On this page
- Quick answer
- What "weight" actually measures — and what it does not
- How drape is decided
- The behaviour × requirement matrix
- What happens when you change fabric on the same pattern
- Target GSM is not enough — agree the tolerance
- From the factory floor: why fabric has to be fixed at sampling
- How to specify fabric in a brief so the factory can answer
- Four misjudgements that produce "the bulk isn't the sample"
- What to confirm before you order
- FAQ
- Conclusion
A buyer approves a sample in a mid-weight crepe. It fits, it hangs the way she imagined, the waist sits where she drew it. Then, six weeks later, the bulk arrives in the satin she chose to "upgrade" the range — same pattern, same measurements, same approved grade rules — and the dress does not look like the same garment. The waist has dropped. The skirt has gained a shine that reads as volume. The hem that broke cleanly on the sample now collapses against the legs.
Nothing in the file was changed. The pattern was not touched. The measurement chart is identical. And yet the dress is different, because the one variable that was treated as interchangeable — fabric behaviour — is the variable that decides whether a pattern works at all.
This is not a quality failure. It is a specification failure, and it happens because buyers are shown fabric in a way that hides behaviour: as a flat swatch, on a screen, or in a photo. Weight is the parameter that is easiest to write down and the hardest to interpret, and it is the one most often mistaken for the whole story. This article is about what weight actually tells you, what it does not, and how to specify around it.
Quick answer
GSM tells you how much mass a fabric carries per unit area. Whether that mass produces a fluid or a structured drape depends on how the fabric bends, shears and recovers — weight is one input, not the answer. A pattern is drafted against a specific combination of those behaviours, so the fabric that produced the approved fit is a specification in its own right, not a weight class.
| What you change | What actually changes |
|---|---|
| Mass per unit area (GSM) | The gravitational load the fabric carries; how fast it settles and how much it resists air movement |
| Bending stiffness | Whether the fabric holds a line or follows the body — independent of weight |
| Shear behaviour | How the fabric moves diagonally; how it drapes over a hip or a bust |
| Surface and finish | How the silhouette reads in light and in photographs |
What "weight" actually measures — and what it does not
GSM is a mass measurement: how many grams one square metre of the fabric weighs. It is objective, repeatable, and commonly listed on mill specifications — which is exactly why it should be recorded in the approved fabric specification rather than remembered from a conversation.
It is also incomplete, and the gap between what GSM measures and what a buyer needs to know is where most specification mistakes start.
Two fabrics can share the same GSM and behave nothing alike. A 110 GSM crepe and a 110 GSM satin sit in the same weight class on paper. In the hand they are different materials: one absorbs light and falls with a matte edge, the other throws light back and slides over itself. Put both on the same pattern and you get two garments.
What GSM does not tell you:
- Stiffness. A fabric's resistance to bending is a separate property. Some lightweight fabrics are stiff (organza, taffeta); some mid-weight and heavy fabrics are soft (washed viscose, fluid satin).
- Recovery. How a fabric returns to shape after stretching at a seam, a waistband or a strap. Two fabrics of identical weight can differ sharply here.
- Surface. Matte or reflective, smooth or textured. This is the property that most changes how a silhouette reads in photos, and it is invisible in a number.
- Directional behaviour. Woven fabrics often behave differently along warp and weft. The same pattern cut on the cross-grain can hang differently from the straight grain.
So the working rule is this: GSM tells you the class. Hand, stiffness and shear behaviour tell you the behaviour. You need both, and only one of them is on the spec sheet.
Choosing between fabrics is a separate decision from predicting how one will behave. How to choose fabric for a dress design covers the selection side — which fabric suits which garment. This article covers the behaviour side — what a given fabric will actually do to a fit once the pattern is cut.
