Depending on the solution method you choose, you may be asked to provide an estimated productivity for the project. The Design to Goal and Time Boxed, Fixed Team methods require PI as an input to the estimate. The Feasibility and Trend Based solution methods work “in reverse”: they use either duration and resource goals or average duration and effort from the project’s trend group to calculate the productivity needed to execute the work, given the specified size, schedule, and resources. So for these methods, PI is actually an output that shows the productivity required to execute the project.
The Productivity Index (or PI) is a macro measure of a team’s development capability. Allowed values range from .1 to 40. Low values are associated with inexperienced staff, poor working environments and tools, or high product complexity. High values are associated with capable and experienced staff, good environments/tools/management, and well-understood, straightforward projects.
Of all the inputs to your estimate, PI and system size will have the most impact on the resulting schedule, effort, and quality. System size measures the magnitude of what you are building. Just as building a 236-room mansion requires more planning and coordination than constructing a one-room cabin, large software systems tend to have a higher number of separate components, each of which must communicate and interface with the others.
In addition to system size, there is another crucial input: process productivity or efficiency. Process efficiency or PI (all other factors being equal) greatly affects how long the project takes and how much it will cost. This efficiency or productivity level is affected by many factors: application complexity, task dependencies, team communication skills, overall environment, team skill level and experience, management influence, algorithm complexity, and tools/technology available.
If you have reasonably good historical data, it is always preferable to use a PI that has been validated in your environment and reflects your proven ability to deliver software. At project closeout, the actual productivity level (or PI) can be calculated from your completed project data using three simple inputs: system size, Phase 3 effort, and Phase 3 duration. Calculated PIs from your completed projects can be used to tune Collaborate’s estimation algorithms to the way your organization plans, develops, and deploys software.
If you have not yet captured closeout data to calculate actual PIs achieved on your completed projects, using industry data or trends to select an estimated PI is the next best thing.