Dyelot Variation Inside One Style: Three Shades of Navy

Background

A European menswear brand ordered 24,000 zippers for a single navy chore jacket. The order was large enough that the supplier split dyeing across three lots to fit the dyehouse schedule. All three lots were approved against the same master, and each individual lot looked acceptable when inspected on its own.

The zippers arrived at the garment factory in three shipments over five weeks. The factory cut and sewed each shipment as it arrived, which meant that jackets from different shipments were assembled at different times but packed into the same cartons for the same retail orders.

What Went Wrong

At the brand's warehouse, a visual audit of finished cartons found three distinct navies: one slightly greener, one slightly darker, and one close to the master. The differences were small enough to pass a single-lot inspection but obvious when jackets from different lots hung side by side on a retail fixture. Store staff reported customers comparing two sizes and choosing the one with the preferred shade.

The brand rejected about 6,000 jackets and asked the factory and supplier to explain how three approved lots could produce a visible mismatch. Both parties had inspection records showing each lot within tolerance, which made the complaint hard to resolve at first.

• Three dye lots, each individually approved.
• Visible shade difference when garments hung together.
• About 6,000 jackets rejected at the warehouse.
• Store-level customer behavior affected by shade choice.

Root Cause

The tolerance was applied to each lot independently. If the master sits at a delta-E of zero, Lot A at 0.9 and Lot B at 1.1 are both acceptable, but the difference between A and B is 2.0, which is visible side by side. The specification controlled lot-to-master deviation but not lot-to-lot spread, which is the number that matters when garments from different lots share a retail floor.

Cutting practice made it worse. Because each shipment was cut separately, garments from a single lot were packed together, so a carton could contain a run of jackets that all leaned the same way. If the factory had mixed lots across the cutting table, the variation would have been distributed and far less noticeable.

How It Was Resolved

The supplier re-measured all retained samples and confirmed the lot-to-lot spread. The brand and supplier agreed a new rule for multi-lot orders: every lot must be within delta-E 1.0 of the master and within 1.2 of every other lot in the same purchase order. Any lot outside the second condition is re-dyed or held for a different program.

For the rejected jackets, the factory re-cut plackets on the affected units where possible and reworked the rest into a separate sku sold through an outlet channel with a clear shade note. The brand retained the master and a physical sample from each lot for future comparisons.

Cutting practice changed as well. When multiple lots are in play, the factory now spreads them across the cutting table so that each bundle contains a mix. Retail cartons are packed by bundle rather than by lot, which distributes any residual variation across stores instead of concentrating it in one location.

The retail impact was larger than the rejection count suggested. Because each carton contained jackets from a single lot, some stores received only the green-leaning navy while others received only the dark one. Customers comparing sizes in the same store saw two shades on the same fixture, which produced the complaints that triggered the audit. The problem was not the average color but the distribution of variation.

The re-cut solution was chosen over scrapping because the placket allowed access to the zipper without destroying the jacket. The factory removed the tape, re-set a zipper from a single re-dyed lot, and re-finished the seam. Units that could not be reworked went to an outlet channel with a shade note, which preserved most of the value.

The supplier also changed its dye planning for large orders. Where volume allows, an order is scheduled as one dye lot even if it spans several weeks, with the tape held in a climate-controlled area until assembly. When a split is unavoidable, the lots are assigned in a defined sequence and the lightest and darkest are reserved for different programs rather than mixed in one purchase order.

The brand's shade approval workflow now requires a visual check of all lots side by side before any lot is released. The supplier sends a bench sample from each lot to a single location, where they are viewed together under D65 and TL84. A lot that looks acceptable alone but stands out in the group is rejected at that stage, which is cheaper than discovering the spread after the garments are sewn and packed.

Lessons Learned

• Control lot-to-lot spread, not only lot-to-master deviation.
• Define a maximum delta-E between any two lots in one purchase order.
• Mix dye lots across the cutting table when multiple lots are unavoidable.
• Pack retail cartons by bundle, not by dye lot.
• Retain a physical sample from every lot for side-by-side review.

Three individually acceptable lots can still be collectively unacceptable. The number that matters is the distance between the darkest and lightest piece a customer can see at the same time. Since this episode, the brand writes a lot-to-lot limit into every multi-lot order, and Donglong Zipper plans large orders as a single dye schedule wherever the volume allows.