GSM, thickness, stiffness and drape are different things
These four terms get used interchangeably in conversation and mean completely different things in production. Keeping them apart is the single most useful habit a buyer can build.
| Term | What it measures | Typical unit | What it does not tell you |
|---|---|---|---|
| Weight / mass | Mass per unit area | g/m² (GSM) | Stiffness, thickness, drape |
| Thickness | Caliper — distance between the two faces | mm | Weight, stiffness, drape |
| Stiffness | Resistance to bending | (measured, rarely quoted) | Thickness, recovery |
| Drape | The shape the cloth takes under gravity | — | Nothing on its own: it is the result |
A light organza is thin, light and stiff; a heavy viscose satin is thicker, heavier and fluid. Weight, thickness and stiffness move independently, which is why a specification that lists only GSM has described roughly a third of the fabric a pattern was drafted against — see how satin, crepe and chiffon differ in behaviour for a side-by-side of three fabrics often treated as interchangeable.
How drape is decided
Drape is the shape a fabric takes when it is allowed to fall under its own weight. It is not a fixed property so much as a relationship: the fabric's mass against the stiffness and shear behaviour that resist it. Add mass and you add the force that pulls the fabric down into folds. Add stiffness and you add the force that holds it away from the body.
That relationship produces three recognisable behaviours, and each one belongs to a different construction.
Stiff drape — the fabric holds its own shape
Stiff-drape fabrics fall in broad, structural folds and stand away from the body. They keep a line the pattern drafted, which makes them forgiving of a simple construction and unforgiving of a complicated one: every added seam, dart or gather has to be resolved cleanly because nothing hides in the fall.
This is the behaviour behind most structured garments — a bodice that holds its own shape, a peplum that stays out, a collar that stands. It is also where weight is most visible: a stiff fabric that is too light will buckle at the fold lines, and one that is too heavy will drag its own shoulder seams out of position.
Fluid drape — the fabric follows the body
Fluid fabrics fall in narrow, close folds and move with the wearer. They skim rather than hold, which is exactly what makes them flattering in the right construction and merciless in the wrong one. There is nowhere for a fluid fabric to hide a construction decision: the fabric reads the body underneath, so every fit choice becomes visible.
Fluid drape is the behaviour behind slips, bias-cut dresses and anything described as "liquid". How a fluid garment reads on the body is influenced not only by the pattern, but also by the fabric's mass, stiffness and shear behaviour as it falls over the wearer — the reason a fluid dress that suits one wearer can look entirely different on the next.
Semi-drape — the working middle
Most commercial womenswear sits here: enough mass to hold a soft shape, enough softness to move. Semi-drape is the hardest to get consistent, because it sits close enough to both ends of the scale that a small change in behaviour moves the garment from one category to the other. A semi-drape fabric specified loosely — "mid-weight crepe" — is one of the most common causes of "the bulk doesn't match the sample".
The practical consequence: clearly structural fabrics can be easier to control within an approved behaviour range. Semi-drape fabrics usually need tighter specification, because small changes in stiffness, shear or finish become visible quickly.
The behaviour × requirement matrix
This is the table worth keeping open while you brief a factory. Read it as: pick the behaviour you want, then check what the pattern and construction have to do to support it. Note that the mass column describes a typical relationship, not a rule — behaviour is set by stiffness and shear as much as by weight.
| Drape behaviour | Typical relationship to mass | What actually governs it | What the pattern must do | Typical failure if it is wrong |
|---|---|---|---|---|
| Stiff, structural | Mass is not the deciding factor — stiffness dominates, and can be light or heavy | High bending stiffness | Hold a clean neckline and waist; avoid tight gathers | Folds buckle; shoulder seams drag; the garment reads "cardboard" |
| Semi-drape | Sits mid-scale; small changes move it | Balance of mass and stiffness | Balance ease at waist and hip; control hem width | Drifts between stiff and fluid; bulk stops matching sample |
| Fluid, following the body | Can be heavy (fluid satin, viscose) or light | Low stiffness, low shear resistance | Limit darting; work on the bias or in panels; design the lining in | Construction shows through; cling at stress points; hem collapses |
| Surface-driven / very light | Lightest end | Low mass, high surface effect | Add structure or lining deliberately; treat as a layered build | Transparency changes; the garment has no shape of its own |
> Note the pattern: as behaviour shifts towards fluidity, construction has to do more work. The weight does not disappear from the garment — it moves into the build.
