Chapter 3 Managing Systems Projects

Chapter
3 Managing
Systems Projects

Chapter 3 is the final chapter in the systems planning The chapter includes four “Case in Point” discussion
phase of the SDLC. This chapter describes project questions to help contextualize the concepts described
management and explains how to plan, schedule, in the text. The “Question of Ethics” considers the
monitor, and report on IT projects. implications of raising awareness of a project going astray
even when the project manager is reluctant to do so.

Learning Objectives Chapter Contents

When you finish this chapter, you should be 3.1 Introduction
able to: 3.2 Overview of Project Management
• Explain project planning, scheduling, monitoring, 3.3 Creating a Work Breakdown Structure
and reporting Case in Point 3.1: Parallel Services
Case in Point 3.2: Sunrise Software
• Draw a project triangle that shows the 3.4 Identifying Task Patterns
relationship among project cost, scope, and time 3.5 Calculating the Critical Path
• Create a work breakdown structure, identify task 3.6 Project Monitoring and Control
patterns, and calculate a critical path 3.7 Reporting
• Explain techniques for estimating task completion 3.8 Project Management Examples
times and costs 3.9 Project Management Software
• Describe various scheduling tools, including Gantt Case in Point 3.3: Census 2010
charts and PERT/CPM charts 3.10 Risk Management
3.11 Managing for Success
• Analyze task dependencies, durations, start dates,
Case in Point 3.4: Spring Forward Products
and end dates
3.12 The Bottom Line
• Describe project management software and how it A Question of Ethics
can be of assistance 3.13 Chapter Summary
• Control and manage project changes as they Key Terms
occur Chapter Exercises
• Discuss the importance of managing project risks
• Understand why projects sometimes fail
 Phase 1  Systems Planning

3.2  Overview of Project Management 71­

3.1  Introduction
Many professionals manage business and personal projects every day, but do not
always give it much thought. To manage a large-scale IT project, specific tools and
techniques are needed. A project manager is also needed, someone who is responsible
for overseeing all relevant tasks. No matter which tools are used, the idea is to break
the project down into individual tasks, determine the order in which the tasks need to
be performed, and figure out how long each task will take. With this information,
Gantt charts or PERT/CPM charts can be used to schedule and manage the work.
Microsoft Project is a popular project management tool that can help create and then
monitor the project plan, report progress, and use risk management to make the
whole process easier for everyone.

3.2 Overview of Project Management

The management process for developing an information system or working on a con-
struction project is much the same. The only difference is the nature of the project.
Project management for IT professionals includes planning, scheduling, monitoring
and controlling, and reporting on information system development.

3.2.1 What Shapes a Project?

A successful project must be completed on time, within budget, and deliver

T
a quality product that satisfies users and meets requirements. Project manage-

S
CH
FA
ment techniques can be used throughout the SDLC. System developers can

EA
initiate a formal project as early as the preliminary investigation stage, or

P
later on, as analysis, design, and implementation activities occur. GOOD
Systems development projects tend to be dynamic and challenging. There
is always a balance between constraints, which was discussed in Chapter 2, PICK ANY TWO
and interactive elements such as project cost, scope, and time.
Figure 3-1  You can’t get
everything you want; you have
3.2.2 What Is a Project Triangle? to make choices.
Figure 3-1 shows a very simple example of a project triangle. For each
project, it must be decided what is most important, because the work
cannot be good and fast and cheap.
When it comes to project management, things are not quite so
simple. Decisions do not need to be all-or-nothing, but recognize
that any change in one leg of the triangle will affect the other
two legs. Figure 3-2 represents a common view of a project
triangle, where the three legs are cost, scope, and time. The
challenge is to find the optimal balance among these factors.
st

