Studio open4 Products in MarketMentziAurumOSDepoZerocodeAICTO Partner — VDOlocal5 Patents FiledBuilt for BharatBooking Q3–Q4 2026Studio open4 Products in MarketMentziAurumOSDepoZerocodeAICTO Partner — VDOlocal5 Patents FiledBuilt for BharatBooking Q3–Q4 2026
AurumOSCase Study

Gold Loss Analytics: How One Mumbai Workshop Found ₹8 Lakh in Hidden Losses

A 22-karigar Zaveri Bazaar workshop discovered ₹8 lakh in annual gold loss hiding in plain sight. Here is exactly how gold loss analytics surfaced it.

Sitio Labs Team8 min read4 topics

A workshop that thought it had no problem

In the lanes off Zaveri Bazaar, a family-run workshop with 22 karigars had been operating for three generations without a single recorded gold dispute. The owner believed his wastage was "normal" — the 3 to 4 percent that every jeweller in Mumbai quietly accepts as the cost of doing business. At a gold price of roughly ₹73,000 per 10 grams in late 2024, that acceptance was quietly draining his margins. He agreed to a 90-day pilot mostly to humour a younger cousin who had heard about gold loss analytics.

Where the gold actually disappears

Gold loss in a jewellery workshop is never one big leak — it is a hundred small ones. Polishing dust on the floor, filing residue trapped in workbench cloth, solder splatter, and the unmeasured difference between gold issued to a karigar and gold returned in finished pieces all add up. Traditional workshops reconcile this on a weekly tola basis with handwritten registers, which means a 0.2 gram discrepancy per job card simply vanishes into rounding. Across 4,000 job cards a year, that rounding alone hid nearly 800 grams.

What the data revealed in 90 days

Once every gold issue and return was logged at the job-card level, three patterns emerged that no register could show. One bench consistently returned 0.4 percent less than the workshop average, traced to a worn polishing motor throwing excess dust. A specific 22-karat chain design lost twice the gold of comparable items because of an inefficient soldering step. And reconciliation gaps clustered on Saturdays, when the senior karigar who weighed returns left early. Together these accounted for ₹8.1 lakh in annual loss — recoverable, not theft.

Trust, not surveillance

The owner was clear that he would not introduce anything that made his karigars feel accused. Gold loss analytics works precisely because it depersonalises the conversation — the data points to a motor and a process step, not to a person. When the senior karigar saw that his bench was flagged for equipment, not dishonesty, he became the loudest advocate for the system. In an industry built on inter-generational trust, this distinction between measuring process and policing people is the entire game.

The takeaway for India’s jewellery MSMEs

India’s jewellery manufacturing sector loses an estimated thousands of crores annually to unmeasured wastage, most of it in workshops with fewer than 50 karigars who lack any analytics layer. The fix is rarely capital expenditure — it is visibility. A workshop that knows its loss per bench, per design, and per day can recover one to two percent of throughput within a quarter. For a Zaveri Bazaar unit, that recovered percentage often exceeds the entire annual profit of a slower year.

Frequently Asked Questions

How much gold do jewellery workshops typically lose?

Most Indian jewellery workshops lose between 2 and 4 percent of gold throughput to wastage, dust, soldering, and reconciliation gaps. At current gold prices, even 3 percent loss in a mid-size Zaveri Bazaar workshop can exceed ₹8 lakh per year.

What is gold loss analytics?

Gold loss analytics is the practice of logging gold issued and returned at the individual job-card and bench level, then analysing the discrepancies to find recoverable loss. It pinpoints which design, machine, or process step is wasting gold rather than blaming individuals.

Is gold loss usually due to theft?

No. In the vast majority of cases gold loss is process-related — worn polishing equipment, inefficient soldering, uncaptured dust, and weekly rounding errors. Analytics surfaces these mechanical and procedural causes, which are recoverable without confrontation.

How long does it take to find hidden gold loss?

A typical workshop sees clear patterns within 60 to 90 days of logging gold movement at the job-card level. That window is enough to capture variation across benches, designs, and weekdays.

Will karigars resist gold loss tracking?

Resistance drops sharply when the system is framed around process, not surveillance. Because analytics flags equipment and procedure rather than people, senior karigars often become its strongest supporters once they see their bench cleared of suspicion.

More from AurumOS