There are streets right by subway stations where shops fail — places where foot-traffic counts and actual sales diverge, because people passing and people staying are different data. A street that thousands pour through without once turning their heads: that is what we have been calling "a good location."
Station-area is the language of radius. Draw a circle of 300 or 500 meters and call everything inside it good land. But people do not move in radii; they move in paths. Which way the station exits face, into which alley the evening commute turns, where a crosswalk severs the flow — within the same radius, one street becomes a river and another an island. The radius exists on the map; the path exists on the ground.
So we look at dwell, not flow. Resident and workplace populations tell us the depth of the catchment; anchor facilities tell us why people gather; the hourly distribution of sales tells us how many hours a day the street is actually alive. Layer these and the effective demand — the demand that actually opens its wallet — emerges. A side street with a deep catchment outperforming a station-front block with a shallow one is not the exception. It is common.
When data and intuition collide, we doubt the data first, and then we doubt the intuition. There are things data cannot catch — a rail line about to open, a major employer preparing to relocate, the early signs of a street changing hands between generations. And intuition carries the past — the memory of a street that was once good tries to stand in for present judgment. The conclusion is simple: numbers speak of the present, the site speaks of what is coming, and both must be heard.
And while a location cannot be changed, a program can. On a site with a weak trade area, the answer is often a use that does not depend on it — destination businesses people seek out on purpose, residential, office. Highest and best use is always a function of location and program. Truly bad land is rarer than it looks. Land being used wrongly is everywhere.