A 500-stone parcel arrived for a pavé programme. The first tray matched the master set. The fourth tray was visibly warmer, and by the last tray the setter was rejecting one stone in six. The parcel had been packed in order of sorting, and the drift had accumulated silently from the top of the box to the bottom.

Loose moissanite stones spread across a sorting tray for parcel inspection

Parcel consistency is a specification problem more often than a supplier problem. A buyer who specifies a grade and nothing else has specified a range without limits, and a range without limits is what arrives.

Why parcels drift

A parcel is assembled from multiple growth runs. Each run has a slightly different colour and a slightly different proportion tendency, and sorting assigns stones to grades by comparison against a master set.

The three sources of drift

  • Run-to-run colour variation within the same nominal grade
  • Proportion variation from different cutting fixtures
  • Packing order, which can concentrate the outliers at one end

The third source is the most avoidable and the most commonly missed. A supplier who packs in sorting order produces a box that starts clean and ends warm, and a buyer who samples only the top never sees it.

Specifying tolerance bands

A grade without a band is a word. A grade with a band is a specification.

AttributeCommercial bandPremium band
ColourHalf grade either sideQuarter grade either side
Diameter+/- 0.10 mm+/- 0.05 mm
Depth+/- 0.15 mm+/- 0.10 mm
Carat per stone+/- 8 percent+/- 4 percent

State the band in the purchase order and state the consequence of a breach. A band without a consequence is a preference, and preferences are not enforced at packing.

How the consequence should read

A workable clause: stones outside the stated band are rejected and replaced at the supplier's cost, or accepted at a stated discount per stone. Both outcomes are clear, and either one gives the buyer a remedy.

The sampling plan

Sampling is the only way to know what arrived without inspecting 500 stones.

Draw across the parcel

Take 10 percent, and take it from the top, the middle and the bottom of the box. If the parcel is packed in trays, take stones from each tray rather than from the first tray only.

Measure, do not eyeball

Measure 20 stones with a gauge and record the diameters. A 0.1 mm drift across 20 stones is a band failure that the eye reads as normal variation.

Grade under controlled light

Grade the sample against the master set under daylight-equivalent light at 5,000 to 6,500 K. A warm bench lamp will make a half-grade drift invisible, which is exactly the drift that ruins a matched halo.

  1. Take 10 percent across the parcel depth
  2. Measure 20 stones with a gauge
  3. Grade the sample against the master set under daylight
  4. Record the pass and fail counts by tray
  5. Photograph the failing stones next to a passing stone

The fifth step turns a complaint into evidence. A photograph of a warm stone beside a matching stone is the document that supports a replacement claim.

What drift costs in production

Drift is not a cosmetic issue. It changes setting time and it changes the finished look.

Setting time

A halo with a half-grade colour drift shows a visible transition across the ring. The setter either re-sorts the stones by hand, which adds hours, or sets them as they arrive, which produces a ring the customer notices.

Finished look

Colour drift is more visible in moissanite than in diamond at the same grade spread, because moissanite's body colour reads more strongly under daylight. A parcel that passes a diamond-style tolerance can still fail visually in moissanite.

Reducing drift at the source

Three purchasing practices reduce drift before it reaches the bench.

Buy from one run

Ask whether the parcel comes from a single growth run. A single-run parcel drifts less than a parcel assembled across runs, and the premium is usually small at 500 pieces.

Ask for pre-packing

Ask the supplier to pack in mixed order rather than sorting order. A mixed pack spreads the outliers across the box, which does not fix drift but does prevent a buyer from missing it in a top sample.

Set a re-order reference

Keep 20 stones from an approved parcel as a reference set. A re-order can then be matched against the physical reference rather than against a description, and the matching is decisive.

Parcel cost math

A parcel price per carat hides the yield question. The number that decides profitability is the cost per usable stone.

Working the numbers

Take 500 stones at 0.06 ct each, 30 ct total at 45 USD per carat, or 1,350 USD. If 6 percent fail the colour band and 4 percent fail the diameter band, 50 stones are unusable and the cost per usable stone rises from 2.70 USD to 3.00 USD.

A parcel at a higher price with a tighter band can be cheaper per usable stone. This is the same arithmetic as melee sorting loss, applied to moissanite, and it is the reason two quotes at different prices are not necessarily different in cost.

What to ask before ordering

  • The expected reject rate for the stated band
  • Whether rejects are replaced or credited
  • The count and weight reconciliation method

A supplier who can answer the first question has measured their own yield. A supplier who cannot is quoting a price without knowing its cost.

Re-order rhythm

Consistency between orders matters as much as consistency within one. A second parcel that drifts from the first breaks a programme that the first parcel supported.

The reference set

Keep 20 stones from the approved parcel in a labelled box. Every re-order is matched against that box, not against a description. Colour memory is unreliable, and a physical reference settles a disagreement in seconds.

Order frequency

For a programme running 200 rings a month at 40 stones each, monthly re-orders keep the reference set relevant. A single annual order saves freight and costs consistency, because the same nominal grade a year later may be produced from different runs, and a yearly order leaves no opportunity to correct a drift before it reaches a season of production.

Tray-level inspection

A parcel packed in trays has a different drift profile from one packed loose. Each tray is sorted at a different moment, and the last tray filled can differ from the first.

Inspect tray by tray

Open every tray and grade five stones from each. A parcel of 500 stones in twenty trays needs one hundred stones graded, which is a twenty-minute task and the only reliable way to catch tray-to-tray drift. Skipping it means the drift is discovered by the setter, one rejected stone at a time, across a full production run.

A single inspection sheet per parcel is enough. It records the date, the parcel number, the trays opened and the pass and fail counts by attribute. Three months later, that sheet is what allows a buyer to compare supplier performance on evidence rather than on impressions.

Record the tray number

When a tray fails, record its number on the inspection sheet. That number lets the supplier identify which sorting batch produced it, and a supplier who can trace a tray can fix the cause rather than just replace the stones.

What our parcel terms look like

We quote moissanite parcels with a colour band, a diameter band and a carat tolerance stated in the purchase order, and we pack in mixed order for orders above 200 stones. Minimum order is one parcel for loose stones and 50 pieces per design for jewelry.

Stock ships in three to five days, bulk in fifteen to twenty days under 500 pieces. Payment is a 30 percent deposit with the balance before shipment, by T/T, PayPal or Western Union. Parcels above 500 stones include a count and weight reconciliation and a 20-stone reference set on request.

Frequently Asked Questions

Should I reject a whole parcel for a few off-band stones?

Usually not. Reject the off-band stones and ask for replacement. A whole-parcel rejection costs time and shipping, and the supplier will price the next order higher.

Does calibration fix colour drift?

No. Calibration fixes diameter, not colour. Colour drift requires sorting against a master set, which is a separate step.

Can drift be corrected by mixing trays?

It can be masked, not corrected. Mixing spreads the outliers so no single ring shows a transition, which works for a scatter setting and not for a graduated one.

How do I compare two suppliers on consistency?

Ask each for the band they guarantee in writing, then buy a 100-stone trial from both and sample identically. The trial costs little and settles the question.

Next step

Send the setting type, stone count and colour band, and we will quote with the band written into the order. Start with our stone desk, or read the guide index for setting and care context.