Billy Beane revolutionized the Oakland A’s with analytics. For seven years it worked. Then everyone read the book, hired their own analysts, and the A’s winning percentage dropped from .586 to .470.

The insight became infrastructure. The edge became the baseline.

This is the pattern that’s playing out across every industry right now. And the built environment is not exempt.

What Moneyball actually teaches us

The lesson most people take from Moneyball is “use better data.” The actual lesson is more consequential: competitive advantages in information don’t last.

The moment a new methodology becomes legible (when it can be described, documented, and replicated) it stops being an edge and becomes a cost of entry. Every team ran analytics after Beane. Every investment bank ran algorithmic trading after the hedge funds proved it worked. Every retailer optimized their supply chain after Walmart showed how.

The firms in AEC who got ahead of digital tools in the 2010s had an edge. For a while. Now BIM is mandatory on most commercial projects. Digital drawing coordination is expected. Those tools are infrastructure, not differentiation.

AI is following the same curve. The firms treating AI adoption as a competitive advantage today are right, for now. In three to five years, AI-enabled workflows will be the floor, not the ceiling. What matters is what you build on top of the floor.

Where judgment becomes the moat

When data and process tools become widely available, the competitive premium shifts to judgment. To knowing which data matters and why. To making good decisions in situations where the data is ambiguous or incomplete. To building relationships and trust that no algorithm captures.

In construction project delivery, this has always been true and is becoming more true. The superintendent who has seen this failure mode before and knows to look for it three months earlier than anyone else, that’s not a tool. That’s pattern recognition accumulated over decades. AI doesn’t have it. AI helps the person who has it work faster.

The firms that will lead the next decade of the built industry are the ones building the organizational infrastructure to develop that kind of judgment at scale: mentorship systems, knowledge transfer mechanisms, post-project learning that actually changes future behavior. Not just deploying better tools.

What this means for owners hiring partners

If you’re evaluating which firms to bring onto your project, the technology question is necessary but not sufficient. Yes, ask whether they’re using AI in procurement analysis, schedule modeling, and document review. That’s table stakes.

The deeper question is: what do they do with the time AI gives back? What does a project manager at this firm do with two extra hours a day that aren’t spent on administrative work?

Do they use it to be proactively ahead of problems? To give you earlier warnings? To do more thorough coordination review? Or do they use it to service more projects with the same team, which is great for their firm and may not be great for yours?

The Moneyball teams that stayed competitive past the first wave didn’t just have better data. They developed better judgment about how to use it. They built organizational cultures that kept learning after the initial edge was gone.

The real question for 2026

The industries facing the hardest questions right now are the ones that built their competitive position around information advantages (being the expert, the insider, the holder of specialized knowledge) that AI is now making widely accessible.

AEC is one of those industries.

The firms that survive and lead the next phase won’t be the ones with the best AI tools. Those will be commoditized quickly. They’ll be the ones with the best judgment about what to do when AI surfaces a problem. The best relationships. The deepest institutional knowledge about their specific building types, markets, and client challenges.

That’s not built by buying software. It’s built by investing in people and the systems that make them better over time.

The edge is always temporary. The question is what you’re building for when it runs out.