Updated August 27, 2026
Everyone in this business repeats the same rule: larger buildings trade lower per door. It is the first thing offered when a per-unit figure looks wrong.
It is also testable, and this desk's record is large enough to test it. So here is the test, and here is where the rule earns its keep and where it does not.
| Building size | Buildings | Median per unit | Range |
|---|---|---|---|
| 4–10 units | 54 | $300,000 | $104,286 – $788,889 |
| 11–25 units | 50 | $248,125 | $83,750 – $1,083,619 |
| 26–50 units | 30 | $220,757 | $57,967 – $485,588 |
| 51+ units | 11 | $177,778 | $92,188 – $268,644 |
Every step down in size band is a step up in price per unit, across all four bands, with no exception. From the smallest band to the largest the figure falls by well over a third.
The mechanism is not mysterious. A four-unit building at two million dollars is reachable by individual buyers, small partnerships, exchange buyers and first-time investors. A sixty-unit building at fifteen million is not. Fewer buyers means less competition, and per-unit rents do not rise with unit count to make up the difference. The result is a lower number per door on a larger building of equal quality.
This is the part the rule usually leaves out, and it is the part that matters if you own one building in one neighborhood.
Of the ten submarkets profiled on this site, 9 have enough closings across enough size bands to test the pattern at all. It descends cleanly in 4 of them. In the others the bands cross over — sometimes mildly, sometimes emphatically. In Sherman Oaks the largest band prices highest of the three.
That is not a market mystery. It is sample size. A single town's band rests on somewhere between two and twelve buildings. At those counts one unusual closing — a renovated building, a corner site, a distressed seller, a redevelopment bid — moves the median for the whole band. Pool the towns and the noise cancels. Split them and it does not.
Because the rule alone is what produces the mispricing it is supposed to prevent.
An owner reads a submarket band table, finds the row matching their building's size, and prices from it. In four towns that lands close. In five it can be a long way off, and the owner has no way to know which they are in — because the table looks identical either way.
A number published without its limits is a claim. Published with them, it is evidence.
Use the pooled table above to understand the direction, which is real and which you should expect a buyer to apply.
Do not use a single submarket's band as a price. Use it as context, then find the closings that actually resemble your building — same size, same corridor, same regulatory position, closest in time. Several of those are published on this site, by name, with dates and prices.
And where a same-street or same-building comparison exists, it beats every band. The strongest evidence in this entire record is not a median at all: it is one Sherman Oaks building that sold twice, ten months apart, and two West Hollywood buildings facing each other across one street.
145 buildings across ten submarkets from one desk's record, not a market index. The archive holds addresses, unit counts, prices and dates — no cap rates, rents, income or condition — so this measures what buildings sold for by size, and nothing about why any individual one did. The full methodology.
Request a free evaluation of your building →
Thinking about selling? Get a no-obligation evaluation on your building.
Request Free Evaluation →