What we pull from public sources to build a market page in 24 hours

One of the most common questions we hear during onboarding is “where does the data on my market pages actually come from?” The short answer: public sources. The slightly longer answer: a stack of public sources, joined together, with your point of view layered on top.

Here is roughly what we pull, per submarket, when we build a market page.

Demographics and population trends. US Census, American Community Survey, and Bureau of Labor Statistics. We pull population, median income, age distribution, employment by sector, and five-year trend lines. This becomes the “who lives and works here” section of the page.

Major employers and employer movement. State and county economic development sites, press releases, and SEC filings for any publicly traded companies headquartered or expanding in the submarket. This becomes the “what is driving demand” section.

Permits and pipeline. Where available, we pull permit data from the relevant county or city site, focusing on commercial construction and large multifamily. This becomes the “what is coming online” section.

Comps and transactions. Public records via county assessors and recorders for sale transactions, plus any press coverage of leasing activity. This becomes the “what is actually trading” section. Note that we do not pull from proprietary CRE databases — your CoStar credentials are yours and we do not need them.

Local context. News mentions, Chamber of Commerce activity, school district performance, transit access. These become the small details that make a page sound like an agent wrote it rather than a model.

Once the data is pulled, our AI stitches it together into a narrative — but the narrative respects two constraints. First, every claim links back to its source where possible, so the page is verifiable. Second, your voice and point of view are layered on top: how you read the trend, where you think the market is going, what you are advising clients to do.

The result is a market page that takes us about 24 hours to ship and would take a research analyst working full time the better part of a week to assemble manually. Multiplied across the three to seven submarkets a typical agent covers, that is the difference between having locally-authoritative content and not.