Two candidate sites, one metro
A metal buildings manufacturer approached us with a decision that looked straightforward on paper. It had narrowed a new location down to two candidate sites in the same market, Kansas City, and needed to know which one would produce more revenue. Both sites were workable. Only one would get built next.
Decisions that look this simple are exactly where costly errors take root. A location is a commitment measured in years, not quarters. Choose well and the advantage compounds. Choose badly and the underperformance quietly weighs on the entire portfolio.
The choice the numbers seemed to make
Judged by the figures anyone can assemble in an afternoon, the answer looked obvious.
| Site A | Site B | |
|---|---|---|
| Households nearby | 69,000 | 11,000 |
| Businesses nearby | 4,400 | 900 |
Site A sat in the busier, denser part of the metro, with roughly six times the households and nearly five times the businesses within reach. More people, more commercial activity, more of everything a quick scan rewards. Most site-selection tools, and most instincts, would have committed to Site A without a second look.
What the model concluded
Our Real Estate Site Selection model reached the opposite conclusion.
| Site A | Site B | |
|---|---|---|
| Predicted first-year revenue | $7.8M | $10.3M |
Site B, despite serving a fraction of the surrounding population, was projected to generate about 32% more revenue, roughly $2.5 million more in its first year. The site that trailed on every visible metric was the one the analysis ranked first.
Why population counts mislead
The reason comes down to what actually drives this manufacturer’s revenue, which is not the size of the crowd.
Our model does not grade a location by how many people are nearby. It grades it by how many of the right people and businesses are nearby, and it defines “right” using the company’s own performance history. Trained on where this manufacturer already earns and what its strongest locations have in common, the model reads the specific blend of households, adjacent industries, and purchasing behavior that turns into orders.
Held to that standard, Site B’s smaller market was the stronger market. It contained far fewer prospects overall but a much denser concentration of the exact buyers this company converts. Much of Site A’s larger population was, in revenue terms, noise: households and businesses that inflate a count yet rarely become customers.
Count heads, and you choose Site A. Understand the market, and you choose Site B.
The takeaway for site selection decisions
The $2.5 million gap between the two sites was invisible to anyone reading the obvious numbers. It surfaced only when each location was measured against the client’s own revenue rather than against generic population and business tallies.
That gap is the entire value of a predictive approach. Not every household within reach is worth the same, and a location’s real potential lives in the concentration of the right ones, not the size of the crowd. The goal is not to describe a place. It is to tell you what your business will actually do there, before a lease is signed or ground is broken. The bigger market is not always the better market, and the only reliable way to tell them apart is to let your own results define what “better” means.
Ready to find out which of your candidate sites is quietly the stronger one? Start the conversation with the data-driven team at Ambient Array.