Sc
Co

Most successful project managers rely on personal experience,

op
e

communication ability, and resourcefulness. For example, if an

extremely time-critical project starts to slip, the project manager
might have to trim some features, seek approval for a budget
increase, add new personnel, or a combination of all three
actions.
On its website, Microsoft offers an interesting suggestion for Time
project managers who have a project at risk: Find the “stuck Figure 3-2  A typical project triangle includes
side” of the triangle. Microsoft states that in most projects, at cost, scope, and time.
Chapter 3 Managing Systems Projects
72­ 3.3  Creating a Work Breakdown Structure

least one side of the triangle is fixed and unlikely to change. It might be a budget cast
in stone, a scope that is inflexible, or a schedule driven by factors beyond the firm’s
control. Whichever side is fixed is probably critical to the project’s success. The leg
where the problem resides must also be identified: cost, scope, or time. Microsoft sug-
gests that if the problem is in the fixed leg, work on the other two legs. For example,
if the project must not exceed the budget and it is starting to run over, adjust the
schedule, or the scope, or both. However, if the problem is not related to the fixed leg,
the adjustment might have to be in the remaining leg. So, when faced with an inflexi-
ble budget (fixed leg) and the schedule is slipping (problem leg), the project’s scope
(remaining leg) might have to be adjusted. Explaining this situation to management is
sometimes the most difficult task of all.

3.2.3 What Does a Project Manager Do?

Whether a project involves a new office building or an information system, good
leadership is essential. In a systems project, the project manager, or project leader,
usually is a senior systems analyst or an IT department manager if the project is large.
An analyst or a programmer/analyst might manage smaller projects. In addition to
the project manager, most large projects have a project coordinator. A project coordi-
nator handles administrative responsibilities for the team and negotiates with users
who might have conflicting requirements or want changes that would require addi-
tional time or expense.
Project managers typically perform four activities, or functions: planning, schedul-
ing, monitoring, and reporting:
• Project planning includes identifying all project tasks and estimating the com-
pletion time and cost of each.
• Project scheduling involves the creation of a specific timetable, usually in the
form of charts that show tasks, task dependencies, and critical tasks that might
delay the project. Scheduling also involves selecting and staffing the project
team and assigning specific tasks to team members. Project scheduling uses
Gantt charts and PERT/CPM charts, which are explained in the following
sections.
• Project monitoring requires guiding, supervising, and coordinating the project
team’s workload. The project manager must monitor the progress, evaluate the
results, and take corrective action when necessary to control the project and
stay on target.
• Project reporting includes regular progress reports to management, users, and
the project team itself. Effective reporting requires strong communication skills
and a sense of what others want and need to know about the project.
The following sections describe the project planning and scheduling steps: how to
create a work breakdown structure, identify task patterns, and calculate the project’s
critical path.

3.3 Creating a Work Breakdown Structure

A work breakdown structure (WBS) involves breaking a project down into a series of
smaller tasks. Before creating work breakdown structures, the two primary chart
types should be understood: Gantt charts and PERT/CPM charts.
 Phase 1  Systems Planning

3.3  Creating a Work Breakdown Structure 73­

3.3.1 Gantt Charts
Henry L. Gantt, a mechanical engineer and management consultant, developed Gantt
charts almost 100 years ago. His goal was to design a chart that could show planned
and actual progress on a project. A Gantt chart is a horizontal bar chart that repre-
sents a set of tasks. For example, the Gantt chart in Figure 3-3 displays five tasks in a
vertical array, with time shown on the horizontal axis. The position of the bar shows
the planned starting and ending time of each task, and the length of the bar indicates
its duration. On the horizontal axis, time can be shown as elapsed time from a fixed
starting point, or as actual calendar dates. A Gantt chart also can simplify a complex
project by combining several activities into a task group that contains subsidiary
tasks. This allows a complex project to be viewed as a set of integrated modules.
A Gantt chart can show task status by adding a contrasting color to the horizontal
bars. For example, a vertical red arrow marks the current date in Figure 3-3. With a
fixed reference point, it is easy to see that Task 1 is way behind schedule, Task 2 is
only about 80% done and is running behind schedule, Task 3 should have started, but
no work has been done, Task 4 actually is running ahead of schedule, and Task 5 will
begin in several weeks.

Figure 3-3  In this Gantt chart, notice the yellow bars that show the percentage of task completion.

Gantt charts can present an overview of the project’s status, but they do not pro-
vide enough detailed information, which is necessary when managing a complex proj-
ect. Some project managers may find that PERT/CPM charts, which are discussed in
the following section, are better tools for managing large projects.

