How to Evaluate a 3D Laser Scanning Provider

TL;DR

• A 3D laser scanning engagement is really three separate decisions — hiring a capture crew, buying a scanner, or outsourcing modeling — and confusing them is the most common procurement mistake.
• Set the spec (accuracy target, LOD by discipline, point density, control tie, coordinate system, deliverable formats) before contacting a single provider, not after.
• Six categories of questions separate a defensible bid from a vague one: accuracy and method, deliverables, software fit, capacity and logistics, commercial terms, and evidence.
• A weighted scoring matrix, not a gut-feel comparison, keeps a multi-bid review honest and defensible to a client or partner who asks how the decision was made.
• Zealot Reality Capture wrote this guide and sells the service it describes; the disclosure and the "when to pick someone else" section below are meant to be read, not skipped.

Decide what you're buying first

Three different procurement decisions get treated as one, and that confusion is where most bad scans of work start. A firm can hire a capture crew for a single project, buy a scanner and bring capture in-house, or outsource the entire modeling pipeline from raw scan data to a finished Revit model. Each is a different vendor conversation, a different cost structure, and a different set of questions to ask.

A firm evaluating "3D laser scanning providers" without first picking one of these three lanes ends up comparing quotes that aren't actually comparable — a scanner lease quote against a full-service scan-to-BIM proposal, for instance. For a fuller breakdown of the buy-versus-hire tradeoff, see Should GCs Buy a Scanner or Hire.

Set the spec before you shortlist

The single highest-leverage step in this process happens before any provider is contacted: writing down what "accurate enough" and "complete enough" mean for this specific project. Providers price and staff differently once the spec is explicit, and a vague RFP produces vague, unequal bids.

At minimum, the spec should state:

• Accuracy target and confidence level. Architectural work is commonly evaluated against a registered accuracy figure such as ±5 mm; industrial and process environments often carry a tighter figure such as ±6 mm. State which applies and how it will be verified.
• LOD by discipline. A single LOD number applied to an entire multi-discipline model hides the fact that MEP, structure, and architecture rarely need the same level of development. LOD 300 model faces to registered cloud is commonly held to a tolerance such as ±10 mm; state that tolerance per discipline, not once for the whole model.
• Point density. Density requirements differ for a corridor walk-through versus a mechanical room needing pipe-run verification.
• Control tie. State whether the scan needs to tie into a known coordinate system or site control network, and who is responsible for establishing that control.
• Coordinate system. Local, state plane, or project-specific — decided in advance, not discovered at handoff.
• Deliverable formats. Registered point cloud, Revit model, CAD, or all three, and in what file formats (E57, RCP, IFC, RVT).
• Occlusion allowance. Every real building has areas a scanner cannot see cleanly — furniture, stacked material, active machinery. State how those areas will be handled and disclosed rather than silently modeled from assumption.

A provider's response to this written spec is itself diagnostic: a provider who pushes back with sharper questions is engaging with the project; a provider who simply quotes a price is not. For the full procurement document built around these eight elements, see Scope of Work Template for Scan-to-BIM.

The evaluation questions, with acceptable answers

Every question below comes with a green (accept), yellow (probe further), and red (walk away) answer. Reading a provider's answers against this key, rather than against a general impression, is what turns a sales conversation into a procurement decision.

Accuracy and method

"What control tie does the scan use, and who establishes it?"

• 🟢 Green: A defined method (survey control points, GPS-tied targets, or existing site control) is named, and responsibility is assigned in writing.
• 🟡 Yellow: "We usually just use our own targets" with no mention of tying to an external control network — acceptable for some interior work, not for anything needing to align with existing survey data.
• 🔴 Red: No answer, or "our scanner is accurate enough on its own" — a scanner's internal accuracy is not the same as registered, control-tied accuracy on a real site.

"What registration method do you use, and what's the reported residual?"

• 🟢 Green: A specific method is named (cloud-to-cloud, target-based, or control-tied) and the provider reports a numeric registration residual on delivery.
• 🟡 Yellow: A method is named but residuals aren't reported by default — ask for one to be included in the deliverable.
• 🔴 Red: "We don't really track that" or an unwillingness to produce a registration report at all.

"Will the accuracy figure include specific conditions?"

• 🟢 Green: The provider states an accuracy figure with a defined qualifier — for example, a registered accuracy figure tied to a specific method and building type.
• 🟡 Yellow: A figure is given with no qualifier, but the provider clarifies readily when asked.
• 🔴 Red: A "sub-millimeter" accuracy claim for terrestrial LiDAR capture with no qualification — see Red Flags below.

"How is accuracy verified after the fact, not just claimed before the job?"

• 🟢 Green: Independent check points, spot measurements against known dimensions, or a written registration report accompany the deliverable.
• 🟡 Yellow: Verification happens only on request or only for larger projects.
• 🔴 Red: No verification process exists at any project size.

Deliverables

"What file formats are included, and is that negotiable later?"

