Install Field Report · Clarity SI.
PH—009
Property intelligence is a dashboard problem disguised as an AI problem. Clarity SI proved that the right aggregation beats the fancier model most of the time. One screen. The number leadership wanted.
I What We Installed
Clarity SI operates in property intelligence for a small-portfolio real estate buyer segment. The install scope was deliberately narrow: one dashboard, one number on it.
Install ran six days. Two capabilities online at full scope: ingestion and cost analysis. The other four ran in reduced mode, sufficient for the dashboard’s needs. This was a deliberate choice. Not every install needs every capability lit at full brightness.
II The One Number
The number was not clever. It was a ratio. Gross yield over adjusted acquisition cost, filtered for the portfolio’s holding window and geographic heat. Leadership had been approximating it for years in Excel. They had never seen it computed correctly.
When they did, three things happened in the first week. One acquisition under contract was terminated. One property in the existing portfolio was flagged for refinancing. A third, previously deprioritized, was accelerated to offer.
The number did not require an agent. It required correctly structured data and the discipline to compute it every day.
III What the Dashboard Replaced
Four separate spreadsheets, three PDF quarterly reports, and two weekly meetings. The meetings did not disappear. They shortened. The spreadsheets disappeared.
The intelligence layer underneath runs on Terrein’s ingestion capability. It pulls from the company’s accounting system, its property management system, and two regional MLS feeds. Normalized to one schema. Updated daily. Surfaced as one number.
IV What the Dashboard Did Not
The install did not solve leadership’s harder question, which was how to originate better deals. That is a different install. We did not pretend otherwise. One of the reasons the install held is that we were explicit about scope.
Property intelligence is dashboard-shaped. Origination is agent-shaped. Conflating the two would have wasted everyone’s time. The trade-off every install faces: do the narrow thing correctly, or do the broad thing badly. We did the narrow thing.