Data-driven culture is AI key to delivering the UK’s infrastructure

McKinsey has set out the four changes necessary for the UK to improve infrastructure project delivery. A recent blog post by four McKinsey partners details the four changes: a stable portfolio of projects; a disciplined and accountable data-driven culture; the embrace of digital and AI tools; and the maintenance of consistent project leadership.

The partners state: “Our analysis across many capital projects shows that projects are typically managed through a variety of systems, often with inconsistent data definitions and reporting standards. While the UK has made progress on encouraging digital information management on buildings, it has proven challenging to extract the value of data at scale to manage complex programmes. Information on cost, scheduling and progress is often collected in silos at different intervals. It is typically focused on backwards-looking, high-level milestones, making it hard for project leaders to identify emerging risks early or to act decisively when delivery issues arise.

“Perhaps even more importantly, teams and leaders often fail to interrogate the analysis to spot issues, determine root causes, understand the impact on delivery, or identify when and how it can be resolved. Those who routinely accept poor data quality and analysis as good enough are missing insights and losing accountability – and a better way is rapidly possible.”

The partners continue that a strong culture of data and decision-making discipline should be embedded before construction. “Improving delivery outcomes will come from project leaders taking the simple but hard actions of enforcing better decision-making, owning the numbers, and interrogating facts to hold everyone accountable for delivery. This includes establishing consistent definitions, developing structured reporting, and prioritising accountability for using data to make decisions,” they add.

“Good practice entails tracking and steering critical-path activities at a granular level, including daily task adherence or weekly production rates. All projects should have a single, up-to-date view of project performance. This means project leaders can hold their stakeholders accountable.”

The McKinsey blog also notes that the data-driven culture should not end when the project ends, but concedes that the demand for post-completion evaluation is mixed because benefits often only materialise many years later.