To mitigate the upfront risks Princeton Information has developed a process that more accurately estimates the size, duration and costs of software projects. The process is implemented in a 4-step approach that starts with sizing the effort.
Step 1: We utilize the International Function Point User’s Groups (IFPUG) methods to determine the size of the effort. Software sizing utilizing IFPUG is a highly reliable technique that produces a size that is independent of language, technology, or system.
Step 2: The next step in the process is to convert the size counts into an effort-driven estimate. For this step, Princeton Information utilizes various Industry norms (parametrics) that include language, developer skills, use of automated methods, required quality (response times, turnaround times, depth and breadth of testing, etc.), re-usability, complexity, security, and required delivery dates (constraints).
Risk factors such as schedule, resources, technology, business, and others are also included in determining the estimate. An estimate confidence level is then calculated to produce the final estimate with a range of confidence between 90 to 95% accuracy.
Step 3: The third step in the process is to perform a comparative, bottoms-up estimate utilizing prior experience with like development projects. This estimate is driven by the assigned architect/designer/lead developer and is reflective of their experience with projects of the same technology and of similar size efforts. The two estimates are compared and are considered valid when the two estimates are within ±20%.
Step 4: The final step is to complete a staffing matrix based on the estimated duration. For this exercise the project’s period (phases, months, or weeks) are populated with the required staffing levels (e.g. project managers, architects, developers, testers, etc.) producing as output a balanced project staffing requirement for the life cycle of the effort.