This scenario is fictional. The numbers are chosen to show how a buyer and supplier can structure a quality decision; they are not a prediction of legal outcome.
08:00 Monday — the hard data
A U.S. distributor receives 5,000 coated metal components from an overseas supplier. The contract price is $18 per unit, total $90,000. The purchase order incorporates a drawing that requires a coating thickness of 70–90 μm and a named visual reference sample.
The first warehouse check finds visible finish variation in 12 of 80 opened units.
The distributor has a customer production run scheduled in five days. Stopping the line could cost more than the invoice. Using defective components could create returns.
The commercial team has 48 hours to decide whether to:
- release the lot;
- sort locally;
- demand rework/replacement;
- negotiate a credit;
- quarantine and escalate.
09:30 — three test results create three stories
Quality pulls 30 units from three cartons.
Internal gauge: 22 units between 72–88 μm; 8 units between 55–64 μm.
Supplier's retained sample: reported 76–84 μm.
Third-party rapid test on five couriered units: 3 pass, 2 fail at 61 and 63 μm.
The parties now have three stories:
- buyer: “A material part of the lot is off-spec”;
- seller: “Your test method or storage is wrong”;
- third party: “Some failures exist, but five pieces cannot define the whole lot.”
Nobody yet has enough evidence to say 8/30 means 26.7% of all 5,000 units are defective.
10:15 — variable one: what exactly controls?
Procurement produces:
- signed master supply agreement;
- purchase order;
- drawing Rev F;
- approval email for a golden sample;
- a later engineering email mentioning Rev G but not clearly incorporating it.
The supplier says Rev G relaxed the lower limit to 60 μm. The buyer says Rev G was never approved commercially.
This is the first turning point. The dispute is not yet “Did the factory produce bad goods?” It is “Which requirement became contractual?”
The team creates a hierarchy sheet rather than arguing by email.
11:00 — variable two: are the tests comparable?
The buyer's gauge was calibrated last month. The supplier asks about substrate preparation and measurement location. The third-party report uses the same instrument type but a different sampling plan.
A technical lead writes a one-page protocol:
- instrument and calibration check;
- measurement locations per unit;
- number of readings;
- conditioning period;
- carton/batch selection;
- visual-reference procedure;
- how borderline results will be handled.
The buyer and supplier agree to witness a new test remotely.
This does not “settle” the case. It makes the next evidence comparable.
13:30 — variable three: what is the commercial deadline really worth?
Sales says the downstream customer needs 2,800 units in five days, not all 5,000.
Operations estimates local 100% sorting at $0.62 per unit plus two shifts of labor. A local coating shop says selected failed units can be reworked for about $2.40 each, subject to a trial and no guarantee on cosmetic match.
Air-freighting 2,800 replacement units would be dramatically more expensive than ocean replacement and still may miss the line start.
The case changes again. The question is no longer “accept or reject 5,000?” It becomes:
Can we identify at least 2,800 conforming units quickly, preserve evidence on the rest, and negotiate who pays the resulting cost?
15:00 — the first five questions
Before anyone sends a demand, the team answers five questions.
1. What is definitely known?
Some tested units are below 70 μm under the buyer's current method.
2. What is not known?
The true lot-wide failure rate and which revision legally controls.
3. What becomes irreversible in the next 24 hours?
Reworking or consuming units without preserving representative originals.
4. Which decision is time-critical?
Whether enough conforming units can be released for the customer's line.
5. Which legal issue can wait until facts improve?
The final damages number.
That prioritization prevents the lawyers from being asked to answer a statistical question and the laboratory from being asked to interpret a contract.
17:30 — counterexample one: “off-spec” does not automatically mean total rejection
Suppose the repeat inspection shows failures clustered in cartons from one production shift. The remaining cartons test consistently within range.
A whole-lot rejection may be commercially unnecessary and, depending on the contract and governing law, legally disproportionate.
Under the CISG, if applicable, conformity and remedies turn on its requirements and facts; the Convention does not reduce every variance to a binary “reject all” rule.
08:00 Tuesday — repeat inspection
The parties jointly select 120 units across production dates and carton locations.
Results:
- 94 within 70–90 μm;
- 19 between 65–69 μm;
- 7 below 65 μm.
Visual variation is concentrated in the same production date as the lowest readings.
The supplier still disputes the contractual status of Rev F, but it now accepts that the lot is not uniform.
The buyer photographs, labels and sets aside representative pass/borderline/fail samples.
10:00 — counterexample two: a regulatory pass does not erase a contract miss
The supplier produces a compliance certificate for a separate product-safety requirement and argues that the goods are therefore acceptable.
That certificate may be important for regulatory compliance, but it does not answer whether coating thickness meets the purchase specification.
The reverse is also true: contractual non-conformity does not automatically prove a breach of a public technical regulation.
