Analytics needs to be well embedded into business culture, from strategy making to project management.IT project management is a critical step in strategy execution as it translates business priority into technique requirement and deliver the solution to improve business processes or delight customers directly. However, PM success rate is very low, how analytics may be incorporated alongside project management practices, and how a PMO may use analytics to create value for the organization?
Effort estimation - How do the method of estimation and type of project correlate with the actual effort? What is the deviation in effort like? Can you do some analytics in requirement gathering? If a business had historical project data, you could use analytics for resource demand, work, task duration, portfolio selection, resource availability, etc,. By its very nature, each project is unique, which means it runs into unique problems, unique risks, unique stakeholders, unique cultural differences, etc. In order to provide useful predictive analytics, you need to have a very decent sampling size, and you'll have to do some complex statistical analysis.
Monitor & Control: The use of analytics in a project would more likely fit into the monitor & control phase of a project that demanded the level of metrics. In order to successfully use it, you'd have to first identify your goals, set your project on a path to meet those goals, then report on the progress in regular intervals (every week, month, quarter, etc.) and address any shortcomings along the way through a change process. You may also use analytics to some degree in the risk management of a project to identify when it's not meeting the goals and implement a mitigation plan.
Measuring compliance to the guideline, or methodology of the organization would be critical data for the PMO and sponsor, so that you can show direct correlation between following or not following a Governance Model and the 'value-add' of the project. The PMO needs to define how it can add value to the organization and help the project, governance and management teams to deliver what your customer wants, as well as achieving the organization’s strategic vision by capturing and analyzing these metrics, and mapping their outcomes against KPIs where appropriate. The ability to effectively capture the metrics for analysis is as important as defining the metric/KPI, which may require changes to business practices and buy-in from stakeholders.
Thus, there are varying areas of project management where analytics could help bring out insights: Understand scope and end state. Strategy always drives Tactics. Understand and map current KPIs and SLAs. Build the same for the target environment. Measure your budget and schedule and honestly report it every week to your teams, SMEs, Stakeholders and PMO. There would be additional value gained as the metrics would provide earlier visibility to potential risks and issues. Another example would be that projects that trend toward extensive use of the governance and change processes ( especially around requirements) would certainly be something that might indicate the 'Requirements Gathering and Documentation Process' needs to be revisited and perhaps tweaked a bit. Still, Even if you do all of these, there is no guarantee that the analytics are going to give you a way of predicting how a project will progress because there is a very human aspect of running a project ... almost an intuition which is developed over years of experience.