the given amount of labor. For example, a task to unload a freight truck may take eight hours with two people assigned to the task. If the effort is increased by adding more labor to the task (in this instance, more people), then the duration of the task is decreased. Some activities, however, have a fixed duration and are not affected by the amount of labor assigned to the task. For example, to install a piece of software on a computer will take the same amount of time if one computer administrator is completing the work or if two computer administrators are attempting to complete the work.

• Resource capabilities The abilities of the project team members must be taken into consideration. Consider a task in an architectural firm. Reason says that if a senior architect is assigned to the task, he will be able to complete it faster than if a junior architect were assigned to the same task. Material resources are also considered to influence activity time. Consider predrilled cabinets versus cabinets that require the carpenter to drill each cabinet as it is installed. The predrilled cabinets allow the job to be completed faster.

• Historical information Historical information is always an excellent source for information on activity duration estimates. Historical information can come from several sources:

• Historical information can come from project files of other projects within the organization.

• Commercial duration estimating databases can offer information on how long industry-specific activities should take. These databases should take into consideration the materials, the experience of the resources, and define the assumptions the predicted work duration is based upon.

• Project team members may recollect information regarding the expected duration of activities. While these inputs are valuable, they are generally less valuable than documented sources such as other project files or the commercial databases.

• Identified risks We'll discuss risk in detail in Chapter 11. Risks, good or bad, can influence the estimated duration of activities. The risks on each activity should be identified, analyzed, and then predicted as to their probability and impact. If risk mitigation tasks are added to the schedule, the mitigation activities will need their duration estimated and then sequenced into the schedule in the proper order.

Applying Expert Judgment

The project manager and the project team should utilize expert judgment if possible to predict the duration of project activities. Expert judgment can come from subject matter experts, project team members, and other resources, internal or external to the performing organization, that are familiar with the activities the project demands.

Estimating durations is not easy as there are many variables that can influence an activity's duration. Consider the amount of resources that can be applied to the resources, the experience of the resources completing this type of work, and their competence with the work packages.

On the Job 

A big dose of reality is also needed with activity duration estimates. Imagine an activity that has been estimated to take 40 hours. While on paper that looks like a typical workweek, it's pretty unlikely the task will be completed within one week. Why? Consider all the phone calls, impromptu meetings, e-mail, and other interruptions throughout the day. These slivers of time chip away at the actual productive hours within a workday. The project manager should find a base of actual productive hours per day based on typical interruptions, meetings, and so on; for example, six productive hours out of eight working hours is typical. Based on this assumption (that six hours out of a day are productive), this means a task slated to last 40 hours will actually take nearly seven working days to complete.

Creating an Analogy

Analogous estimating relies on historical information to predict what current activity durations should be. Analogous estimating is also known as top-down estimating and is a form of expert judgment. To use analogous estimating, the activities from the historical project are similar in nature and are used to predict what the similar activities in the current project will take.

A project manager must consider if the work has ever been done before, and if so, what help will the historical information provide. The project manager must consider the resources, project team members, and equipment that completed the activities in the previous project compared to the resources available for the current project. Ideally, the activities should be more than similar; they should be identical. And the resources that completed the work in the past should be the same resources used in completing the current work.

When the only source of activity duration estimates is the project team members, instead of expert judgment and historical information, your estimates will be uncertain and inherently risky.

Exam Watch

Analogous estimating uses historical information and is more reliable than predictions from the project team members.

Applying Quantitative Estimates

Quantitatively-based durations use mathematical formulas to predict how long an activity will take based on the 'quantities' of work to be completed. For example, a commercial printer needs to print 100,000 brochures. The workers include two pressman and two bindery experts to fold and package the brochures. Notice how the duration is how long the activity will take to complete, while the effort is the total number of hours (labor) invested because of the resources involved. The decomposed work, with quantitative factors, is shown in Table 6-1.

Table 6-1: Decomposed Work, with Quantitative Factors

Workers	Units per hour	Duration for 100,000	Effort
Pressman (two)	5,000	20 hours	40 hours
Bindery (two)	4,000	25 hours	50 hours
Totals		45 hours	90 hours

Exam Watch

Duration is how long an activity takes, while effort is the billable time for the labor to complete the activity. Consider an activity that is scheduled to last 40 hours. The project manager must consider the cost of the person's time assigned to complete project work-for example, a senior full-time engineer versus a part-time person, at a lower cost. The senior engineer may be able to complete the activity in 40 consecutive work hours, but the cost of this employee's time may be more than the value of the activity. The part-time employee may be able to complete the task in two segments of 20 hours, but their time is billed at a substantially lower rate.

