ORDREL / LEARN · ENGLISH
5 Whys in Manufacturing: Follow Evidence, Not a Script
Use 5 Whys to develop and test a causal explanation of a specific problem. Each link needs evidence. Five is a prompt, not a required number of answers, and a completed worksheet does not prove a root cause.
State the observed gap
Separate the event from your theory. Record what was unacceptable, where it occurred, the time window, the comparison basis and what remains unknown. “Carelessness” is not a useful starting observation. Where immediate containment is needed, follow the site’s existing escalation process while the investigation continues.
Build the explanation with the people who know the process. Review samples, records and the actual method. A plausible story becomes stronger only when observations distinguish it from alternative explanations.
A fictional packaging investigation
Fictional manufacturing teaching example. These figures are not ORDREL customer results.
A team finds 12 cartons with the wrong instruction leaflet among 400 inspected cartons in one defined run. The observed fraction is 12 ÷ 400 = 3%. This defines the inspected sample; it does not establish the rate in uninspected stock.
| Working explanation | Evidence to check | What would weaken it |
|---|---|---|
| The wrong leaflet variant reached the packing station. | Retained cartons and leaflet identification records. | The leaflets were correct when packed and changed later. |
| Two visually similar variants were staged together. | Contemporaneous staging records and operator observations. | Only the correct variant was present during that run. |
| The changeover check did not verify leaflet identity. | The actual instruction and completed changeover record. | The check required identity and was performed correctly. |
| The check was designed around quantity, not variant matching. | Review the approved method and its change history. | A current variant check exists but a different failure occurred. |
These are fictional hypotheses and checks, not a proven chain. Stop when evidence stops. Branch if more than one mechanism fits. “Train everyone” is not a finding, and “the operator failed” does not explain why the process permitted the error.
Specify a countermeasure that can be tested
If the evidence supports a variant-matching gap, propose a check between the work requirement and leaflet identity before release. Define who performs it, what counts as a mismatch and the existing escalation route. Verify that the check works with the actual materials and identifiers; do not assume a written instruction implements a technical interlock.
A trial should include a matching variant, a deliberately mismatched test item and an unreadable identifier under controlled, approved conditions. Record expected and actual outcomes. Preserve product controls and do not introduce defects into saleable production to run the exercise.
Review the result without overstating it
In a fictional follow-up sample, two wrong-leaflet findings occur among 400 inspected cartons: 0.5%. The difference from 3% is 2.5 percentage points, an 83.33% relative reduction in these sample proportions. Neither figure proves the countermeasure caused the change. Check product mix, sampling, staffing, detection and other concurrent changes.
Assign an owner, review date and an outcome measure before the trial. If the problem continues, revisit the explanation rather than editing the worksheet until it appears successful.
Use ORDREL as the investigation record
Open 5 Whys investigation to record the problem, evidence, why links and follow-up. Use Fishbone if several cause families need parallel investigation. The separate 5 Whys Word playbook is a one-time paid file; it is not required to use the free on-screen tool.
- Do I need exactly five links? No—use only the supported links needed to explain the mechanism.
- Can there be several causes? Yes; record dependencies and alternatives.
- Does a cause become verified because everyone agrees? No. Agreement should be tested against evidence.
Sources and method notes
Sources support the method definitions. Examples, interpretation and tool instructions are ORDREL teaching material, not an endorsement by the referenced organizations. See methods and corrections.