TL;DR
TL;DR: The scanner is the cheap part — operators, processing hours, QA discipline, and utilization are the real cost of ownership. Buy only if you'd scan weekly for internal verification; hire providers for consequential deliverables (models, permits, disputes) and specialty capture. Most GCs land on the hybrid: own quick checks, contract certified work.
Every year more GCs price out a scanner of their own, and the sticker math looks tempting: a capable terrestrial unit or mobile system versus per-project provider fees. But the scanner is the smallest part of the capability. Ownership means operators, training, processing software and hardware, registration QA, data management, and keeping all of it utilized — and the honest answer for most GCs is a hybrid: own the quick-verification capability if volume justifies it, hire out anything that produces deliverables people will stake decisions on. Here's the build-vs-buy breakdown.
What does owning a scanner actually require?
The capture device is step one of six:
- Hardware — the scanner itself, plus targets, control equipment, and the workstation class needed to process point clouds (they are enormous files; see formats).
- Software — capture, registration, and processing licenses, renewed annually.
- Trained operators — capture technique determines data quality, and the learning curve is real. Bad capture looks fine on site and fails in registration.
- Processing time — for every field hour, expect multiples in registration, cleanup, and QA. This is the cost GCs most underestimate.
- QA discipline — accuracy claims mean nothing without control verification and documented residuals. The ±5mm question — "registered, verified how?" — applies to your own crew too.
- Utilization — the machine only pays while scanning. Idle weeks are depreciation without revenue.
When does owning make sense for a GC?
When frequency is high and tolerance for deliverable polish is low. The strong ownership cases:
- Frequent quick verification — daily/weekly layout checks, pour verification, punch documentation where speed matters more than certified accuracy.
- [Progress documentation](/services/progress-scanning) at volume — if your projects want weekly capture and you run many simultaneous jobs, in-house capture removes scheduling friction.
- Self-perform trades with prefab pipelines — mechanical contractors who fabricate to field conditions capture constantly; the scanner is production equipment for them.
If your scanner would run several days a week across a season, ownership math starts working. If it would live in a case between monthly uses, it's an expensive way to feel modern.
When does hiring a provider win?
- Deliverables with consequences. Revit models, floor plans for permits and leases, documentation for disputes — anything stamped, signed, or litigated wants verified registration, documented accuracy, and someone whose whole business is that chain.
- Peak capacity and speed. A provider's crew captures 80,000–120,000 sq ft per field day and hands back registered data in 3–5 business days — without pulling your people off billable work.
- Specialty capture. Above-ceiling MEP, live industrial environments, occupied hospitals — each has a playbook that occasional operators don't develop.
- No utilization risk. You buy outcomes per project; the depreciation clock is someone else's.
The provider checklist covers what to demand when you do hire — registered accuracy verified against control, documented QA, open-format delivery.
What does the hybrid model look like in practice?
The pattern maturing across the industry: GCs own fast, forgiving capture for internal verification, and contract survey-grade capture and modeling for anything downstream teams consume. The two coexist cleanly when a provider establishes the control network and base documentation, and in-house quick scans reference the same coordinates. Your team answers "did the wall land where the model says" on a Tuesday morning; the provider delivers the model the wall was checked against.
How should a GC decide?
Count last year honestly: how many times would the scanner have left its case, and how many of those uses needed certified deliverables versus quick checks? Multiply the quick checks by friction saved; weigh against hardware, software, training, and the processing hours your PMs don't have. If the answer is "we'd scan weekly and mostly for ourselves" — buy, train seriously, and QA like a provider. If it's anything else, put a provider on speed dial and spend the capital on work you self-perform. Fixed quotes in one business day make the per-project math easy to run.