Project performance is one of the pillars of measurement inside a Project Management Office (PMO), and although it iss not only focused on gathering data for measuring project processes (faster product lifecycle, project delivery or others), it has received a lot of attention with automated workflows and applied Artificial Intelligence to it. So how we do it? Just to give an example, whether you have automated gatings or you rely on crowdsourcing evidence, you gather if all the delivery process or product lifecycle is followed (more about compliance than quality - for the purpose of this example). So you measure how much/many of the deliverables or gates are achieved compared to the full process. With that data you can apply data science methodologies to do classification (to understand who skips phases, deliverables, gates, or has the highest compliance, and according to what profile characteristics); then clustering (the do something similar to classification but by observations of natural groups); and then regression (linear regression to analyze variable correlations that will help for predictive analysis). The result is knowing and anticipating which kind of projects will comply more with a product lifecycle, the struggles of project process/delivery (according to size, stakeholders or other profile characteristics), but above all... the pain points of your Project Performance Lifecycle (standards and processes) that may be slowing the performance of Project/Program/Portfolio Managers or showing the areas where the PMO should improve (processes, reports, visibility) to accelerate project performance and value generation.
Comments