• 🟢 Green: Multiple formats (point cloud plus native BIM plus a neutral exchange format) are listed as standard.
• 🟡 Yellow: One primary format is standard, with others available at extra cost — acceptable if disclosed upfront.
• 🔴 Red: IFC-only export, or any deliverable structure that locks the client out of the native modeling software.

"What LOD applies to which discipline, in writing?"

• 🟢 Green: LOD is broken out by discipline (architecture, structure, MEP) in the proposal itself.
• 🟡 Yellow: LOD is discussed verbally but not written into the contract.
• 🔴 Red: One LOD number quoted for an entire multi-discipline model.

"Who owns the source scan data, and is it delivered or held?"

• 🟢 Green: Raw and registered point cloud data is delivered to the client as part of the engagement, with ownership stated in the contract.
• 🟡 Yellow: Source data is available on request, potentially at additional cost.
• 🔴 Red: Source data is never released, locking the client into that provider for any future rework.

"Can a sample deliverable be reviewed before the full contract is signed?"

• 🟢 Green: A comparable sample from a similarly scoped past project is provided.
• 🟡 Yellow: No sample exists, but references who can speak to deliverable quality are offered.
• 🔴 Red: No samples and no references — an unverifiable deliverable.

Software fit

"Will the model be delivered in project-correct Revit coordinates?"

• 🟢 Green: The provider confirms the model will be placed in the firm's shared coordinates or project base point, not an arbitrary internal origin.
• 🟡 Yellow: Coordinates are placed generically and require the firm's own rework to align.
• 🔴 Red: No process for coordinate alignment at all.

"Does the deliverable match our firm's Revit template and modeling standards?"

• 🟢 Green: The provider asks for the firm's template and standards document before modeling begins.
• 🟡 Yellow: A generic template is used, with revisions available afterward.
• 🔴 Red: No willingness to conform to firm-specific standards under any circumstances.

"How does the point cloud hand off into Navisworks or a coordination platform?"

• 🟢 Green: A defined export path and file structure for Navisworks (or the firm's coordination tool of choice) is described.
• 🟡 Yellow: It's possible but not a routine part of the workflow.
• 🔴 Red: The provider is unfamiliar with coordination-platform handoff entirely.

"Has this provider worked inside our specific software stack before?"

• 🟢 Green: Named prior projects used the same software combination.
• 🟡 Yellow: Different software, but a clear technical bridge is described.
• 🔴 Red: No prior experience and no plan for the gap.

Capacity and logistics

"Are field crews in-house or subcontracted?"

• 🟢 Green: In-house crews are named, or subcontracted crews are disclosed with the same QA standards applied.
• 🟡 Yellow: Subcontracted with limited visibility into QA — acceptable for lower-stakes work.
• 🔴 Red: Subcontracted with no disclosure until after signing.

"What states or regions does this team actually cover?"

• 🟢 Green: A clear, honest coverage area is stated, including any distance-based cost implications.
• 🟡 Yellow: Coverage is described as "anywhere" with no detail on travel logistics or cost.
• 🔴 Red: Vague geographic claims that don't hold up when pressed for specifics.

"What experience exists capturing occupied or live facilities?"

• 🟢 Green: Specific prior work in occupied buildings or live operating environments is named, with a description of how disruption was minimized.
• 🟡 Yellow: Some experience, but not in comparable facility types.
• 🔴 Red: No occupied-facility experience and no plan for minimizing disruption.

"What insurance coverage applies to on-site work?"

• 🟢 Green: Coverage is disclosed clearly and in writing, matched to the project's requirements.
• 🟡 Yellow: Coverage exists but documentation takes repeated requests to produce.
• 🔴 Red: Coverage details are withheld or vague under direct questioning.

Commercial

"Is pricing fixed or time-and-materials?"

• 🟢 Green: The pricing structure is stated clearly, with the basis (square footage, complexity, deliverable scope) explained.
• 🟡 Yellow: A structure is offered but the basis for the number isn't explained.
• 🔴 Red: No clear pricing structure, or a number with no stated basis at all.

"What specifically triggers a change order?"

• 🟢 Green: Change-order triggers are listed in the contract — added square footage, added discipline, occlusion beyond a stated allowance.
• 🟡 Yellow: Change orders are mentioned generally without specific triggers.
• 🔴 Red: No change-order language at all, leaving scope creep undefined.

"Is there a rush premium, and how is it calculated?"

• 🟢 Green: A stated rush premium and the conditions that trigger it.
• 🟡 Yellow: Rush work is possible but priced case by case with no stated formula.
• 🔴 Red: Rush requests are refused outright or priced unpredictably after the fact.

"Who bears the cost if a re-scan is needed due to provider error?"

• 🟢 Green: Re-scan liability for provider-caused error is addressed directly in the contract.
• 🟡 Yellow: Addressed informally, not in writing.
• 🔴 Red: No liability language, leaving the client exposed to paying twice.