What happens when you change fabric on the same pattern
This is the section that saves money, because it is the mistake that costs the most.
Take one style — a fitted bodice with a defined waist and a panel skirt — and specify it two ways. Same pattern, same measurements, same grade rules. Only the fabric changes.
| Element | Version A — mid-weight woven | Version B — lighter fluid woven |
|---|---|---|
| Waist position as worn | Reads at the drafted position | Reads lower; the fabric stretches over the bust and settles |
| Bodice fit | Holds shape; ease reads as intended | Follows the body; ease reads as puckering |
| Skirt movement | Controlled, moderate swing | Full movement, but hem collapses inward at rest |
| Dart behaviour | Darts resolve into the shape | Dart ends telegraph through; visible on light colours |
| Lining requirement | Optional | Effectively required to stop cling and show-through |
| Sample validity | Represents bulk | Sample no longer represents bulk at all |
The last line is the expensive one. Once the fabric changes, the approved sample stops being a reference. From a factory's point of view, a sample approval is an agreement about what the bulk will be — and a fabric substitution signed off later voids that agreement without anyone noticing it happened. The bulk then arrives technically correct and commercially wrong, and nobody can point to the decision that caused it.
How to approve a bulk fabric substitution
When a substitution is unavoidable — a mill runs out, a dye lot fails, a cost target moves — the approval is a comparison, not a weight check. Work through six items against the approved fabric, not against a description of it.
- [ ] Mass / GSM — replacement versus approved, recorded in the same stated condition.
- [ ] Construction — weave or knit, yarn count and sett where it is known.
- [ ] Stretch and recovery — especially at waist, bust and straps, where fit is decided.
- [ ] Finish and hand — how it feels and how it falls, judged on a hanging length, not a swatch.
- [ ] Surface — matte or reflective, before and after.
- [ ] Drape against the approved standard — the replacement on a body or form, compared with the approved sample.
If behaviour falls outside what the approved sample established, re-sample before bulk. Similar weight is not evidence of similar behaviour, and a weight check on its own will pass a fabric that changes the fit. GSM is a control point, not a substitute-approval system.
Target GSM is not enough — agree the tolerance
This is the step most buyers skip, and it is the one that decides whether the bulk that arrives is the bulk that was approved.
A target of "120 GSM" does not mean every roll will measure 120. Mass per unit area varies within a roll, between rolls and between dye lots, and the measured figure also depends on the condition of the cloth when it is weighed. So the specification needs four things, not one:
| What to fix | Why it matters |
|---|---|
| Target GSM | The figure the approved sample was measured at |
| Agreed tolerance | The range the bulk is allowed to fall inside |
| Measurement method and condition | Whether the figure refers to greige or finished cloth, and how the sample is taken |
| What happens if bulk falls outside | The remedy, agreed before production rather than argued after it |
A word on the tolerance itself: there is no universal percentage that fits every cloth, mill and project. Tolerance has to be agreed commercially, per fabric, with the supplier — a stable, compact woven may hold a much tighter band than a loose knit or a pile fabric. What is standardised is the method of measuring mass per unit area, described in ASTM D3776 / D3776M, which sets out how the sample is taken and weighed rather than what the acceptable range should be for your order.
The practical rule: do not rely on a target GSM alone. Record the agreed commercial tolerance, record the condition the figure refers to, and compare sample and bulk under the same stated condition. Without that, "the GSM matched" and "the fabric is different" can both be true at the same time.
From the factory floor: why fabric has to be fixed at sampling
On a factory floor, fabric is not the start of the process — it is the constraint the whole process is planned around.
Pattern and construction are drafted around one behaving material. When a pattern maker works up a style, they are not drawing a shape; they are resolving a specific fabric's stiffness, shear and mass into seam lines. The dart intake on a mid-weight woven is not the same as on a light fluid one, even for the same body. Change the material and those decisions stop being correct — which is why the relationship between fabric and pattern is worth understanding on its own; see how pattern making works in womenswear manufacturing.