3.3.2 PERT/CPM Charts
The Program Evaluation Review Technique (PERT) was developed by the U.S. Navy
to manage very complex projects, such as the construction of nuclear submarines. At
approximately the same time, the Critical Path Method (CPM) was developed by pri-
vate industry to meet similar project management needs. The distinction between the
two methods has disappeared over time, and today the technique is called either
PERT, CPM, or PERT/CPM. This textbook will use the term PERT chart.
PERT is a bottom-up technique because it analyzes a large, complex project as a series
of individual tasks, just as a pyramid is built from the bottom up using individual blocks.
To create a PERT chart, first identify all the project tasks and estimate how much time
each task will take to perform. Next, determine the logical order in which the tasks must
Chapter 3 Managing Systems Projects
74­ 3.3  Creating a Work Breakdown Structure

be performed. For example, some tasks

cannot start until other tasks have been
completed. In other situations, several
tasks can be performed at the same time.
Once the tasks are known, their dura-
tions, and the order in which they must be
performed, calculate the time that it will
take to complete the project. The specific
tasks that will be critical to the project’s
on-time completion can also be identified.
An example of a PERT chart, which
Microsoft calls a network diagram, is
shown in the lower screen in Figure 3-4.
Although a Gantt chart offers a valu-
able snapshot view of the project, PERT
charts are more useful for scheduling,
monitoring, and controlling the actual
work. With a PERT chart, a project
Figure 3-4  The top screen shows a Gantt chart with six tasks. The PERT manager can convert task start and fin-
chart in the bottom screen displays an easy-to-follow task pattern for the same ish times to actual dates by laying out
project. When the user mouses over the summary box for Task 5, the details
become visible. the entire project on a calendar. Then,
on any given day, the manager can com-
pare what should be happening with
what is taking place, and react accordingly. Also, a PERT chart displays complex task
patterns and relationships. This information is valuable to a manager who is trying to
address high priority issues. PERT and Gantt charts are not mutually exclusive tech-
niques, and project managers often use both methods.
Figure 3-4 shows both chart types. The top screen is a Gantt chart with six tasks. The
PERT chart below it shows the same project, using a separate box for each task instead
of a horizontal bar. Although they both show the task patterns and flow, the PERT chart
boxes can provide more information, such as task duration, start date, finish date, and
the names of resources assigned to the task. The PERT chart in Figure 3-4 would be too
small to view the actual details, which are shown in the expanded text box at the bottom
of the figure. How to create PERT charts is explained in a later section.

3.3.3 Identifying Tasks in a Work Breakdown Structure

A work breakdown structure must clearly identify each task and include an estimated
duration. A task, or activity, is any work that has a beginning and an end and requires
the use of company resources such as people, time, or money. Examples of tasks include
conducting interviews, designing a report, selecting software, waiting for the delivery of
equipment, or training users. Tasks are basic units of work that the project manager
plans, schedules, and monitors—so they should be relatively small and manageable.
In addition to tasks, every project has events, or milestones. An event, or milestone,
is a recognizable reference point that can be used to monitor progress. For example, an
event might be the start of user training, the conversion of system data, or the comple-
tion of interviews. A milestone such as “complete 50% of program testing” would not
be useful information unless it could be determined exactly when that event will occur.
Figure 3-5 shows tasks and events that might be involved in the creation, distribu-
tion, and tabulation of a questionnaire. Notice that the beginning and end of each
task are marked by a recognizable event. It would be virtually impossible to manage a
project as one large task. Instead, the project is broken down into smaller tasks,
c­reating a WBS. The first step in creating a WBS is to list all the tasks.
 Phase 1  Systems Planning

3.3  Creating a Work Breakdown Structure 75­

Figure 3-5  Using a questionnaire requires a series of tasks and events to track the progress.
The illustration shows the relationship between the tasks and the events, or milestones that mark the
beginning and end of each task.

