"It's broken" is not a defect report. The conversation that names defects precisely — chips versus crazing, pinholes versus blisters, warping versus distortion — gets them fixed: the cause sits behind the name, and the remedy sits behind the cause. The defect taxonomy is the fleet's shared language: buyer, factory, and inspector reading the same dictionary. This guide builds it.
Why a Taxonomy Matters
The vague defect report fails three ways: the factory cannot fix what it cannot identify (the "quality problem" complaint returns a shrug), the patterns hide (the ten different defects reported as "broken" never sum to a cause), and the disputes stall (the claim without the name is the claim without the clause). The taxonomy's promise: the defect named is the defect halfway to fixed — the report that points to the process stage and the responsible line.
The Body Defects
The body's failures, named:
| | |
|---|
| The line through the body, often hairline | Thermal shock, impact, or the firing's stress |
| The missing fragment at edges and rims | Impact — handling, stacking, the bus tub |
| The piece that rocks, the rim that waves | Firing support, the body's thickness variance |
| The shape that drifted from the mold | Firing temperature or the forming's tolerance |
| The tiny holes in the glaze surface | Firing gases escaping — the glaze's application or the curve |
| The raised bubbles in the glaze | Over-firing or the glaze's thickness |
| The fine crack network in the glaze | The glaze-body fit: expansion mismatch |
| The discoloration that washes won't lift | Porous or crazed surfaces absorbing |
The body defects' reading: each name points to a stage — the forming, the firing, the glazing, or the service. The report that names the defect points at the fix.
The Glaze and Surface Defects
The surface's vocabulary: the glaze defects — the orange peel (the texture that didn't level), the dry foot (the unglazed ring's roughness beyond spec), the color drift (the batch against the golden standard), the speckle's variance (the intentional vs the accidental), and the gloss loss (the wear line or the wash chemistry's verdict). The surface defects' reading: the glaze's chemistry and application are their causes — and the golden standard is their judge. The surface vocabulary is the brand-consistency conversation's language: the drift named is the drift negotiated.
The Structural and Dimensional Defects
The measurement's failures: the dimensional defects — the diameter off (the stack that no longer nests), the capacity's shortfall (the portion spec's miss), the weight's drift (the hand's read), the foot's unflatness (the table's scratch and the wobble), and the handle's alignment (the cup that pours off-center). The dimensional reading: the tolerances' conversation — the spec's numbers, the factory's process capability, and the golden standard's measurements. The dimensional defects are the rubric's lines: scored, thresholded, verdicted.
The Decoration Defects
The brand layer's failures: the decoration defects — the misregistration (the print that slid), the color's shade miss (the batch against the reference), the coverage gaps (the band that thins), the firing's blemishes (the speck in the logo, the scratch through the decal), and the wear's early arrival (the decoration that fades before its season). The decoration reading: the decoration layer's process — the decal's application, the firing's placement, the ink's quality — and the brand's tolerance windows. The decoration defects are the brand's risk line: named precisely, negotiated against the sealed reference.
The Service-Induced Defects
The fleet's own wear, named for the ledger: the service defects — the metal marks (the grey trails from other ware), the knife scoring (the cutting surfaces' patina), the thermal crack (the gradient's verdict), the chip's propagation (the small chip that grew), and the wash erosion (the years' dulling). The service reading: these are not the factory's — they are the fleet's life, and their rates are the attrition forecast's data. The service defects belong in the fleet ledger: named, counted, fed back to the material specs and the handling training.
The Report Format
The taxonomy in use: the report format — per defect: the name (from the taxonomy), the location (where on the piece), the frequency (how many in the sample), the severity (against the threshold), and the photos (the macro that shows it). The format's discipline: one defect per line (the combined reports bury the patterns), the taxonomy's name used exactly (the shared dictionary's discipline), and the batch reference attached (the report joins the batch ledger). The report format is the taxonomy's working clothes: the language in the email, the claim, and the corrective action.
The Corrective Action Loop
The named defect closes its loop: the corrective action — the cause traced (the process stage behind the defect), the fix agreed (the factory's corrective plan, the buyer's verification), and the recurrence watched (the next batches' sampling focused on the fixed line). The loop's discipline: the corrective action logged per defect class (the supplier's scorecard builds), and the recurrence's honesty (the fix that didn't fix is a conversation about capacity, not intention). The loop is the taxonomy's purpose: the defect named, the cause fixed, the fleet protected.
Frequently Asked Questions
Why does naming defects precisely matter? The name points to the stage and the fix: cracks say thermal or firing, crazing says glaze-body fit, pinholes say the curve. The vague report returns a shrug.
What is crazing, and why does it matter? The fine crack network in the glaze — the glaze-body expansion mismatch. It matters because it harbors stains and bacteria: the safety line, not just the look.
Which defects belong to the factory, and which to the fleet? Body, glaze, dimensional and decoration defects are the factory's; metal marks, scoring, thermal cracks and wash erosion are the fleet's service life — the ledger's data, not the claim's.
What does a proper defect report contain? The taxonomy name, the location, the frequency in the sample, the severity against threshold, the macro photos, and the batch reference — one defect per line.
What is the corrective action loop? Cause traced to the process stage, the fix agreed and verified, the recurrence watched in the next batches' sampling — the defect named, the cause fixed, the fleet protected.