The Dashboard Was Green. My Gut Said Otherwise. My Gut Was Right.

I wrote about data analytics a while back and mentioned that our most sophisticated models had failed to catch something a casual conversation with the customer service team surfaced instantly. That experience has stayed with me, because it crystallises a tension I run into constantly: we have more data available now than at any point in the history of project management, and yet the quality of our decisions hasn’t improved proportionally. If anything, the abundance of data has made it harder to know when to trust the numbers and when to trust the instinct built from years of being in the room when things go sideways.

Here’s a specific version of this I’ve lived through. A project dashboard shows everything tracking ahead of schedule. Every metric is green, every RAG is green, the velocity looks fine. But something feels off. The team is too quiet in standups. The questions that should be coming up aren’t being asked. A senior developer who usually pushes back on timelines has gone quiet. None of this shows up in the data, not because it isn’t real, but because the data doesn’t have vocabulary for it. That gut sense isn’t mysticism. It’s compressed experience reading signals that no dashboard was designed to capture. I’ve learned to take it seriously rather than wait for the metrics to confirm what I already knew.

The mistake I see most often is treating data and instinct as competing inputs, where you have to pick a side. They’re not competing. They’re two different instruments measuring different things. Data is exceptional at surfacing patterns across large information sets, spotting trends you’d miss in the noise, and providing objectivity when emotions run high or stakeholder pressure is distorting the room. Instinct is exceptional at reading context, sensing interpersonal dynamics, and catching things that are real but not yet measurable. The best decisions I’ve made used both, deliberately.

Where it gets genuinely hard is when they contradict each other. The data says proceed. Your gut says pause. In those moments, I’ve learned to ask one specific question: what does the data not know? Usually it doesn’t know about the conversation that happened after the meeting ended. It doesn’t know that the vendor’s key person just gave notice. It doesn’t know that the executive sponsor is quietly losing confidence in the team. When I can name what the data is missing, the right call almost always becomes clear and I can explain it to a stakeholder without it sounding like a hunch.

The PMs who’ll navigate the AI era best aren’t the most data-literate or the ones with the sharpest instincts. They’re the ones who know which tool to reach for in which moment. Data when the question is what’s happening. Instinct when the question is what does this mean. The discipline is knowing the difference, and that’s not something any algorithm can shortcut. It comes from showing up, paying attention, and being honest about what you’re actually seeing not just what the numbers are telling you to see.