Factoring in Reserve Time

Parkinson's Law states: 'Work expands so as to fill the time available for its completion.' This little nugget of wisdom is oh-so-true. Consider a project team member that knows an activity should last 24 hours. The team member decides, in his own wisdom, to say the activity will last 32 hours. This extra eight hours, he figures, will allow plenty of time for the work to be completed should any unforeseen incidents pop-up. The trouble is, however, that the task will magically expand to require the complete 32 hours. Why does this happen? Consider the following:

• Hidden time Hidden time, the time factored in by the project team member, is secret. No one, especially the project manager, knows why the extra time has been factored into the activity. The team member can then 'enjoy' the extra time to complete the task at leisure.

• Procrastination Most people put off starting a task until the last possible minute. The trouble with bloated, hidden time is people may wait through the additional time they've secretly factored into the activity. Unfortunately, if something does go awry in completing the activity, the work result is later than predicted.

• Demands Project team members may be on multiple projects with multiple demands. The requirement to move from project to project can shift focus, result in loss of concentration, and require additional ramp-up time as workers shift from activity to activity. The demand for multitasking allows project team members to take advantage of hidden time.

• On schedule Activities are typically completed on schedule or late, but rarely early. Users that have bloated the activity duration estimates may finish their task ahead of what they promised, but have a tendency to hold the results until the activity was due. This is because workers aren't usually rewarded for completing work early. In addition, workers don't want to reveal the inaccuracies in their time estimates. Workers may believe future estimates may be based on actual work durations, rather than estimates, so they'll 'sandbag' the results to protect themselves-and finish 'on-schedule.'

So what's a project manager to do? First off, the project manager should strive to incorporate historical information and expert judgment to predicate accurate estimates. Second, the project manager should stress a genuine need for accurate duration estimates. Finally, the project manager can incorporate a reserve time.

A reserve time is a percentage of the project duration or a preset number of work periods and is usually added to the end of the project schedule. Reserve time may also be added to individual activity durations based on risk or uncertainty in the activity duration. When activities are completed late, the additional time for the activity is subtracted from the reserve time. As the project moves forward, the reserve time can be reduced or eliminated as the project manager sees fit. Reserve time decisions should be documented.

Evaluating the Estimates

The end result of estimating activities provides three things:

• Activity duration estimates Activity duration estimates reflect how long each work package will take to complete. Duration estimates should include an acknowledgement of the range of variance. For example, an activity whose duration is expected to be one week may have a range of variance of one week ± three days. This means the work can take up to eight days, or as little as two days. This is assuming a week is five days.

• Basis of estimates Any assumptions made during the activity estimating process should be identified. In addition, any historical information, subject matter experts, or commercial estimating databases that were used should also be documented for future reference.

• Activity list updates During the estimating process, there may be discoveries of missing activities within the activity list. The project manager should confirm that the new work packages are reflected in the activity list for the project.

Inside the Exam

There's a ton of information in this chapter-all of it important-but there are some key things you must know to pass the PMP exam. For starters, you should understand how activity estimates are created.

Analogous estimates use historical information to predict how long current project activities will take place. These estimates are considered top-down estimates and are part of expert judgment. Quantifiable estimates, on the other hand, use a quantity to predict how long activities will take. Consider any unit such as square feet painted per hour or number of units created per day.

GERT is the only network diagram that allows for loops and conditional branching based on what the project has experienced to date. System dynamics is another example of conditional advancement.

When developing the schedule, the most common method is the CPM, though PERT and GERT may also be used. Lag is a positive time added to a task to indicate waiting. Lead is negative time added to a task to 'hurry up.' Fast tracking arranges activities to happen in tandem rather than in succession-this increases risk. Crashing adds more resources to activities to decrease their duration, which typically adds cost.

Monte Carlo Analysis is typically a computer program to estimate the many possible variables within a project schedule. Monte Carlo simulations predict probable end dates, not an exact end date. Another tool the project manager can use is resource leveling. Resource leveling smoothes out the project schedule so resources are not over-allocated. A result of this is that projects are often scheduled to last longer than initial estimates.

The critical path in a project has zero float, and is the path with the longest duration to completion. There can be more than one critical path in a network diagram. Should delays happen on noncritical paths, and all float is consumed, the critical path may change.

The project schedule is a calendar-based system used to predict when the project, and work, will start and end. Gantt charts map activities against a calendar and may show the relationship between activities. Milestone charts show when key deliverables are expected; they do not show the relationship between activities.

---