Listing the Tasks:  While this step sounds simple, it can be challenging because the
tasks might be embedded in a document, such as the one shown in the first version of
Figure 3-6. One trick is to start by highlighting the individual tasks, as shown in the
second version. Adding bullets makes the tasks stand out more clearly, as shown in the
third version. The next step is to number the tasks and create
a table, similar to the one shown in Figure 3-7, with columns
for task number, description, duration, and predecessor tasks, First version
which must be completed before another task can start. First, reserve the meeting room. Then order
the marketing materials and brief the
managers. After the briefings, send out
Estimating Task Duration:  Task duration can be customer emails and burn sample DVDs.
hours, days, or weeks—depending on the project. Because the When the emails are sent and the DVDs are
following example uses days, the units of measurement are ready, load the new software. When the
marketing materials have arrived and the
called person-days. A person-day represents the work that one
software is ready, do a dress rehearsal.
person can complete in one day. This approach, however, can
present some problems. For example, if it will take one person Second version
20 days to p­erform a particular task, it might not be true that
First, reserve the meeting room. Then order
two people could complete the same task in 10 days or that the marketing materials and brief the
10 people could perform the task in two days. Some tasks can managers. After the briefings, send out
be divided evenly so it is possible to use different combinations customer emails and burn sample DVDs.
When the emails are sent and the DVDs are
of time and p­eople—up to a point—but not all. In most systems ready, load the new software. When the
analysis tasks, time and people are not interchangeable. If one marketing materials have arrived and the
analyst needs two hours to interview a user, two analysts also software is ready, do a dress rehearsal.
will need two hours to do the same interview.
Third version
Project managers often use a weighted formula for estimat-
ing the duration of each task. The project manager first makes • First, reserve the meeting room.
three time estimates for each task: an optimistic, or best-case • Then order the marketing materials and brief
the managers.
estimate (B), a probable-case estimate (P), and a pessimistic, or
• After the briefings, send out customer emails
worst-case estimate (W). The manager then assigns a weight, and burn sample DVDs.
which is an importance value, to each estimate. The weight can • When the emails are sent and the DVDs are
vary, but a common approach is to use a ratio of B = 1, P = 4, ready, load the new software.
and W = 1. The expected task duration is calculated as • When the marketing materials have arrived
follows: and the software is ready, do a dress rehearsal.

Figure 3-6  The three versions show how to

 (B + 4P + W)  transform a task statement into a list of specific
6 tasks for a work breakdown structure.
Chapter 3 Managing Systems Projects
76­ 3.3  Creating a Work Breakdown Structure

Task Description Duration Predecessor

No. (Days) Tasks
1 Reserve the meeting room
2 Order the marketing materials
3 Brief the managers
4 Send out customer emails
5 Burn sample DVDs
6 Load the new software
7 Do a dress rehearsal
Figure 3-7  In this table, columns have been added for task number, description,
duration, and predecessor tasks, which must be completed before another task can start.

For example, a project manager might estimate that a file-conversion task could be
completed in as few as 20 days or could take as many as 34 days, but most likely will
require 24 days. Using the formula, the expected task duration is 25 days, calculated
as follows:
(20 + (4*24) + 34)
= 25
6

Case in Point 3.1:  Parallel Services

The project management team at Parallel Services is having a debate about how to define
tasks in the work breakdown structure (WBS). Ann, the project manager, wants to break
tasks down into the smallest possible units. For example, she objected to a broad task
statement called Develop a training schedule. Instead, she suggested three subtasks:
(1) Determine availability of training room, (2) Determine availability of attendees, and
(3) Select specific dates and training times.
Karen, another project team member, disagrees. She feels that the broader task state-
ment is better because it allows more flexibility and will produce the same result. Karen
says that if you break tasks into pieces that are too small, you risk over-managing the work
and spending more time on monitoring than actually performing the tasks. As a member of
the team, would you tend to agree more with Ann or Karen? What are the pros and cons
of each approach?

