Scan-to-estimate is a workflow that collapses two jobs into one. Traditionally, an estimator measures a structure — by tape, by wheel, by aerial report, or by hand off a set of plans — and then, in a separate sitting, prices that measurement against an assembly catalog to produce a bid. Scan-to-estimate fuses those steps: the operator captures the geometry of the structure in the field with a 3D sensor, and software turns that capture directly into a priced, line-item estimate. The takeoff and the bid become a single motion.
The phrase describes the whole pipeline, not just the sensor. A LiDAR scan that produces a point cloud you then export to a measurement tool is not scan-to-estimate — it's scan-to-measurement, and the estimate is still a second job. Scan-to-estimate means the output is the estimate: line items, quantities, waste factors, and dollars, in the trade's own assemblies, ready to turn into a proposal. The distinction matters because most of the market sells the first half and calls it the whole thing.
The three steps it compresses
Step one is measurement — establishing the dimensions, areas, heights, and spatial relationships of the structure being bid. On a commercial flat roof that means gross and net area, parapet heights, drain and curb placement, penetrations through the membrane. In a server or rack room it means wall runs, ceiling height, conduit pathways, and equipment clearances. Historically this was the slowest, most error-prone step, because it depended on a person physically measuring and transcribing.
Step two is quantity takeoff — translating raw measurement into the materials and labor a specific assembly requires. A measured roof area isn't a bid; it becomes a bid only when it's multiplied through a membrane assembly with its insulation package, fasteners, adhesive, flashing, walkway pad, and the waste factor that the geometry actually warrants. This is where trade knowledge lives, and where two estimators measuring the same roof can still produce two different bids.
Step three is pricing — applying current material costs and labor rates to the quantities, producing the line-item dollar figure the customer sees. Scan-to-estimate is the claim that all three steps can run off a single in-field capture, fast enough that the operator leaves the site with a defensible number instead of a homework assignment.
Why it matters for commercial installation contractors
The economics of commercial bidding are brutal on slow estimating. A commercial installation contractor — roofing, security and fire, AV and low-voltage, solar, HVAC, electrical — wins work by bidding more jobs accurately than the competition can. Every estimate that takes a day to produce is a day the bid sits unsubmitted, and every measurement done by hand carries an error budget that either pads the bid until it loses or undercuts it until it bleeds.
Manual takeoff also doesn't scale with the estimator's calendar. The constraint on most commercial shops isn't sales appetite — it's how many bids one or two people can produce before quality slips. Scan-to-estimate attacks that constraint directly: if the field tech can capture the structure during the walkthrough and the estimate writes itself from the capture, the bid throughput of the shop stops being gated by the estimator's transcription speed.
And the bid that comes off a ground-truth capture is the bid you can stand behind in front of an engineer or a general contractor's PM. A number built from measured geometry — not a guess off a stale photo or a plan that doesn't match what got built — survives review. That's the difference between bidding from the truck door and bidding from the roof you just walked.
Scan-to-estimate vs. manual takeoff
Manual takeoff is the incumbent, and it has one real virtue: a skilled estimator with a tape, a wheel, and a set of plans can measure anything, including geometry no sensor handles well. There's no substitute yet for a veteran estimator's judgment on a strange roof, and there are surfaces and angles a person reaches with a tape that a depth sensor reads poorly. Where the geometry is hostile to capture, manual takeoff remains the right call, and an honest scan-to-estimate workflow keeps a manual override for exactly those cases.
What manual takeoff costs is time and consistency. It's hours per bid, it varies estimator to estimator, and the transcription step — copying tape readings into a spreadsheet, the spreadsheet into the estimate — is where the quiet errors enter. Scan-to-estimate trades the universality of the tape for speed and repeatability: the same structure captured twice produces the same geometry, and the measurement never gets retyped because it never leaves the system.
Scan-to-estimate vs. aerial measurement
Aerial-imagery measurement — the satellite-and-drone report you order and receive back later — is the other thing operators reach for, and it's genuinely useful for properties a field tech can't safely access. If the roof is too steep, too tall, or too dangerous to walk, an aerial report beats not measuring at all, and a good scan-to-estimate workflow treats aerial as the fallback for exactly that situation.
But aerial measurement has two structural limits that scan-to-estimate doesn't. First, the imagery is only as fresh as the last flyover, so a building that was retrofitted recently can measure as the old structure. Second, aerial can't see what's behind a parapet, under an overhang, or inside a courtyard, which is precisely the commercial geometry where measurement precision pays. Scan-to-estimate captures the structure the contractor walked through this morning. The thing under their feet is the thing being measured, and the estimate reflects current reality rather than a remembered one.
What it can't do — the honest limits
Scan-to-estimate is a measurement workflow, and it inherits the limits of whatever sensor feeds it. LiDAR-based capture doesn't read glass and mirrored surfaces reliably; the infrared pulse passes through or reflects in ways the time-of-flight calculation can't resolve. Direct sun on a hot membrane adds noise. Sub-millimeter detail — a hairline crack, a bolt-pattern offset to surveyor precision — is out of reach of a room-and-structure-scale sensor.
It also doesn't replace judgment. The capture produces geometry and a first-pass quantity takeoff; it doesn't know that this customer always specs a heavier walkway pad, or that the AHJ in this jurisdiction wants a detail the catalog assembly omits. Scan-to-estimate gets the estimator to a defensible draft in a fraction of the time — it does not get them out of reviewing the draft. The contractors who get the most out of it treat the generated estimate as a strong starting line, not a finish line, and the workflow has to support editing it as such.
Anyone selling scan-to-estimate as a hands-off magic number is overselling it. The honest pitch is narrower and more valuable: capture the geometry once, in the field, accurately, and never retype it — so the estimator spends their time on the judgment calls instead of the transcription.
How Forge does scan-to-estimate (Hyperion)
Hyperion is Forge's scan-to-estimate instrument — an iPad and iPhone LiDAR scanning system that runs the full pipeline. The operator walks the structure; the sensor captures its geometry and the penetrations and obstructions on it. Forge's trade assemblies, waste factors, and material costs apply to that geometry, and a line-item estimate comes back in field conditions — measurement, takeoff, and pricing in one motion rather than three.
On commercial flat and low-slope geometry, Forge publishes an approximate ±0.8% accuracy figure for Hyperion; on tighter, more complex geometry that figure widens to about ±2%. That figure is currently published as a methodology and sample plan at /proof/hyperion-accuracy, not as an independently verified benchmark — independent third-party verification is not yet complete, and the number stays hedged on the proof page until it lands. Forge publishes it the way it actually stands rather than the way it would market best; it graduates to a measured benchmark only after outside verification.
The reason Hyperion produces an estimate and not just a measurement is the federation — the same word that runs through everything Forge builds. Estimating, scheduling, CRM, documents, Treasury, and field ops are drawers of one chest that share architecture rather than trading files across integrations. So the geometry Hyperion captures lands in the estimate, the estimate becomes a proposal, the won proposal becomes a scheduled job, and the job's labor flows into certified payroll — without anyone exporting and re-importing at each seam. The scan is the first link in a chain that's already connected.
Hyperion is a $399/mo module on Forge Core ($499/mo flat) — included in the Working Stack and Full Platform packages, and in the Charter founder program. It is not sold separately. The spatial-intelligence and photogrammetric-LiDAR-augmentation methods behind it are among the 16 provisional USPTO applications assigned to Dominus IP Holdings LLC.