WTO TBT materials help distinguish mandatory technical regulations, standards and conformity assessment at a policy level. The private sales dispute still requires its own contract and governing-law analysis.
11:30 — build the decision model
The team uses four variables:
| Variable | Current estimate | Decision effect |
|---|---|---|
| units needed immediately | 2,800 | defines minimum releasable quantity |
| verified severe failures | 7/120 sample | supports targeted quarantine, not automatic whole-lot figure |
| sorting cost | ~$3,100 plus management | compare with delay cost |
| rework cost | ~$2.40 per selected unit | possible fallback if trial succeeds |
They add two legal unknowns: controlling revision and effect of notice/use under governing law.
13:00 — counterexample three: Incoterms does not answer the quality question
The contract uses CIF to a named port.
One manager says, “CIF means the seller carried the risk, so it is the seller's defect.” That is too simple. Incoterms 2020 allocates defined delivery, cost and risk responsibilities; it does not decide whether factory-applied coating met the sales specification.
If the coating was correct at shipment but damaged in transit, risk allocation matters. If it was manufactured below spec before shipment, conformity analysis is different. Both questions can exist at once.
14:30 — the commercial proposal
The buyer proposes:
- joint accelerated sorting of all 5,000 units;
- immediate release of conforming units;
- quarantine of sub-70 μm units pending final technical review;
- supplier funds sorting and reasonable local rework of a defined subset;
- both sides reserve legal positions on Rev F versus Rev G and any additional loss;
- no destruction or bulk rework of representative evidence until written agreement.
The supplier counters with shared sorting cost and a capped credit.
Now there is something concrete to negotiate.
16:00 — what local law still needs to answer
The team cannot responsibly resolve these points from a generic article:
- which document controls the specification;
- whether the buyer's notice was legally sufficient and timely;
- whether use of conforming units affects rights over the remainder;
- what damages are recoverable;
- whether the seller has a right to cure;
- how limitation/exclusion clauses operate;
- forum and dispute procedure.
For example, the UK Sale of Goods Act has its own national rules around description, quality and sample where applicable. The CISG has a different international framework when it governs. Another domestic law may differ again.
20:00 — the 48-hour decision
The parties do not settle every legal issue.
They agree operationally to sort, release verified conforming units, preserve failed samples and run a rework trial. Commercial teams reserve the remaining money dispute for a five-day review after the full sorting data is available.
That is a good outcome for this stage because it prevents the urgent production problem from being held hostage by the slower legal accounting problem.
It is not a guaranteed outcome, and in a real case either party might reasonably choose a different path.
Before the final decision, finance adds one more control: no one may book a “settlement reserve” from the 7/120 severe-failure figure as though it were the final lot failure rate. The team keeps three numbers separate—observed sample result, current working estimate, and legally claimed quantity. That prevents a provisional technical finding from hardening into a financial or legal assertion simply because it appeared in a spreadsheet first.
The transferable lesson
The useful sequence was:
hard data → controlling document → comparable test method → commercial deadline → reversible action → legal analysis → quantified proposal.
What they did not do was equally important:
- they did not extrapolate 8 failed samples to all 5,000 units without a defensible basis;
- they did not let a regulatory certificate answer a private specification question;
- they did not let CIF answer a manufacturing-conformity question;
- they did not destroy representative evidence while rushing to keep production moving.
Bottom line
A quality dispute becomes manageable when the team separates what must be decided in 48 hours from what can wait five days or five weeks. Use the deadline to prioritize facts, not to invent certainty.
This is a fictional educational scenario, not legal advice or a prediction of case outcome. Inspection, notice, conformity, risk, remedies and damages depend on the contract, governing law and jurisdiction.
Related Reading
- The Goods Arrived, but Are They Legally ‘Wrong’? A Practical Primer on Quality & Specification Disputes
- Before You Reject a Shipment: 18 Decisions in a Quality & Specification Dispute
- The Defect Is Small—Until These 12 Signals Appear: When a Quality Dispute Is Escalating
Sources and Scope Notes
- United Nations Convention on Contracts for the International Sale of Goods (Vienna, 1980) (CISG) — UNCITRAL; checked 2026-10-03. Boundary: Applies only when its scope and connecting rules are satisfied; parties may exclude or vary its application subject to applicable law. It is not a universal sales law.
- Sale of Goods Act 1979 — legislation.gov.uk; checked 2026-10-03. Boundary: UK statute used only as a concrete national-law example. Scope, amendments and transaction type must be checked locally.
- Incoterms® 2020 — International Chamber of Commerce; checked 2026-10-03. Boundary: Contractual trade rules allocating defined delivery, cost and risk responsibilities when incorporated. They do not replace the whole sales contract or decide product-quality law.
- Technical Barriers to Trade — World Trade Organization; checked 2026-10-03. Boundary: Intergovernmental framework concerning technical regulations, standards and conformity-assessment procedures; it does not decide a private quality claim by itself.