3.3.4  Factors Affecting Duration

When developing duration estimates, project managers consider four factors:
1. Project size
2. Human resources
3. Experience with similar projects
4. Constraints

Project Size:  As described in Chapter 1, information systems have various charac-

teristics that affect their complexity and cost. In addition to considering those factors,
a project manager must estimate the time required to complete each project phase.
 Phase 1 Systems Planning

3.3  Creating a Work Breakdown Structure 77­

To develop accurate estimates, a project manager must identify all project tasks, from
initial fact-finding to system implementation. Regardless of the systems development
methodology used, the project manager must determine how much time will be needed
to perform each task. In developing an estimate, the project manager must allow time
for meetings, project reviews, training, and any other factors (e.g., scheduled vacations
or unscheduled medical leave) that could affect the productivity of the development
team.

Human Resources:  Companies must invest heavily in cutting-edge technology

to remain competitive in a connected world. In many areas, skilled IT professionals
are in great demand, and firms must work hard to attract and retain the talent they
need. A project manager must assemble and guide a development team that has the
skill and experience to handle the project. If necessary, additional systems analysts or
programmers must be hired or trained, and this must be accomplished within a spe-
cific time frame. After a project gets under way, the project manager must deal with
turnover, job vacancies, and escalating salaries in the technology sector—all of which
can affect whether the project can be completed on time and within budget. The proj-
ect manager also has to accommodate official holidays, family emergencies, and other
events that may affect the schedule.

Experience with Similar Projects:  A project manager can develop time and
cost estimates based on the resources used for similar, previously developed informa-
tion systems. The experience method works best for small- or medium-sized projects
where the two systems are similar in size, basic content, and operating environment.
In large systems with more variables, the estimates are less reliable.

Constraints:  Chapter 2 explained that constraints are defined during the pre-
liminary investigation. A constraint is a condition, restriction, or requirement that
the system must satisfy. For example, a constraint might involve maximums for one
or more resources, such as time, dollars, or people. A project manager must define
system requirements that can be achieved realistically within the required constraints.
In the absence of constraints, the project manager simply calculates the resources
needed. However, if constraints are present, the project manager must adjust other
resources or change the scope of the project. This approach is similar to the what-if
analysis described in Chapter 12.

Case in Point 3.2:  Sunrise Software

A lively discussion is under way at Sunrise Software, where you are a project manager.The
main question is whether the person-days concept has limitations. In other words, if a task will
require 100 person-days, does it matter whether two people in 50 days, five people in 20 days,
10 people in 10 days, or some other combination that adds up to 100 performs the work?
Programmers Paula and Ethan seem to think it does not matter. On the other hand,
Hector, a systems analyst, says it is ridiculous to think that any combination would work. To
support his point, he offers this extreme example: Could 100 people accomplish a task esti-
mated at 100 person-days in one day?
Is Hector correct? If so, what are the limits in the “people versus days” equation? Taking
the concept a step further, is there an optimum number of people to be assigned to a task?
If so, how would that number be determined? You need to offer some guidance at the next
project team meeting. What will you say?
Chapter 3 Managing Systems Projects
78­ 3.4  Identifying Task Patterns

3.3.5  Displaying the Work Breakdown Structure

After the task durations are entered, the work breakdown structure will look like
Figure 3-8. Task groups can be used to manage a complex project with many tasks, just as
with a Gantt chart, to simplify the list. Note that the WBS shown in Figure 3-8 is still

Task Description Duration Predecessor

No. (Days) Tasks
1 Reserve the meeting room 1
2 Order the marketing materials 9
3 Brief the managers 2
4 Send out customer emails 3
5 Burn sample DVDs 3
6 Load the new software 2
7 Do a dress rehearsal 1

Figure 3-8  Task durations have been added, and the WBS is complete except for
predecessor task information. The predecessor tasks will determine task patterns and
sequence of performance.

incomplete: It does not show fields such as Start Date, End Date, Task Name, Duration,
and Predecessors—fields that can be key for project managers. With Microsoft Project,
the WBS (including some of these missing fields) might resemble Figure 3-9.

Figure 3-9  This Microsoft Project screen displays the same WBS, including task number, task name,
duration, and predecessor tasks.

3.4  Identifying Task Patterns

Tasks in a work breakdown structure must be arranged in a logical sequence called a
task pattern. This section explains how to understand and create graphical models of
these patterns.

