Page 510 - Using MIS

478 Chapter 12 Information Systems Development

System denition
1.1 Dene goals and scope
1.1.1 Dene goals
1.1.2 Dene system boundaries
1.1.3 Review results
1.1.4 Document results
1.2 Assess feasibility
1.2.1 Cost
1.2.2 Schedule
1.2.3 Technical
1.2.4 Organizational
1.2.5 Document feasibility
1.2.6 Management review and go/no-go decision
1.3 Plan project
1.3.1 Establish milestones
1.3.2 Create WBS
1.3.2.1 Levels 1 and 2
1.3.2.2 Levels 3+
1.3.3 Document WBS
1.3.3.1 Create WBS baseline
1.3.3.2 Input to Project
1.3.4 Determine resource requirements
1.3.4.1 Personnel
1.3.4.2 Computing
1.3.4.3 O‘ce space
1.3.4.4 Travel and Meeting Expense
1.3.5 Management review
1.3.5.1 Prepare presentation
1.3.5.2 Prepare background documents
1.3.5.3 Give presentation
1.3.5.4 Incorporate feedback into plan
1.3.5.5 Approve project
1.4 Form project team
1.4.1 Meet with HR
1.4.2 Meet with IT Director
Figure 12-16 1.4.3 Develop job descriptions
Example Work Breakdown 1.4.4 Meet with available personnel
Structure (WBS) 1.4.5 Hire personnel

In Figure 12-16, the overall task, System definition, is divided into Define goals and scope,
Assess feasibility, Plan project, and Form project team. Each of those tasks is broken into smaller
tasks until the work has been divided into small tasks that can be managed and estimated.
Estimate Time and Costs

As stated, it is exceedingly difficult to determine duration and labor requirements for many de-
2
velopment tasks. Fred Brooks defined software as “logical poetry.” Like poetry, software is not
made of wood or metal or plastic; it is pure thought-stuff. Some years ago, when I pressed a sea-
soned software developer for a schedule, he responded by asking me, “What would Shakespeare
have said if someone asked him how long it would take him to write Hamlet?” Another popular
rejoinder is, “What would a fisherman say if you ask him how long it will take to catch three fish?
He doesn’t know, and neither do I.”
Organizations take a variety of approaches to this challenge. One is to avoid scheduling
problems altogether and never develop systems and software in-house. Instead, they license
packages, such as ERP systems, that include both business processes and information systems
components. As stated in Chapter 7, even if the vendor provides workable processes, those
processes will need to be integrated into the business. However, the schedule risk of integra-
tion activities is far less than those for developing processes, programs, databases, and other
components.
But what if no suitable package exists? In this case, companies can admit the impossibility
of scheduling a date for the completion of the entire system and take the best result they can get.

2 Fred Brooks was a successful executive at IBM in the 1960s. After retiring from IBM, he authored a classic book
on IT project management called The Mythical Man-Month. Published by Addison-Wesley in 1975, the book is
still pertinent today and should be read by every IS manager. It’s informative and quite enjoyable to read as well.