Sampling is where the behaviour gets translated into numbers. This is why sampling is scheduled before fabric is booked for bulk, and why we hold a sample lead time of 7–10 days from tech pack approval rather than treating it as a formality. The sample is the only stage where fabric and pattern are tested against each other before money is committed; sampling is the commercial version of that stage.
Bulk scheduling depends on how the fabric handles. Very light or fluid fabrics often need tighter handling control during cutting, sewing and pressing, because machine settings that resolve a mid-weight fabric will mark or distort a light one. Very heavy, thick, pile or surface-sensitive fabrics introduce a different set of constraints — velvet and bouclé are not easier to handle than chiffon, they are differently difficult. That is a scheduling fact, not a marketing one, and it is one of the reasons production timing on some styles is quoted as a range rather than a single date.
Inspection reads the fabric. Surface faults, shading and directional marks are found by handling the cloth, which is why fabric inspection happens on receipt, before cutting. A shading issue caught on the roll costs a conversation. The same issue caught in finished garments costs the garments.
How to specify fabric in a brief so the factory can answer
Most fabric briefs fail for the same reason: they describe a look and leave the behaviour open. A factory can quote an open brief — but what it quotes is a range, and the range is what produces the argument three months later. A brief that can be answered precisely has four parts.
1. The behaviour the garment needs. Say what the finished piece has to do: hold its own shape at the bodice, fall without clinging, stay opaque in daylight, survive a working day without creasing at the seat. Behaviour is what the pattern is drafted against, and it is the part a mill can genuinely search its range for.
2. The weight class. Give a figure if you have one, and say whether it is a target or a ceiling. "Around 120 GSM, softer than it is stiff" is a specifiable request. "Mid-weight" is not, because one supplier's mid-weight is another supplier's light — and the middle of the scale is exactly where small differences change the behaviour.
3. The surface. Matte or reflective. This single choice changes how a silhouette photographs, and it is the most common item left unstated — even though it is the easiest to fix.
4. The constraints. Cost ceiling, minimum quantity, colourways, and the market's care or compliance requirements. These do not change what the fabric is, but they change which fabrics are available, and a brief that omits them produces quotes against fabrics you cannot actually buy. If shortlisting fabric is not work you want to do yourself, that is the service described under fabric sourcing.
| How the brief is written | What a factory can do with it | Where it usually ends up |
|---|---|---|
| "Something like this, but softer" | Nothing specific | Two rounds of swatches, then a weight change that breaks the fit |
| "Mid-weight crepe" | Shortlist, quoted as a range | The range is too wide; the sample lands mid-scale and the bulk drifts |
| "Around 120 GSM, matte, must not cling" | Search a class, propose two or three, explain each | The brief becomes materially easier to source and compare |
| "Same as last season, but cheaper" | Guess | The silhouette changes and nobody agreed to the change |
A worked example: one dress, two briefs
A buyer is developing a fitted midi dress with a defined waist. She sends two versions of the same request.
Brief A: "I want a fluid satin, something premium, around the price of the crepe we used last time."
What comes back is a satin swatch in a weight class lighter than the crepe the pattern was drafted around. It looks right on the card. It goes to sampling. The sample is made, tried on, and the waist reads low because the fabric stretches over the bust and settles — the same failure described at the start of this article. Fixing it means either re-drafting the bodice for a different behaviour or moving to a heavier satin, which is a different price conversation, and the two weeks spent on the first sample are gone.
Brief B: "Fitted midi, defined waist, woven. Needs to hold the waist without boning; must not cling at the hip; matte is preferred but a low sheen is acceptable. Target around the mass of the crepe we used last time, not lighter."
What comes back is two or three options inside the same behaviour envelope, each with a note about which is closest to the brief and where it departs from it. One is chosen, it samples cleanly, and the bulk matches because nothing about the fabric's behaviour was left to chance.
The difference between the two briefs is not expertise. It is four sentences that name behaviour, mass, surface and constraint. Brief B costs ten minutes and saves a sample round.
