GeoBusinessIQGeoBusinessIQ

8D problem solving: writing an argument a customer will accept

What this answers

Can we show a customer that this defect is contained, understood and stopped, without them having to take our word for it?

A customer asks for an 8D when a defect has reached them and an informal explanation will no longer do. The document is not a form to populate but an argument to win: this is what we contained, this is why it was made, this is why we failed to catch it, and this is the production evidence that the change works. Treated as paperwork, it yields a report nobody believes and a defect that comes back.

Written for: customer quality engineers, manufacturing engineers, supplier quality teams.

Containment is judged on the completeness of the net

The first obligation is to make the customer safe from further defective product, and it is assessed separately from anything you later learn about cause. That means sorting or quarantining every location the affected material could be: finished goods, work in progress, material already dispatched, stock sitting at a distribution point, parts in transit, and inventory held on the customer's own line. Plants routinely lose credibility here by containing the plant and forgetting the pipeline. State clearly what was checked, what was found, and what remains unverified — an honest gap in the net is recoverable, a discovered one is not.

The team determines how good the answer can be

A quality engineer working alone produces a plausible narrative rather than a cause. The people who can actually tell you what changed are the operators who run the process, the setter who last adjusted the tooling, maintenance, and whoever handles the material before and after. Give the group a leader with authority to stop and rerun the process for a trial, and a sponsor who can release the time. The practical failure is assembling this team on paper and then investigating by email, which reproduces the single-author problem while adding names to the signature block.

Two causal chains: why it was made and why it escaped

Every defect that reaches a customer represents two separate failures, and answering only the first leaves the plant exposed. The occurrence chain explains what in the process produced the condition. The escape chain explains why the inspection, gauge, control plan or operator check that should have detected it did not — the check did not exist, was at the wrong frequency, could not physically see the feature, or was performed and the result ignored. Customers with experience read the escape section first, because it tells them whether anything else has been slipping through unnoticed.

Proving the fix before you claim it, then spreading it

Verification means deliberately running the process under the new condition and showing the defect absent, ideally including an attempt to reproduce the original failure with the fix in place. Once proven, the change has to land in the documents that govern the process: the control plan, the work instruction, the setup sheet, the maintenance schedule, the training record. A fix living only in the memory of one shift is undone by the next holiday cover. The read-across step then asks which other parts, tools or lines share the same failure mode and applies the change there before it is needed.

Closure, and the cost of closing early

The customer decides when the report is accepted, and their bar rises with each repeat. A defect that recurs against a closed 8D typically triggers escalation: additional inspection at your cost, containment activity monitored by the customer, restricted access to new business, or an on-site audit at short notice. Weighed against that, holding the report open for another few production runs to gather evidence is cheap. The team should also record what the whole event cost, since that figure is the only reliable argument for the prevention work that stops the next one.

Frequently asked questions

When is an 8D the wrong tool for the problem?
When the cause is already known and trivially fixable, the method adds administration without insight — fix it, record it, move on. It also struggles with chronic low-level losses spread across many causes, where a structured improvement project with data over time works better. The method earns its weight on a discrete, serious event with a customer waiting for an answer and a defect whose mechanism is genuinely unclear.
How quickly does containment have to be in place?
Before the customer has to ask again. In practice the expectation on a safety or functional defect is same-day protection of their line, communicated in writing even if the sort is still running. Containment does not need the cause; it needs a defined boundary of suspect material, a method of identifying good from bad, someone doing the sorting, and clear marking of what has been cleared so nothing gets checked twice or missed entirely.
Our customer rejected our 8D. What is usually wrong with it?
Most rejections come from three places. The cause is stated as an operator error or a supplier issue with no explanation of why the system permitted it. The escape chain is missing entirely, so the report explains the defect but not the detection failure. Or the verification is a statement of intent rather than evidence from production after the change. Adding data to those three sections resolves the majority of returned reports.

Data limitations

  • Standards are referenced, never reproduced. Pages describe what a standard governs and point to the issuing body; they do not restate its requirements, and conformity is determined by the standard itself and by an accredited assessment, not by anything here.
  • Manufacturing figures are operator-supplied inputs, not market data. GeoBusinessIQ holds no factory costs, production volumes, yields, cycle times, tooling prices or capacity data and does not estimate them — every result reflects only the figures you enter.

Explore the graph

Sources

  • NIST Manufacturing Extension Partnership NIST MEP (accessed )
    Covers: A public programme supporting small and medium manufacturers with operational, quality and technology adoption practice.
    Does not cover: Results attributable to any specific manufacturer, or improvement figures transferable to another plant.
    Why it matters: Cited for the operational practice it publishes for smaller manufacturers, not for benchmarks or outcome claims.
    Review cadence: annual
  • United Nations Industrial Development Organization UNIDO (accessed )
    Covers: Industrial development analysis, industrial statistics methodology, and manufacturing capability programmes across member states.
    Does not cover: Company-level data, factory costs, supplier information, or real-time production statistics.
    Why it matters: The United Nations agency for industrial development; used for structural framing of how manufacturing sectors develop, never for point figures.
    Review cadence: annual

Educational and operational information only — not legal, engineering, safety, customs, tax, or financial advice. Requirements vary by jurisdiction, product, process, and contract; confirm with the relevant authority or a qualified professional before acting.

Last updated: