IT Project Selection and Portfolio Management

• Decision makers consider size, cost, purpose, length, risk, and scope of a project.
• The economic value of projects are challenged to select projects that provide the highest return on IT investments.
• Project portfolio management selects, prioritizes and monitors project results; this is critical for IT departments.
• System development project must undergo a planning and controlling process for: time, cost, and desired output.
• A realistic work assessment is a critical project management success factor.

Project Selection

• Systems projects are evaluated within the context of an entire portfolio.
• Determining a project's contribution reinforces the need for a feasibility study.
• Portfolio management considers the different projects in an organization.
• An approval committee allocates resources selectively due to limited funds.
• If there are several potentially high-payoff projects with the same risk, one project may be selected.

Waterfall Development

• A development approach with advantages that include identifying requirements long before programming begins, and limiting changes to those requirements as the project proceeds.
• Waterfall develpment, however, has disadvantages that include: the long time between analysis phase and system delivery, user can forget the original purpose of the system, and is not conducive to a dynamic business environment.
• Design must be completely specified before programming begins.
• A long time elapses between system proposal (analysis phase) and system delivery.
• Testing is treated as an afterthought during the implementation phase.
• Expensive post-implementation programming could be needed.
• Users may forget the system's original purpose due to the time elapsed.

Parallel Development

• Parallel development reduces time for delivering a system.
• It can result in voluminous deliverables.
• Integrating the subprojects can be challenging, if the subprojects are not independent.

V-Model

• Its advantages are attention to testing and a straightforward, simple and high quality system.
• Its primary disadvantage is the rigidity of the waterfall method.

Rapid Application Development: Iterative Development

• It improves speed and quality.
• However, managing user expectations is a potential disadvantage.

Rapid Application Development: System Prototyping

• It is useful for users with difficulty expressing requirements.
• It has a lack of careful analysis before design implementation.

Throwaway Prototyping

• It produces a stable system and is more reliable.
• Taking longer to deliver a final system is a disadvantage.

Agile Development

• A group of programming-centric methodologies streamline the SDLC.
• It includes interactive face-to-face communication.
• Extreme Programming (XP) emphasizes customer satisfaction and teamwork.
• In Agile development, a project starts with user stories that describe the system needs.
• Programmers then code in small modules and test to meet those needs.
• Standards minimize confusion by having teams use a common set of names, descriptions, and coding practices.
• XP projects deliver results sooner than RAD and are rarely bogged down by gathering the system requirements.
• XP works very well for small projects and highly motivated, cohesive, stable, and experienced teams.

Agile Development Process

• Development follows a continuous improvement cycle, exposing flaws faster and reducing waste
• Shorter development cycles, wider market windows, early customer feedback, and continuous improvement are advantages.

Selecting an appropriate development methodology.

• System Prototyping and Throwaway Prototyping are excellent with unclear user requirements.
• Iterative, System Prototyping and Throwaway Prototyping are excellent with unfamiliar technology.
• With projects that are complex, Waterfall, Parallel and V-Model are all rated good.
• V-Model is rated excellent when the projects are reliable.
• With short time schedules, Iterative, System Prototyping and Agile Development are all excellent.
• Iterative, System Prototyping and Agile Development are excellent with schedule visibility.

Estimating the Project Time Frame

• Two basic ways to estimate the time required to build a system are:
  • Use the amount of time spent in the planning phase to predict the time required for the entire project.
• Typical industry standards for business applications divide the project steps into:
  • Planning, 15%.
  • Analysis, 20%.
  • Design, 35%.
  • Implementation, 30%.
• Estimates based on actual figures for the first stage of SDLC were:
  • Planning, 4 person-months.
  • Analysis, 5.33 person-months.
  • Design, 9.33 person-months.
  • Implementation, 8 person-months.