Ask most people what makes a ten-year program hard and they'll say the schedule. It isn't. It's that the world keeps moving while you build.

The New Terminal One opens in phases through 2030, with a commercial brief built around New York: first-to-airport luxury retail, food and beverage drawn from the city, cultural integration borough by borough. It's the right ambition. It also contains a problem nobody names in the concept deck.

"Locally inspired" is a moving target. The restaurant that defines a Queens neighborhood today may not exist in 2030, and might not want an airport even if it does. A brand is only first-to-airport once. By the time the last phase opens, the concept that won the pitch will be six years old — in food and retail, roughly a generation.

So the discipline isn't executing the vision. It's sorting the decisions that have to be locked from the ones that must not be.

Lock the things that are ruinous to change: how many units and where, where anchors sit relative to passenger flow, MEP capacity, grease and ventilation routes, service access. Get those wrong and no amount of leasing talent saves you — the building has already decided what it can hold.

Keep open the things that cost little to change late: tenant identity, menu formats, finishes, digital content, the mix of global and local.

The common failure is inverting the two. Teams lock the tenant list first, because a tenant list is legible and makes a good board slide, and leave the infrastructure flexible, because deciding on infrastructure is unpleasant and expensive. Then the phase opens with a 2024 concept inside a base build that can't accommodate the change you now urgently need.

Avoiding that isn't a better plan. It's somebody holding both lists, every week, for years, and willing to say that a decision made in year two is no longer a good one in year five — without reopening the ones that genuinely are closed.

A plan for 2030 that hasn't changed by 2028 isn't a good plan. It's an unattended one.