Where fabric is the driver of a price change, it is worth reading what actually moves a womenswear manufacturing price before committing — fabric is usually the largest single lever, and the cheapest fabric is rarely the cheapest garment once construction and rework are priced in.
Four misjudgements that produce "the bulk isn't the sample"
1. Judging fabric from a swatch. A swatch is a small square on a card. Weight, drape and surface only appear at scale and under gravity. A swatch can tell you colour, texture and roughly how it feels between fingers. It cannot tell you what a 90 cm skirt length will do. Where drape matters, ask for a length of at least a metre, hung, not folded.
2. Treating "same composition" as "same fabric". Two 100% polyester wovens can differ in weight, weave, finish and recovery. Composition is a starting point for care and durability, not a specification for behaviour.
3. Specifying by price tier instead of by behaviour. "A cheaper version of this" is not a specification. The useful form of that request is: same look, more mass, or same behaviour, softer hand — and then decide which one the pattern can absorb.
4. Approving the sample and the fabric separately. If the sample is approved in one fabric and the bulk is booked in another, the approval covers a garment you are not making. Approve the pair, and record the tolerance with it.
None of these are exotic. They are the ordinary ways a specification comes apart between a photograph and a carton.
What to confirm before you order
- The GSM of the sample fabric, written down, not remembered.
- The same figure for the proposed bulk fabric — and the agreed tolerance around it.
- The stated condition the figure refers to (greige or finished), so both numbers mean the same thing.
- A hung length of the bulk fabric, seen falling, not a swatch.
- The behaviour the pattern was drafted for, agreed in writing rather than implied by a weight class.
- Whether the lining decision was made against the sample fabric or the bulk fabric.
- Whether the surface (matte/reflective) matches what the sample was approved on.
- If any of the above changes after sample approval: run the substitution checklist, and re-sample if behaviour falls outside the approved standard.
That list is short on purpose. Every item on it is a variable that can change the garment while every number on the measurement chart stays correct.
FAQ
Can two fabrics with the same GSM really fit differently? Yes. GSM is a mass measurement, not a behaviour measurement. Stiffness, shear and recovery are separate properties, and they are what a pattern is actually drafted against. Two fabrics of identical weight can produce visibly different fit on the same pattern.
How do I ask for a heavier fabric without changing the look? Describe the behaviour you want, not the number: less flighty, less cling, more opaque. Then check whether the mill offers the same construction at higher mass. Changing mass while holding the look is possible; changing mass without changing behaviour usually is not.
Does a lining fix a fabric that is too light? It changes the problem rather than removing it. A lining adds mass to the assembly and can control cling and show-through, which is often enough. It does not give the outer fabric stiffness, so a pattern that needs structure will still need structure.
Why can't I just approve the sample and choose the bulk fabric later? Because the sample approval is a statement about a specific fabric. Approve one and produce another and the approval no longer describes the goods. If the fabric is genuinely open at approval, record the approval as provisional so the change is a decision rather than an accident.
Is a heavier fabric always better quality? No. Higher mass correlates with durability and opacity in many constructions, and it is also a design choice. A fluid dress made in a heavy fabric is not a better dress — it is a different one, and likely a worse one for its purpose.
What should go in the tech pack about fabric? Composition, mass in GSM, weave or knit, finish, colour reference, and the behaviour the pattern was drafted for — plus the agreed tolerance if bulk is not yet booked. What a fashion tech pack should include covers the field list in full.
Conclusion
Fit is usually discussed as a pattern problem, and pattern is where it gets fixed. But pattern is only a set of instructions, and fabric is the material those instructions are written for. Mass, stiffness, shear, surface and tolerance decide whether the instructions produce the garment that was approved or a garment that merely carries the same measurements.
The practical version of this is unglamorous: write the GSM down, agree the tolerance, look at a hanging length, fix the fabric at sampling, and treat any later change as a new approval rather than a like-for-like swap. Do that, and "the bulk doesn't match the sample" stops being a risk you manage and becomes a situation you simply do not create. If you want the fabric and the pattern tested against each other before you commit to a bulk run, send us the brief with your target fabric and we will tell you honestly whether the two belong together.