Troubleshooting

Common Mistakes in Jewelry and Watch Production and How to Catch Them

The recurring mistakes that quietly bleed margin in precious metal and precision luxury work, each paired with its symptom, root cause, and a numeric fix.

The costliest mistake in precious metal work is mixing fine-metal and alloy-metal weights in the same yield check. Symptom: your Precious Metal Yield reads a suspicious 58 percent when the run looked clean. Root cause is weighing finished pieces in 14K alloy grams (0.585 fine) while charging in fine grams, or the reverse. The fix is one rule: measure both the finished weight and the charged weight on the identical basis. Convert with the karat factor before dividing, so 210 alloy grams against 250 alloy grams reads 84 percent, not a phantom number that sends you chasing losses that never happened.

Counting sprues and buttons as finished product is the second yield trap. Symptom: yield jumps to 96 percent yet your scrap bin keeps filling. The buttons, sprues, and gates got weighed as saleable metal because they had not been cut from the tree yet. A gold tree's runner system alone commonly carries 20 to 35 percent of poured weight. Trim every piece from the tree before you weigh, and log runner metal as recoverable scrap headed to the refiner. Feed that scrap weight into Scrap Recovery Value so the metal is tracked, not double-counted as both yield and recovery.

Treating rack fill as a proxy for cost torpedoes plating quotes. Symptom: a 40-piece rhodium run that should cost roughly $3.50 a piece settles at over $6. The bath's fixed maintenance, filtration, and analysis cost, say $120, does not shrink when the rack runs half empty. On 200 pieces that $120 spreads to $0.60 each; on 40 pieces it becomes $3.00 each. Run Plating Bath Cost on the actual piece count before quoting, and set a minimum economical batch where per-piece cost stays within about 15 percent of the variable rate.

Applying one blended setting rate across mixed goods wrecks setting labor estimates. Symptom: a pave-heavy order booked at 40 hours consumes 62. The default 3 stones per minute suits clean melee, but micro-pave drops below 1 per minute and fancy-cut bezels slower still. Setting 500 stones at a blended 3 per minute predicts 2.8 hours; if half are micro-pave at 0.8 per minute, the real figure is closer to 6 hours. Split every mixed job in Stone Setting Labor by setting type and run each segment on its own rate before summing.

Skipping the sorting and rework allowance is a quiet 15 percent underquote on both setting and polishing. Symptom: bench schedules that always run over, and jobs that look profitable on paper but never are. Base setting time of 40 hours ignores stone grading, chipped-goods swaps, and resetting crooked stones. Always quote from scheduled labor, not base time: a 15 percent allowance turns 40 hours into 46. For colored or included stones that break more often, raise the allowance to 20 or 25 percent rather than pretending the diamond rate holds.

Assuming metal below your target yield is lost forever leads shops to over-price scrap loss. Symptom: cost accounting books a 6 point yield gap as a 6 percent hard material loss. In reality sprues, filings, and polishing sweeps refine back at 90 to 98 percent for clean scrap and 85 to 92 percent for sweeps and dust. The true cost is refining fees plus tied-up working capital, not the metal. Run Scrap Recovery Value on the segregated scrap so the recoverable value offsets the yield gap instead of being written off twice.

Overloading the casting tree to chase utilization backfires into porosity rejects. Symptom: tree fill hits 100 percent but 3 or 4 pieces per flask come out with shrinkage porosity or misfill. Feed channels get starved when patterns crowd the sprue. Utilization above 90 percent on heavy pieces often trades a few percent of overhead savings for double-digit reject rates. Use Casting Tree Utilization with an 85 percent target for standard pieces, and read it alongside Finishing Defect Cost so the reject cost of the last two patterns is weighed against the flask savings they bring.

Leaving traceability, inspection, and repair reserve out of the number is the last common miss. Symptom: certified serialized pieces that priced fine still lose money after shipping. Serial engraving and provenance logging via Serialized Item Traceability can add several dollars per unit, magnified inspection via Inspection Magnification Workload adds bench minutes, and warranty work lands after the sale. If 3 percent of pieces need a $45 repair, Repair Reserve sets aside about $1.35 per unit. Bury none of these in overhead; each belongs as a line on the quote.

Published 2026-07-01.