3.4.1 Task Patterns
In any project, large or small, tasks depend on each other and must be performed in a
sequence, not unlike the commands in a software program. Task patterns can involve
dependent tasks, multiple successor tasks, and multiple predecessor tasks. In larger
 Phase 1 Systems Planning

3.4  Identifying Task Patterns 79­

projects, these patterns can be very complex, and an analyst must study TASK BOX FORMAT
the logical flow carefully.

Task Name
3.4.2 Using Task Boxes to Create a Model
In a PERT/CPM chart, project tasks are shown as rectangular boxes,
arranged in the sequence in which they must be performed. Each Start Day/Date Task ID
re­ctangular box, called a task box, has five sections, as shown in Finish Day/Date Task Duration
Figure 3-10. Each section of the task box contains important inf­
ormation about the task, including the Task Name, Task ID, Task Figure 3-10  Each section of the task
Duration, Start Day/Date, and Finish Day/Date. box contains important information
about the task, including the Task Name,
Task Name:  The task name should be brief and descriptive, but it Task ID, Task Duration, Start Day/Date,
and Finish Day/Date.
does not have to be unique in the project. For example, a task named
Conduct Interviews might occur in several phases of the project.

Task Id:  The task ID can be a number or code that provides unique
identification.

Task Duration:  The duration is the amount of time it will take to complete a
task, which is not necessarily the same as the elapsed time. For example, a task that
takes eight hours of effort to complete would be done in one day by a person dedi-
cated 100%, but if the person assigned this task is only working 50% on this project,
the task would take two days elapsed time to complete. All tasks must use the same
time units, which can be hours, days, weeks, or months, depending on the project. An
actual project starts on a specific date, but can also be measured from a point in time,
such as Day 1.

Start Day/Date:  The start day/date is the time that a task is scheduled to begin.
For example, suppose that a simple project has two tasks: Task 1 and Task 2. Also
suppose that Task 2 cannot begin until Task 1 is finished. An analogy might be that a
program cannot run until the computer is turned on. If Task 1 begins on Day 1 and
has duration of three days, it will finish on Day 3. Because Task 2 cannot begin until
Task 1 is completed, the start time for Task 2 is Day 4, which is the day after Task 1
is finished.

Finish Day/Date:  The finish day/date is the time that a task

is scheduled for completion. To calculate the finish day or date,
add the duration to the start day or date. When doing this, be
very careful not to add too many days. For example, if a task
starts on Day 10 and has duration of 5 days, then the finish
would be on Day 14—not Day 15.

3.4.3 Task Patterns
A project is based on a pattern of tasks. In a large project the
overall pattern would be quite complex, but it can be broken
down into three basic patterns: dependent tasks, multiple suc-
cessor tasks, and multiple predecessor tasks.
Figure 3-11  In a relay race, each runner is
dependent on the preceding runner and cannot
Dependent Tasks:  When tasks must be completed one start until the earlier finishes.
after another, like the relay race shown in Figure 3-11, they William Perugini/Shutterstock.com
Chapter 3 Managing Systems Projects
80­ 3.4  Identifying Task Patterns

EXAMPLE OF A DEPENDENT TASK are called dependent tasks because one de-
pends on the other. For example, Figure 3-12
Prepare Outline Create Document shows that Task 2 depends on Task 1, because
Task 2 cannot start until Task 1 is completed.
Start: Day 1 ID: 1 Start: Day 6 ID: 2 In this example, the finish time of Task 1, Day
Finish: Day 5 Dur: 5 Finish: Day 14 Dur: 9 5, controls the start date of Task 2, which is
Day 6.
Figure 3-12  This example of a dependent task shows that the
finish time of Task 1, Day 5 controls the start date of Task 2, which is Multiple Successor Tasks:  When
Day 6. several tasks can start at the same time, each
is called a concurrent task. Often, two or
more concurrent tasks depend on a single
prior task, which is called a predecessor
task. In this situation, each concurrent task
is called a successor task. In the example
EXAMPLE OF MULTIPLE SUCCESSOR TASKS shown in Figure 3-13, successor Tasks 2
and 3 both can begin as soon as Task 1 is
finished. Notice that the finish time for Task
Arrange
IdentifyInterviews
Needs
1 determines the start time for both Tasks
2 and 3. In other words, the earliest that
Start: Day 31 ID: 2
3 Task 1 can finish is Day 30, so Day 31 is the
Develop Plan
earliest that Tasks 2 and 3 can start.
Finish: Day 60
35 Dur: 30
5