Evidence

"Can this provider name comparable projects, not just categories of work?"

• 🟢 Green: Specific named projects, with details a client could independently verify.
• 🟡 Yellow: General project types described without names or specifics.
• 🔴 Red: Only generic claims — "we've done hospitals, plants, and campuses" — with no names attached.

"Do the numbers on this provider's own materials agree with each other?"

• 🟢 Green: Accuracy figures, turnaround times, and coverage claims are consistent across the proposal, website, and case studies.
• 🟡 Yellow: Minor inconsistencies that are explained when raised.
• 🔴 Red: Contradictory numbers across a provider's own materials — a sign the figures aren't being tracked carefully anywhere.

"Are references available, and will they actually speak?"

• 🟢 Green: References are offered proactively and respond when contacted.
• 🟡 Yellow: References exist but take effort to obtain.
• 🔴 Red: No references offered, or references who won't confirm basic project facts.

"Does this provider disclose bias in its own comparison content?"

• 🟢 Green: Any published comparison or "best providers" content discloses that the publisher is also a vendor.
• 🟡 Yellow: Bias isn't disclosed but is obvious from context.
• 🔴 Red: A roundup or comparison where the author ranks itself first with no disclosure at all — see Red Flags below.

A weighted scoring matrix

Comparing bids by gut feel produces defensible-sounding decisions that fall apart under scrutiny. A weighted matrix — scoring each provider 1 to 5 against each category, then multiplying by a category weight — turns six categories of qualitative answers into a single comparable number, and gives a documented basis for the decision if a client or partner asks how it was made.

These weights should shift by project type. A live process plant with 143 pipe runs that don't match the record drawings weights capacity and logistics higher, because the field team's ability to work around live operations matters as much as the modeling accuracy. A historic building with tight tolerance requirements for adaptive reuse weights accuracy and method higher. There is no universal weighting — the matrix is a structure for making the tradeoffs explicit, not a fixed formula.

Red flags

Some patterns should end a conversation immediately, regardless of how strong the rest of a proposal looks:

• An accuracy claim stated without conditions. A number with no method, no confidence level, and no qualifying context is not a specification — it's a marketing line.
• "Sub-millimeter" claims for terrestrial LiDAR. Terrestrial laser scanning does not operate at sub-millimeter registered accuracy on real buildings under real field conditions; a claim at that level should be treated as disqualifying on its face.
• A single LOD number applied to an entire multi-discipline model. Architecture, structure, and MEP rarely need — or get — the same level of development from the same capture pass.
• No registration report offered with the deliverable. If accuracy can't be shown numerically after the fact, it wasn't verified during the project.
• IFC-only export. A deliverable structure that locks the client out of native BIM software formats limits future flexibility and should be disclosed and negotiated, not discovered after signing.
• Contradictory numbers across a provider's own site. Different accuracy or turnaround figures on different pages of the same provider's materials suggest the numbers aren't being tracked carefully anywhere internally.
• Roundups where the author ranks itself first, without disclosure. A "best providers" or "top companies" list published by a vendor, that happens to rank the publisher first, with no disclosure of that conflict, is not a comparison — it's an advertisement dressed as research.

Our own bias, stated

Zealot Reality Capture wrote this guide, and Zealot sells 3D laser scanning and scan-to-BIM services. That should be stated plainly rather than left for a reader to discover later, and it should factor into how much weight this guide carries relative to an independent source.

That said, a bias disclosed is different from a bias hidden. Every question, acceptable answer, and red flag above applies equally to Zealot's own proposals — a reader is encouraged to hold this content to the same standard it recommends holding any other provider to, including asking Zealot the same questions listed here.

There are also situations where a different provider is a better fit than Zealot:

• Projects outside the geographic range a provider can reasonably service without excessive travel cost.
• Projects requiring a specialty method — deformation monitoring at extremely high frequency, for instance — outside a provider's stated capability.
• A firm that has already standardized on a different provider's software ecosystem or file structure, where switching costs would exceed the benefit of changing vendors.
• Projects where an existing, verified relationship with another provider already meets every question above at a green level — there's no reason to re-run a vetting process that's already been satisfied.

The goal of this guide is a better-informed buyer, not a guaranteed sale. For a shorter, checklist-style version of this same evaluation process, see Choosing a 3D Laser Scanning Provider: A Buyer's Checklist, and for a broader look at the national field, see How to Evaluate 3D Laser Scanning Services in the United States.

Related reading

• Vet Point Cloud Accuracy Before You Hire
• What Does ±5mm Accuracy Actually Mean?
• 10 Things to Know About 3D Laser Scanning Services
• 3D Laser Scanning Scope of Work Template for Scan-to-BIM
• How Much Does a Scan-to-BIM Revit Model Cost?
• /services/scan-to-bim
• /services/building-3d-laser-scanning
• /services/industrial-3d-laser-scanning