Our method
Built on the public record, and honest about its edges.
Tubman reads the same public data anyone can request and measures it at a level the market usually skips. Here is what goes in, and what we refuse to claim from it.
What it is built on
Three things go in.
Public records only
City permits and business licenses, county assessor sales, safety records, federal Census and price indexes. Nothing proprietary, nothing scraped, nothing bought from a data broker.
Measured below the average
A community area is scored, and then the commercial corridors inside it are scored separately against each other. The gap between those two answers is usually the whole story.
Cross-checked before it ships
Every published figure is validated against an independent source or an independent method before it reaches the site. That gate has caught real errors in this work.
What we will not do
The limits are part of the product.
Anyone can publish a confident number. The harder and more useful thing is to publish what it cannot tell you, next to it, where you will actually read it.
We do not claim cause
Prices and permits describe what happened, not why. Where something moved, we say it moved, and we do not tell you what caused it.
We publish scenarios, never forecasts
A scenario is one stated assumption carried forward from an observed rate. It is not a prediction, and we never dress it up as one.
We do not publish what we cannot support
Where the data is too thin to carry a claim, the claim does not appear. What the model cannot do is stated next to what it can.
The specific limitations that govern each study are published on that study, beside the figures they apply to. See the South Shore and Uptown studies.