Start: Day 1 ID: 1 Multiple Predecessor Tasks:  Suppose

Finish: Day 30 Dur: 30 that a task requires two or more prior tasks to
Design Survey
be completed before it can start. Figure 3-14
shows that example because Task 3 cannot
Start: Day 31 ID: 3 begin until Tasks 1 and 2 are both completed.
Finish: Day 40 Dur: 10 Since the two tasks might not finish at the
same time, the longest (latest) predecessor
Figure 3-13  This example of multiple successor tasks shows that the task becomes the controlling factor. Notice
finish time for Task 1 determines the start time for both Tasks 2 and 3. that the start for Task 3 is Day 16, not Day
6. Why is this so? Because Task 3 depends on
two predecessor tasks, Tasks 1 and 2, Task
3 cannot begin until the later of those tasks
is complete. Therefore, the start time for a
EXAMPLE OF MULTIPLE PREDECESSOR TASKS successor task must be the latest (largest)
finish time for any of its preceding tasks.
Obtain Authorization In the example shown, Task 1 ends on Day
15, while Task 2 ends on Day 5, so Task 1
Start: Day 1 ID: 1 controls the start time for Task 3.
Conduct Interviews
Finish: Day 15 Dur: 15
3.4.4 Identifying Task Patterns
Create Job Description
Start: Day 16 ID: 3 Task patterns are identified by looking care-
Finish: Day 45 Dur: 30 fully at the wording of the task statement.
Words like then, when, or and are action
Start: Day 1 ID: 2
words that signal a sequence of events. Here
Finish: Day 5 Dur: 5 are three simple examples:
Figure 3-14  This example of multiple predecessor tasks shows • Do Task 1, then do Task 2 describes
that the start time for a successor task must be the latest (largest) dependent tasks that must be completed
finish time for any of its preceding tasks. In the example shown, Task 1 one after the other.
ends of Day 15, while Task 2 ends on Day 5, so Task 1 controls the
start time for Task 3.
 Phase 1 Systems Planning

3.5  Calculating the Critical Path 81­

• When Task 2 is finished, start two tasks: Task 3 and Task 4 describes multiple
successor tasks that can both start as soon as Task 2 is finished.
• When Tasks 5 and 6 are done, start Task 7 indicates that Task 7 is a multiple
predecessor task because it cannot start until two or more previous tasks all are
completed.

3.4.5 Working with Complex Task Patterns

When several task patterns combine, the facts must be studied very carefully to
understand the logic and sequence. A project schedule will not be accurate if the
underlying task pattern is incorrect. For example, consider the following three fact
statements and the task patterns they represent. Examples of the task patterns are
shown in Figure 3-15, Figure 3-16, and Figure 3-17.

6
1 2

1 2 4

Figure 3-15  Dependent tasks. Figure 3-16  Dependent tasks and multiple successor tasks.

3 7

6
1 2 8

Figure 3-17  Dependent tasks, multiple successor tasks, and multiple predecessor tasks.

Dependent Tasks:  Perform Task 1. When Task 1 is complete, perform Task 2.

Dependent Tasks and Multiple Successor Tasks:  Perform Task 1. When

Task 1 is complete, perform Task 2. When Task 2 is finished, start two tasks: Task 3 and
Task 4. When Task 3 is complete, start two more tasks: Task 5 and Task 6.

Dependent Tasks, Multiple Successor Tasks, and Multiple

Predecessor Tasks:  Perform Task 1. When Task 1 is complete, perform
Task 2. When Task 2 is finished, start two Tasks: Task 3 and Task 4. When Task 3
is complete, start two more tasks: Task 5 and Task 6. When Tasks 5 and 6 are done,
start Task 7. Then, when Tasks 4 and 7 are finished, perform Task 8.

