Most organisations are about to have access to the most powerful project analysis capability ever built that should radically change execution effectiveness. Yet many will get very little from it. The reason isn’t the AI. It’s what’s underneath it.
Most projects have a schedule. Far fewer have a baselined schedule that’s been tracked consistently through delivery. That distinction matters more than most people realise – because a schedule without a baseline is just a list of dates. It tells you where things are. It can’t tell you what changed, when it changed, or why. And without that, AI has nothing meaningful to work with.
I ran this recently on a project schedule that I was tracking in Smartsheet and which had slipped. One prompt to AI asking for a performance assessment against baseline. A minute later – a written analysis with root cause and recommended actions.
The one-off isn’t the point. The point is what becomes possible when that logic runs across every project in a portfolio, historically and in real time. Where do we consistently over-commit? Where do we slip? What does that tell us about how we estimate?
The same principle applies to project learnings – a different beast, but the same underlying problem. Learnings have always been captured badly because the friction was too high. Delivery pressure wins against process discipline every time.
But that excuse is largely gone now. A tech-savvy PM working with AI and transcription tools can extract meaningful learnings from meetings across the full project lifecycle almost automatically. Apply that consistently across a portfolio and you build something most organisations have never had: institutional memory that’s actually retrievable.
Two different capabilities. One principle. AI can analyse data at a scale and speed no PM can match – but only if the data exists, is structured, and has been captured consistently.
Is your project infrastructure ready? And are your PMs?
