The second order is the real test

01What a repeat order needs

The first production run proves that a garment can be made once. The repeat proves whether the product exists as a controlled specification instead of a remembered achievement.

First runs receive concentrated attention because many decisions are still live. The pattern is discussed, sample comments are recent, and the approved garment is familiar to the people involved. That attention can carry missing information. On a repeat, fabric may come from another dye lot, meaning a batch dyed together, and production may involve different hands. Anything held only in memory becomes a new decision.

02Record how the first garment was made

Repeatability begins with distinguishing an outcome from its cause. An approved sample records how one garment turned out. It does not automatically record the pattern version, fabric state, sewing method, wash conditions or measurement tolerance that produced it. If a pocket lies flat because an experienced operator adjusted the tension by eye, the sample preserves the flat pocket but not the instruction.

The technical record has to carry that instruction. Points of measure are the named locations used to check a garment, such as chest, body length and opening. Each target needs a tolerance, the acceptable variation around it. The grade rule states how those targets change between sizes. Together they protect the fit more reliably than a note saying “same as approved sample”, because they make drift observable.

Fabric needs comparable control. Gsm, or grams per square metre, records weight, but it does not fully describe hand feel, the tactile quality of a cloth, or drape, the way that cloth falls under its own weight. A retained swatch gives a tactile standard. A lab dip, the small dyed colour submission reviewed before fabric is committed, controls shade before bulk and should be tied to the approved colour reference. Colour fastness records how well the shade resists specified washing, rubbing or exposure conditions. Each control answers a different question.

03An example: a sweatshirt changes between orders

Imagine a loopback cotton sweatshirt whose first run has the intended compact body and clean rib hem. On repeat, the shell fabric meets the recorded gsm and the chest measures within tolerance, yet the garment feels longer and the hem ripples. The specification contains an approved sample and general fabric composition. It does not state the relaxed measurement after washing, the rib recovery or how the shell and rib should be conditioned before assembly.

Nobody needs to make a dramatic error for that garment to drift. A new dye lot can receive a slightly different finish. The rib can recover with a different force. Cutting can occur before the fabric has settled. Washing then changes the shell and rib at different rates. Every individual choice may look reasonable, while the worn result loses the compact line that defined the first run.

04Prepare the next order from clear references

The corrective action is not “match the original more carefully”. Record the material and process conditions that matter. Set the relevant measurements at the state in which they will be inspected. Establish shrinkage allowance, the pattern provision made for expected dimensional change. Name the rib quality and recovery standard. Keep construction details clear enough that a different operator can reach the same result without private instruction.

The pre-production sample is central. This is the signed garment made with approved fabric, trims, labels, construction and finish before bulk begins. It is a physical standard, not a substitute for the written one. When the sample and specification disagree, the discrepancy must be closed before production. Leaving both in circulation asks the production team to choose which approval was meant.

A useful repeat pack also records change. Pattern files need clear versions. Revised points of measure should not sit beside obsolete tables without status. Substituted thread, fusible or elastic must be reviewed for the effect it can have, even when the customer will never see the component. Quiet substitutions are a common route by which a product remains recognisable yet stops feeling the same.

Inspection should be designed around product risk. A loose body may absorb small variation at the chest while a narrow neck opening cannot. A printed placement may tolerate movement on an all-over motif but not when it must align with a pocket. The agreed checks should concentrate on the features that carry fit, function and recognition, then judge them against the approved standards.

Repeat orders also expose whether the bill of materials was specific enough. A generic entry such as “black elastic” permits several technically acceptable replacements with different recovery and surface. Naming the construction, dimensions, performance and approved reference turns a purchasing note into a product control. The customer may never see the elastic, but they will feel what it does.

Before placing a repeat, review the first run as evidence alongside our guide to recording the decisions behind a specification. Which variations appeared during inspection? Which comments were resolved through conversation but never entered in the file? Are the approved colour, fabric, trim and finish standards still available and clearly named? Does the grade still protect the silhouette across the size range? Close those gaps before asking production to recreate the result.

Craft can rescue one run. Only a specification can remember it.