Concepts

The Score

The Buildability Score is a 0–100 integer answering one question — how buildable is this land? It is deterministic (same inputs, same score, every time), versioned (published methodologies are frozen forever), and explainable (every point traces to a factor with a cited source).

The nine factors

The Score is a weighted average of ninefactor sub-scores, each 0–100. Weights vary by state — flood matters more in Florida than in Colorado; wildfire the reverse — and the response tells you which region’s weights applied (result.region).

Factor key	Measures	Primary source
`zoning`	The dimensional standards on record — FAR, lot coverage, minimum lot size. Omitted when only a designation name is known.	Local zoning codes (via parcel data partners)
`flood`	Flood exposure: FEMA flood zone, Special Flood Hazard Area designation, and the share of the lot inside an SFHA.	FEMA National Flood Hazard Layer
`seismic`	Earthquake hazard at the site.	USGS seismic hazard data
`soil`	Soil suitability for building — drainage, hydric rating, series characteristics.	USDA SSURGO
`radon`	EPA radon zone (1 high → 3 low).	EPA radon zoning
`wildfire`	Wildfire risk at the location.	USFS wildfire risk data
`utilities`	Proximity to power, water, and road infrastructure. Omitted when unmapped — never assumed.	OpenStreetMap (Overpass)
`terrain`	Lot-scale ground slope — percent grade at the parcel.	USGS 3DEP
`geometry`	Lot shape regularity (area vs. bounding rectangle), 0–1 from the real parcel ring.	Parcel boundaries (Regrid)

Missing data is never a penalty. A factor with no data comes back with included: false and is dropped from the weighted average — the remaining factors renormalize. Check included before displaying any sub-score.

Grades, bands & status

The same number is expressed three ways for three audiences. These thresholds are part of the published methodology — they change only with a version bump.

Letter grade (`result.grade`)

Score	Grade
90–100	A+
85–89	A
80–84	A-
75–79	B+
70–74	B
65–69	B-
60–64	C+
55–59	C
50–54	C-
30–49	D
0–29	F

Band (`result.band`)

Score	Band
80–100	Excellent
60–79	Good
40–59	Moderate
0–39	Poor

Status (`result.status`)

Score	Status	Display label
70–100	`ic_ready`	“IC Ready” (or “Ready to Build” for consumer audiences)
50–69	`caution`	“Proceed with Caution” (“May Need Review”)
0–49	`high_risk`	“High Risk” (“Challenges Found”)

Data provenance

Every factor carries its lineage in the report’s provenance map: live (fetched from the authoritative source at request time), estimated (neutral assumption or regional estimate), fixture (curated offline dataset), or unavailable. If you show the Score, show the provenance — it is the difference between a citable number and a black box, and it is why agents can quote Buildability with confidence.

Methodology versioning

The current methodology is v2 — it adds the lot-scale terrain and geometry factors, grades zoning from its dimensional standards, and never applies optimistic defaults. Published versions are frozen and addressable forever: a score computed under v1 or v2reproduces exactly from the report’s inputs object, today or in ten years. When the methodology evolves, a new version ships alongside — existing stored scores never silently change. Two rules for integrators:

Persist result.version next to every score you store.
If you cache reports long-term, keep inputs too — it makes any historical score independently auditable.

The full methodology narrative — what each factor rewards and penalizes, and why — is on the Accuracy and Data Sources pages.