Demand has an address.

Ask the owner of a service business where their best customers come from and you will not hear a demographic. You’ll hear neighborhoods: the older subdivisions off the parkway, the golf communities, never the apartments by the highway. Owners carry a map of their own demand, learned one job at a time. The map is usually right. What’s interesting is why.

That question, in a different market, was my doctorate. My dissertation linked millions of property transactions to the demographic, economic, and administrative records of the places around them, to measure how a change in a neighborhood moves the value of the homes in it. Work like that cures you of treating geography as a backdrop. Place behaves like a variable: two houses a mile apart can sit in different markets, and effects travel along school zones and street networks, not in neat circles around a pin.

The same structure runs demand for almost every local service, and housing age is the clearest example. A subdivision isn’t just a place; it’s a cohort. Three hundred homes built in 2005 came with three hundred roofs installed in 2005, and most of the water heaters, driveways, and garage doors to match. Those systems fail on schedules. A neighborhood that was a wave of construction twenty years ago is a wave of replacement demand today — you can read the coming years of work orders in the year the homes went up.

This is a lesson applied economics keeps teaching: averages conceal structure. “The Tampa market” is not a market. It’s thousands of neighborhood-sized markets moving on their own schedules — different housing stock, different incomes, different ratios of owners to renters — and the interesting information is exactly the variation the metro-wide average smooths away.

Most businesses reason about geography anyway; they just do it with a circle. Draw a radius around the shop, call it the service area, done. But a radius assumes demand spreads evenly in all directions, like sound, and it doesn’t. It follows the housing stock and the incomes and the age of the subdivisions, and it stops at boundaries no circle knows about — the interstate, the school-district line, the edge of the flood zone. Two businesses with identical circles can face entirely different markets.

Place is one of the first structures I look for in any business: which neighborhoods its customers actually come from, and what those neighborhoods share — housing vintage, incomes, tenure. That list says more about a business than any average of its market does, because it’s the business as its customers reveal it, not as its owner describes it.

The owner’s mental map was right all along. It was never intuition — it was structure, learned one job at a time.