The instrument may be functional, have a minor fault, require expensive repair, or be missing the component that makes the system useful. Do not translate “untested” directly into either “probably working” or “parts only.” Ask what the evidence establishes, what failure modes remain possible, and what those unknowns should do to the bid ceiling.
“As-is” should change the underwriting
Auction terms vary, but “as-is” is not a tested-working warranty. Current official GSA guidance says GSA does not guarantee sale-item condition and encourages inspection before sale. Its General Sale Terms and Conditions say property is offered “as is” and “where is” unless the invitation says otherwise, and urge bidders to inspect before bidding.
- GSA guidance for citizens seeking surplus property
- GSA Standard Form 114C, General Sale Terms and Conditions (PDF)
This does not mean every GSA instrument is defective. It means a buyer prices the evidence that exists rather than assuming a warranty that does not. See the GSA reseller due-diligence guide.
Use a condition evidence ladder
1. Visually present
Photos establish presence, not operation.
2. Visually complete
Major visible components appear present; function is still unproven.
3. Powers on
The unit receives power and responds, but the evidence is usually narrow.
4. Self-test or limited functional evidence
A particular diagnostic or operation is shown; the exact test matters.
5. Seller-reported operational
Stronger than silence, but not an independent technical inspection.
6. Last known working
A historical state that may not account for storage, transport, batteries, seals, or age.
7. Functional test
Relevant functions were tested recently and deeply enough to support a stronger case.
8. Calibration evidence
Current documentation can materially improve confidence when calibration matters.
A broad phrase such as “working” should not erase these distinctions.
Powers on is not the same as works
Oscilloscope
Power-up does not establish channel accuracy, bandwidth, trigger behavior, probes, or calibration.
Spectrum or network analyzer
A booting display does not establish RF paths, source, ports, dynamic range, calibration, or options.
HPLC
Powered modules do not establish pumps, detectors, seals, flow, autosampler operation, or a usable stack.
Mass spectrometry
Boot state does not establish vacuum, detector/source condition, software, workflow, or calibration.
XRD or XRF
Power does not establish source, detector, motion, interlocks, software, or workflow completeness.
Vacuum equipment
Motor response does not establish vacuum performance, seals, contamination, or controller compatibility.
Build an as-is resale scenario
Blind/as-is resale
What the equipment conservatively supports without new favorable condition evidence.
Tested-working resale
What the equipment might support after relevant testing establishes useful operation.
The blind decision should not use the tested-working value automatically. Account for the cost and risk of moving between those states.
Add a testing and repair reserve
Testing can require technical labor, fixtures, probes, cables, consumables, standards, pumps, service software, outside technicians, calibration, replacement accessories, or shipping to a provider. Category-specific repair exposure should prevent the model from assuming every untested subsystem works.
Do not double-count uncertainty
Condition risk can appear as lower resale, a repair reserve, a risk buffer, or a higher required return. Use enough conservatism to represent downside without mechanically penalizing the same unknown several times. Keep assumptions visible.
Condition and configuration interact
A unit can be operational but incomplete; a complete-looking system can be untested; a tested mainframe can contain unknown modules. Model condition and configuration separately. See the guide to configuration and missing modules.
Decide what evidence could change the bid
- Can the seller confirm an option code or show the detector?
- Does the unit complete a relevant self-test?
- Are the controller, software, license media, cables, probes, or fixtures included?
- Can an inspection be arranged?
- Can a local contractor quote packing and freight?
A REVIEW can be rational when one answer changes the economics. If uncertainty cannot be resolved or bounded, the blind/as-is ceiling may be zero.
Condition evidence in real auction snapshots
The Bertin Precellys 24 analysis keeps visual completeness distinct from a loaded functional test.
The Labomed UVD-2950 analysis preserves limited power evidence without treating it as proof of full performance. Risk can sometimes be priced; it is never described as risk-free.
The Simbionix LAP Mentor III analysis separates last-known-working history from current software and haptic verification.
When to skip—or pursue—an untested lot
Stacked uncertainty in identity, configuration, market support, freight, failure exposure, parts, software, and liquidity can make a cheap lot unattractive.
An as-is lot can still work when identity and configuration are strong, resale evidence is useful, price is well below the conservative ceiling, testing and repair can be bounded, logistics are known, and the buyer has category expertise. Untested status is an input to the economics, not an automatic outcome.
As-is diligence checklist
Evidence
- What has been demonstrated, when, and by whom?
- Is there test output or only a photo?
Failure exposure
- Which expensive subsystems could fail?
- Can they be tested and are parts or service available?
Configuration
- Is the full workflow present?
- Could one missing module dominate the economics?
Logistics
- Could handling cause damage?
- Is packing responsibility clear?
Resale
- What are realistic as-is and tested-working exits?
- What cost and risk separate them?
Bid
- What reserve and ceiling are defensible?
- Which fact would make the ceiling rise or fall?
Continue with the complete acquisition framework or the maximum-bid mechanics.
Evidence in practice