3.5 Calculating the Critical Path

Task patterns determine the order in which the tasks are performed. Once the task
sequence has been defined, a project manager can schedule the tasks and calculate the
critical path.
Chapter 3 Managing Systems Projects
82­ 3.5  Calculating the Critical Path

3.5.1  Critical Path

A critical path is a series of tasks that, if delayed, would affect the completion date of
the overall project. If any task on the critical path falls behind schedule, the entire
project will be delayed. For example, suppose that Joan and Jim are invited to some-
one’s home for dinner. Joan arrives on time, but Jim arrives 30 minutes late. Jim’s
arrival is part of the critical path because the host does not want to start without him,
so the meal will be served 30 minutes later than originally planned.
Project managers always must be aware of the critical path, so they can respond
quickly to keep the project on track. Microsoft Project and other project management
software can highlight the series of tasks that form the critical path.

3.5.2  Calculating the Critical Path

Figure 3-18 shows a training project with five tasks. Notice that the analyst has
arranged the tasks and entered task names, IDs, and durations. The task patterns
should be reviewed first. In this example, Task 1 is followed by Task 2, which is a
dependent task. Task 2 has two successor tasks: Task 3 and Task 4. Tasks 3 and 4 are
predecessor tasks for Task 5.
The next step is to determine start and finish dates, which will determine the criti-
cal path for the project. The following explanation outlines a step-by-step process.
The result is shown in Figure 3-19.

Plan Training

ID: 3
Dur: 5
Obtain Authorization Hire Analyst Announce Training

ID: 1 ID: 2 ID: 5

Dur: 10 Dur: 30 Arrange Logistics Dur: 30

ID: 4
Dur: 25

Figure 3-18  Example of a PERT/CPM chart with five tasks. Task 2 is a dependent task that has multiple successor tasks. Task 5 has
multiple predecessor tasks. In this figure, the analyst has arranged the tasks and entered task names, IDs, and durations.

Plan Training

Start: Day 41 ID: 3

Finish: Day 45 Dur: 5
Obtain Authorization Hire Analyst Announce Training

Start: Day 1 ID: 1 Start: Day 11 ID: 2 Start: Day 66 ID: 5

Finish: Day 10 Dur: 10 Finish: Day 40 Dur: 30 Arrange Logistics Finish: Day 95 Dur: 30

Start: Day 41 ID: 4

CRITICAL PATH:1-2-4-5 Finish: Day 65 Dur: 25

Figure 3-19  Now the analyst has entered the start and finish times, using the rules explained in this section. Notice that the
overall project has duration of 95 days.
 Phase 1 Systems Planning

3.6  Project Monitoring and Control 83­

• Task 1 starts on Day 1 and has duration of 10 days, so the finish date is Day 10.
• Task 2, which is dependent on Task 1, can start on Day 11—the day after
Task 1 ends. With duration of 30 days, Task 2 will end on Day 40.
• Tasks 3 and 4 are multiple successor tasks that can start after Task 2 is done.
Task 2 ends on Day 40, so Tasks 3 and 4 both can start on Day 41. Task 3 has
duration of 5 days, and will end on Day 45. Task 4 has duration of 25 days,
and will not end until Day 65.
• Task 5 depends on Tasks 3 and 4, which are multiple predecessors. Because
Task 5 depends on both tasks, it cannot start until the later of the two tasks is
complete. In this example, Task 3 ends earlier, but Task 4 will not be completed
until Day 65, so Task 5 cannot start until Day 66.

Recall that the critical path is a series of tasks that, if delayed, would affect
the final completion date of the overall project. In this example, Tasks 1 and 2
are the first tasks on the critical path. Now look at Task 5, which cannot start
until both Tasks 3 and 4 are done. In this case, Task 4 is the controlling factor
because Task 4 finishes on Day 65, which is 20 days later than Task 3, which is
completed on Day 45. Therefore, the start date for Task 5 is determined by the
finish date for Task 4.
In contrast, Task 3 has slack time, and could be delayed up to 20 days without
affecting Task 5. Slack time is the amount of time that the task could be late without
pushing back the completion date of the entire project. Tasks 1, 2, 4, and 5 represent
the critical path, which is highlighted with red arrows in Figure 3-19.