Predictive Analysis: An Overview
The value of predictive analytics increasingly takes hold in international organization. Delays are not just costly but also frequent; an analysis of more than 1,800 software projects revealed a 30% on-time completion time. On average, a project was behind around 25%of the original deadline. And while time in itself carries strategic value, the deployment of extra resources needed to deal with overrun costs can be direct. Research on average budget overrun experienced by factory-automation-software projects studied exceeded 10%. One in five projects had overruns exceeding 50%. Then there are the indirect costs.
One of the main reasons for overruns appears to be a misunderstanding by managers and engineers by the complexity of a project. Teams remain unaware of the large impact of features and performance targets and the costs of implementing features into the final product. Psychologically, project progress is thought of as linear, a concept not applicable to reality. Additional to that, the overestimation of productivity of the development team causes additional delays. Project planners appear quick to ignore previous problems and ignore the potential for new issues.
To combat the issue of overruns, analytical models provide a powerful new way to deal with such constraints. Through the use of predictive analytics software, a company is able to model multiple projects running concurrently, and test for factors such as staff demand and possible resource bottlenecks. Planners can use the data to make adequate adjustments, whether it be through contracting, hiring or outsourcing.
Keith Knutsson of Integrale Advisors commented, “we increasingly see predictive analytics enhance human decision-making. The technology in itself carries much room for improvement, but even at it’s current state we are looking at very promising results.”