How Do You Spec a Fabric Before You Source It?

2026-10-31 02:34:09

Fabrics & Materials

A fabric specification POD buyers can hand to a mill is a short written document that fixes weight, width, fibre blend, shrinkage allowance, colour tolerance and test method before a single roll is ordered. It turns a conversation into a measurable target, so a dispute about hand feel or a shifted print has a reference point. Without it, sourcing decisions rest on sample memory and photographs, and the second order rarely matches the first. This guide covers what belongs in the document, which fields cause the most rework, how a spec is checked on receipt, and where a written spec stops protecting you.

Key Takeaways

A spec fixes numbers and methods, not adjectives, so a mill can be held to something.
Weight, width, shrinkage and colour tolerance cause most of the rework in custom programs.
Every field needs a stated test method, or two labs will report different results.
Knitted fabric needs a relaxation clause that woven fabric does not.
A spec is a filter for mills, not a substitute for a sealed sample.

What a Fabric Specification Has to Contain

A usable specification describes the fabric in terms a mill can reproduce and a buyer can verify. That means composition with a tolerance, construction, weight per square metre, usable width, shrinkage after a named wash cycle, colour fastness limits, and the test standard behind each number.

The document also states what the fabric is for. A jersey destined for all-over print needs different stretch and recovery limits than a woven used for a structured jacket, and the intended end use is what makes a mill push back on unrealistic numbers.

Keep the sheet to one or two pages. Long specification packs tend to be copied forward from an old program, and the fields that no longer apply are the ones that create arguments later.

The Fields That Cause Most of the Rework

Weight is the field buyers state most confidently and measure least carefully. A 180 gsm jersey from one mill can feel heavier than a 200 gsm from another, because weight is measured on a conditioned sample and the fabric may be finished differently.

Width comes second. Mills quote usable width, not roll width, and a difference of two centimetres changes the marker and therefore the yield per garment. That change lands on cost, not on the specification sheet.

Shrinkage is the field that generates customer complaints rather than mill disputes. A fabric that passes a lab test at three percent can still move more after a home wash at a higher temperature, and the return arrives long after the mill has been paid.

Colour tolerance is the last common gap. Writing a Pantone reference without a pass or fail limit leaves the mill and the buyer to negotiate every delivery on opinion rather than on a measured difference.

A Working Spec Table for Custom Apparel

Spec field	What to write	How it is verified	Why it matters for print
Composition	Fibre percentages with a plus or minus 3 point tolerance	Fibre analysis on a cut sample	Cotton and polyester take ink differently, so a blend change shifts colour
Fabric weight	gsm at a stated conditioning standard	Cut a known area and weigh it on a calibrated scale	Heavier fabric holds a print flatter and drapes differently
Usable width	Centimetres after finishing, measured inside the selvedge	Measure three places per roll	Width sets the marker and the print area a panel can carry
Shrinkage	Percent after a named wash and dry cycle	Wash three specimens, measure before and after	Shrinkage moves a printed panel relative to the seam
Colour fastness	Minimum grey scale rating for wash, rub and light	Standard test with a graded comparison card	Weak fastness shows as fading around the print edge

The right-hand column is the reason a print supplier should be involved before the fabric is booked. A specification that satisfies a garment technologist can still be a poor match for a given decoration method.

Why Specs Fail Once Production Starts

Most specification failures are not dishonest. They come from a document that was written for a different program, or from a mill quoting against its standard production rather than the buyer's sheet.

Batch variation is the structural cause. Shrinkage, elasticity, thickness, structure and resilience move between lots, and knitted fabric is more sensitive than woven because it carries tension from the knitting machine that releases over time.

Handling inside the factory adds a second layer. Spreading tension, panel placement, seam allowance width, the feed during sewing and the habits of individual operators all influence the finished measurement, so a fabric that met its spec can still produce a garment that measures outside tolerance.

Heat is the third. Washing, drying, heat setting and the press used for print can shrink, extend or distort a panel locally, and the amount differs by fibre and by process.

How to Build the Sheet Before You Source

Start from the garment, not from the mill catalogue. Write down the end use, the decoration method, the wash instructions you intend to publish, and the size range, because each of those constrains the fabric.

Then fix the numbers in the order a mill reads them: composition, construction, weight, width, shrinkage, fastness. Add the test method beside each value, and name the conditioning and the wash cycle so the result can be repeated by a second lab.

Where a value cannot be met, write the concession rather than deleting the field. A stated allowance for a knit that relaxes two percent is more useful than a silent gap.

Send the sheet with the enquiry and ask each mill to confirm or flag exceptions in writing. Replies that ignore a field are a signal about how the order will be run.

Finally, tie the sheet to a sealed sample. The specification defines the envelope, the sample defines the target inside it, and both are needed before a bulk order is released.

Reading Colour on Fabric, Not on Screen

Colour is specified on paper and judged on cloth, and the two do not behave the same way. A screen emits light through RGB pixels, while fabric reflects ambient light, so brightness, colour temperature and the viewing environment all change what a buyer sees.

The fabric changes the result again. Cotton and polyester differ in fibre structure and ink absorbency, and base shade, texture, thickness and surface finish alter how light returns to the eye.

