Measure/Foundation repair leads
Territory access · for contractors
Foundation repair leads, sourced from the ground itself
Foundation repair leads normally come from someone who already noticed a crack. These come earlier: addresses where we measured, from remote 3D measurement, that the ground outside the wall tips toward the house — on which walls, over how many meters, and with how much land draining into them.
The honest framing first. Nothing in this dataset looks at a foundation. It looks at the earth against it. Soil moisture cycling beside a footing is a recognized driver of the movement you get called out to fix, and ground arranged to deliver water to a wall is a condition you can measure from the air. What has already happened inside that wall is what your inspection is for.
What is measured on each wall
Each wall gets its own outward normal — not one slope averaged across the property, which is the number that makes a corner lot look fine when one elevation is a bowl.
- The full first three meters. Slope is measured from 0.75 m to 3.25 m out from the wall, not read off a single pair of points.
- No single reading decides a wall. The measurement is repeated along the wall and combined, so one bad reading cannot carry it.
- Nothing nearer than 0.5 m to the wall. Inside a footprint the bare-earth surface is not ground — it is interpolation across a building that was removed.
- The target is a 5% fall over the first three meters. Positive means the ground rises away from the house and water runs back at it.
Nearby readings are not independent observations, so the stated uncertainty is computed on the information actually present rather than on the count of samples taken. Oversampling buys smoothness, which is real, and buys no false confidence, which is the point.
Then the house is moved
A published building outline is a trace of a roof, and it can sit a meter from the wall it claims to be. That matters more here than almost anywhere: shifting a real Pittsburgh footprint north in 0.2 m steps moved one wall from +5.4% to −23.5%, crossing zero inside 1.4 m.
So the footprint is displaced one meter in each of sixteen directions and every wall is re-measured. Seventeen positions in all. A direction is published only when all seventeen agree on the sign.
It does not demand that the magnitudes agree. A meter of movement on steep ground can legitimately swing a grade from −52% to −13%. That says the hillside is not a plane, and nothing whatever about which way the water goes.
Walls that fail this pass are shipped as too close to call. They do not put a property on your list.
Getting the right building, which is the harder problem
The best terrain model in the world is worth nothing pointed at the neighbor's roof. Identity is decided by graded evidence, strongest first, and the engine stops the moment it has an identification. Which grade answered ships with every measurement; the sequence itself does not, because a building we identified and a building we merely located are not the same claim.
| How the building was identified | What that claim is | How it is reported |
|---|---|---|
| Direct identification | The address is tied to this outline in particular | Identified |
| Corroborated identification | Independent evidence puts the address inside this outline and no other | Identified |
| Nearest outline — the guess, reported as one | The nearest outline is all there is | Located, not identified — never sold as a report |
| Nothing fired | No evidence identifies a structure | Refused, and never billed |
The weakest grade is reported as the guess it is rather than promoted, and the grade that answered ships with the result. We publish no headline identification rate: one figure would hide the thing that decides whether an answer is safe to use, which is what identified the building on this address.
Roughly half the remaining errors are not a different building at all. Ten of the eighteen wrong rows sit within five meters of the addressed structure — our source and the judge tracing the same roof differently. That is a judging convention, and it is worth nothing to you either way.
Territory lists take only high-confidence identifications, so the proximity path in the table above is excluded from them entirely.
The claim, and its boundary
The sentence that ships with an address is "the ground falls toward this structure on N m of its walls". It is a measurement of ground slope at the walls. It is not a finding that the property has a drainage problem, a wet basement, foundation movement, or any defect. Whether water has ever reached this structure was not measured and is not claimed.
We also quote no conversion rate. We tested measured terrain against Chicago's 311 basement-water complaints and it did not separate them — mostly a verdict on the proxy, since an owner who knows their grading is bad calls a contractor rather than the city. It is a reason not to promise you a number we have never observed.
What it costs
Territory access starts at $14 per property: 25 for $450, 100 for $1,600, 500 for $7,000, or $35 each held exclusively to you for twelve months. It is a separate product from the platform plans — Solo at $99 a month for 25 lookups, Pro at $399 for 150 — and it is not metered against them.
A refusal is never billed. If the building is wrong, the property is replaced free. One contractor per territory, and we never sell the same property twice.
Check whether your territory is open
Five questions, then a 30-minute market analysis: which ZIPs we hold measurement coverage for, your own addresses run live including the refusals, and how a territory gets bounded.
We have never measured how many properties qualify in any market, and we will not quote you a figure we have not computed.
Questions
Does this detect cracks, settlement or heave?
No. Nothing here looks at the foundation. It measures the ground outside it. Structural condition is what your inspection is for — this tells you where to send it.
Why should a foundation contractor care about slope?
Repeated wetting of backfill against a wall, and the moisture cycling that follows, is ordinary building science. Measure shows you where the ground is arranged to deliver that water. It does not tell you a particular foundation has moved.
How precise is a wall slope figure?
Measured out to 3.25 m, repeated along the wall and combined, with the uncertainty computed on the information actually present rather than the sample count. The direction is published only if it survives seventeen footprint positions.
What happens when it cannot answer?
It refuses and names the quantity it declined. An address that cannot be found is refused outright rather than resolved to something nearby. A refusal is never billed.
Can I check it against a job I have already done?
That is the right test. Run an address you have already excavated and see whether the walls we call match what you found. A full report on any property is $25.