Worked examples

What 20 seconds actually surfaces.

No customer testimonials here yet — we’d rather show you the engine than invent stories. These are worked examples of what the screen checks, including one real, re-runnable address and our published backtest.

Loss avoided~20 seconds

The floodplain, surfaced first

FEMA NFHL · Flood screen · Scenario

A lot can look perfect from the street and still sit partly inside a FEMA Special Flood Hazard Area. The flood factor scores the share of the parcel inside the SFHA — a creek clipping one corner reads differently than a fully mapped floodway — so the constraint shows up before the offer, not after the appraisal.

The flood factor comes from the FEMA National Flood Hazard Layer, queried live, and the report names the zone and the lot fraction it covers.
Scenario · FEMA NFHL
Data source
FEMA NFHL, queried live
Signal
Share of the lot inside an SFHA
Where it lands
Flood factor + report hazards section
Loss avoided~20 seconds
The five-figure hookup, flagged

OpenStreetMap · Utility gap · Scenario

On rural and edge parcels the make-or-break question is distance to power, water, and road access. The utilities factor scores proximity from mapped infrastructure — and when mapping is too thin to trust, the factor drops out instead of being assumed serviceable.

Unknown is a finding, not a pass: methodology v2 never applies an optimistic default to utilities.

Data source
OpenStreetMap infrastructure
Signal
Proximity to power, water, road
When unmapped
Factor omitted, never assumed
Loss avoided~20 seconds
Slope you can’t see in photos

USGS 3DEP · Terrain check · Scenario

Two adjacent lots can differ by an entire foundation budget because one falls away at 20% grade. The terrain factor reads lot-scale slope from USGS 3DEP — one of the two v2 factors that separates a parcel from its neighbors instead of scoring the whole ZIP code the same.

Terrain carries the heaviest default weight in methodology v2 — 18% — because slope differs house to house while county-level hazards don’t.

Data source
USGS 3DEP, queried live
Signal
Percent grade at the lot
Default weight
18% — heaviest in v2
Loss avoided~20 seconds
The unbuildable sliver, caught

Parcel ring · Lot geometry · Scenario

Acreage alone hides shape. A long, irregular sliver builds very differently than a clean rectangle of the same size. The geometry factor computes shape regularity from the real parcel boundary — area against its bounding rectangle — so a flag lot or remnant strip scores like what it is.

Geometry is the second within-market factor added in methodology v2, computed from the actual parcel polygon rather than a radius guess.

Data source
Parcel boundaries (Regrid)
Signal
Shape regularity, 0–1
Default weight
10%
Score = min(composite, every landmine cap). The cap appears as a named banner in the report — you can re-run this exact address in the app today and see it.
Worked example · Austin, TX
Landmine
Soil rated “Very limited” for septic
Published score
60 · C+ · caution
Mechanism
Score = min(composite, caps)
Loss avoidedRe-run it yourself

92-point land, published at 60

Austin, TX · Septic-limiting soil · Worked example

A real address we use as the docs example — 1207 E 30th St, Austin — scores well on flood, terrain, utilities, and geometry. But USDA rates the soil “Very limited” for septic absorption fields, so the Buildability Formula caps the published score at 60 · C+ · caution instead of letting a strong average bury the finding.

Time saved
7/10

the built lot outranks its never-built neighbor (Austin backtest, methodology v1)

Austin backtest · Methodology v1 · Published test · July 2026

We tested the score against real build outcomes: 76 completed new-construction permits against 369 comparable never-built parcels in Austin, TX, all public data. Within the same neighborhood, methodology v1 ranked the lot that actually got built above its never-built neighbor about 7 times out of 10.

Those numbers are evidence about methodology v1. The current engine, v2, ships engineering-calibrated pending its own backtest — the full design and limitations are on the Accuracy page.

76 built vs. 369 never-built parcelsDesign
Built lot wins 71–77% of pairsNeighbor-matched result
Methodology v1 — v2 backtest pendingApplies to
Upside found~20 seconds
The envelope, in numbers

Municipal code · Zoning envelope · Scenario

A district name alone tells you almost nothing. Where the jurisdiction publishes dimensional standards, the report turns FAR, lot coverage, height, and minimum lot size into a buildable envelope — units, floors, and square feet — and names the constraint that binds first.

When only a designation name is known, v2 omits the zoning factor entirely: a name is not an assessment.

Signal
FAR · coverage · min lot · height
Output
Max units, floors, buildable sf
If standards unknown
Factor omitted, never guessed
Time saved~20 seconds
A line item, not a landmine

EPA zones · Radon check · Scenario

EPA Zone 1 radon isn’t a deal-killer — it’s a mitigation system you should price before you bid. The radon factor reads the county designation and takes a moderate, honest penalty, so the report tells you to budget for it instead of scaring you off the lot.

Every factor carries a plain-English note naming its dataset, so you can tell a budget item from a blocker.

Data source
EPA radon zones
Signal
County zone designation
Reading
Moderate penalty, clearly labeled
Loss avoided~20 seconds
Severity that won’t average away

Federal hazard layers · Wildfire exposure · Scenario

Very-high wildfire hazard potential pulls its factor deep into the red — and in California the wildfire weight rises to 16%, so great zoning can’t paper over a WUI lot. Insurance is where this bites first; the screen puts it in front of you before the insurer does.

Severe findings dominate their factor instead of averaging away — that behavior is published on the Score methodology page.

Data source
Federal wildfire hazard mapping
Regional weight
Up to 16% (CA)
If unmapped
Factor omitted
Upside foundOne question

The entitlement, priced before the hearing

Geo · Upzone modeling · Scenario

Wondering what a rezone is worth? Geo’s run_scenario tool re-scores the parcel under the changed assumption — upzone, utilities, hazards — and returns the before/after score and unit delta, computed by the same versioned engine as every report.

Scenarios run through the real Score engine, not a separate model — the delta is auditable factor by factor.
Scenario · Geo
Tool
run_scenario (Geo)
Output
Before/after score + unit delta
Engine
Same versioned methodology
Get started

Ready to find your edge?

Run your first property analysis free. Nine weighted factors. 5 federal agencies queried live. 20 seconds.

Analyze a property — free.

Every check above starts with one address. Run yours and see what 20 seconds surfaces.

Analyze a property — free

No signup required · Results in 20 seconds