Decoration methods have their own gamut. DTG, DTF, sublimation and all-over print reproduce the same file differently, and highly saturated blues, fluorescent greens and some violets commonly fall outside what a given system can hit.

That is why a colour field in the specification should name the light source and the viewing condition as well as the reference. A match agreed under one lamp can fail under another, and the argument is about the lamp, not the dye.

Tolerance Is a Management Decision

A tolerance is not a confession of poor quality. Every garment program has one, and the difference between a well run program and a chaotic one is whether the number was agreed before production and what happens when it is missed.

Traditional cut-and-sew runs at a tighter band than print on demand, because it cuts large lots from one marker. A program that produces after the order arrives works in smaller batches with more manual handling, so its measurement band is wider by design.

The measurement method belongs in the same paragraph as the number. Flat or hanging, zip closed or open, knit relaxed or stretched, and where the tape starts and ends all change the reading, and soft or elastic areas amplify the gap.

ISO 18890:2018 standardises how garment dimensions are measured rather than how much variation is allowed, which is why a spec sheet still has to state its own limits.

Checking a Delivery Against the Sheet

Incoming inspection should sample across the roll, not from the first metre. Mills cut from the end they can reach, and the outer wrap is the least representative part of a lot.

Weigh a known area, measure width in three places, and cut a specimen for the wash test rather than trusting the certificate alone. A certificate describes the lot the mill tested, not the roll in your warehouse.

Record the result against the batch and keep it with the cutting ticket. When a size complaint arrives three months later, that record is the only way to separate a fabric fault from a sewing fault.

Limitations: When a Written Spec Will Not Protect You

A specification cannot fix a wrong choice. If the fabric is unsuitable for the decoration method, tighter numbers will not make the print adhere or stop a heavy coverage area from stiffening.

It also cannot cover small orders well. Below a certain volume, mills sell from stock and will not run a custom lot against your sheet, so the spec becomes a selection filter rather than a production instruction.

Nor does it replace inspection capacity. A sheet with ten fields and no sampling routine is a document, not a control, and the fields that matter most are the ones someone has to measure.

The last limit is contractual. A spec that is not referenced in the purchase order has little standing when a delivery is disputed, so the sheet needs to travel with the order rather than sit in a folder.

Where Experience Changes the Numbers

In our production, the tolerance that causes the most discussion is garment measurement rather than fabric weight, because a two to three centimetre band is normal for this production model and still surprises buyers who expect cut-and-sew precision.

Our qc team checks the measurement method before the number, since a flat measurement on a relaxed knit and a hanging measurement on the same garment can differ enough to fail a lot that is inside tolerance.

Seam behaviour on all-over print is the other recurring topic. Because the panel is printed before it is cut and sewn, a pattern can sit slightly differently across a side seam, and a half to one inch shift at the seam is treated as normal rather than as a defect.

Those three points belong in the specification as written expectations, not as exceptions negotiated after the first complaint.

Frequently Asked Questions

How detailed should a fabric spec be for a first order?

Six fields are enough to start: composition, construction, weight, usable width, shrinkage after a named wash, and colour tolerance. Add the test method for each so a second lab can reproduce the result.

Should the spec include a shrinkage figure or a pre-shrunk claim?

Use a figure with a wash cycle attached. A claim that fabric is pre-shrunk says nothing about how it behaves after the customer washes it, and the complaint arrives after the mill has been paid.

Who should write the specification, the buyer or the mill?

The buyer writes it and the mill confirms it. A sheet supplied by the mill describes what that mill already makes, which removes the comparison that sourcing depends on.

Does a spec need a colour fastness rating for print as well as fabric?

Yes, and they are separate ratings. Fabric fastness covers the dyed ground, while print fastness covers the ink film, and a garment can pass one and fail the other after repeated washing.

How does fabric choice affect print area on a finished garment?

Stretch and recovery decide how much a panel grows during wear, and that changes where a fixed print area lands. A fabric with high elongation needs a smaller nominal print area to keep the design inside the panel after stretching.

Is a sealed sample enough without a written spec?

No. A sample shows one piece at one moment, while a spec defines the range a production lot may fall inside. Mills reproduce numbers more reliably than they reproduce a hand feel from memory.

Putting the Sheet to Work

A fabric specification pod program only earns its keep when the document is tied to an order, a sample and an inspection record. Write the numbers, name the methods, agree the tolerance, then measure what arrives against it.

Teams building a men's custom apparel line can start with a one page sheet and refine it after the first delivery, which is cheaper than rebuilding a program around a fabric that was never specified. The companion steps are covered in custom fabric sourcing for new brands and in this fabric sourcing timeline for custom programs.

Two habits keep the process honest. Compare mills against the same sheet rather than against each other's samples, and review the red flags in fabric sourcing red flags when vetting mills before a deposit is paid.

Trim and hardware carry the same logic, and the approach in custom trim and hardware sourcing applies to labels, zips and drawcords. Buyers who want a wider view of supplier capability can read this guide to the best t-shirt printing companies for custom apparel.

The return on a written sheet shows up in claims rather than in cost. Programs that specify before they source spend less time arguing about what was agreed, and the print on demand returns prevention through better specs approach is the practical version of that argument.