Quick answer
A women's top fits correctly when four documents agree with each other: the tech pack (which states the measurements and how they are taken), the graded size chart (which states how those measurements step up and down across sizes), the fit sample (which proves the pattern against a real body), and the sealed sample (which becomes the physical reference for bulk). If any of the four disagrees with the others, the bulk order will be judged against whichever one the inspector happens to hold — and that is how a shipment that is correct to the tech pack gets rejected against a sample that was quietly amended. Fixing the four documents at the start costs a few days; reconciling them after production costs the shipment.
Fit approval is where a women's apparel programme is either won or lost, and it is rarely a question of talent. It is a question of whether the documents contradict each other, and whether the contradictions were found while the correction was still cheap.
This guide describes the fit development and approval process we run at Guangzhou Maodi Clothing Co., Ltd. for women's tops, camisoles, tank tops, sleeveless tops and knitwear. It is written for designers, technical designers and sourcing managers who need the approval cycle to converge in two or three rounds rather than six.
1. The tech pack: specify the measurement, not just the figure
A measurement without a stated method is not a specification. A chest width of 46 cm could be measured flat across the front, flat across a laid-flat garment including both layers, or on a form — and those three readings can differ by more than the tolerance the factory is being asked to hold.
Every measurement in a tech pack should carry four things:
| Component | What it fixes | Typical failure when omitted |
|---|---|---|
| Measurement point name | Which dimension is meant | Ambiguous "length" measured from a different start point |
| How to measure | Flat, on a form, with or without stretch | Readings differ by more than the tolerance |
| Nominal value and tolerance | The acceptable band | A correct garment rejected for being 4 mm out |
| Grading rule | How the value steps across sizes | Sizes that fit at the middle and not at the ends |
The fourth row is the one most often left to the pattern maker. It should not be. Whether the waist steps by 4 cm or by 5 cm across sizes, and whether the length stays constant or grades up, is a design decision about who the range is for. Left unstated, the pattern maker applies the house rule — which may be perfectly reasonable and completely unlike what the brand intended.
2. Grading: the step that decides whether the ends of the range sell
Grading is not arithmetic applied to a base size. It is a set of deliberate decisions about how a silhouette evolves, because a body does not scale uniformly. The three questions that resolve most grading disputes:
- Which dimension anchors the range? Usually the bust or the waist for a women's top. The anchor is the dimension the customer uses to choose a size, so it should be the one that is graded most carefully and held to the tightest tolerance.
- Where does the range break? Many women's ranges need a split grade — a different rule below and above a mid-size — because the relationship between bust, waist and hip changes across the range. A single uniform rule applied from XS to XXL is what produces a size at one end that fits nothing.
- Does the fabric change the rule? A rib knit with high recovery will accommodate more than a woven with none, so the same nominal chart can yield a comfortable garment in one fabric and a restrictive one in another. Fabric and grading decisions should therefore be made together, not sequentially.
| Measurement | Typical grade step (women's tops) | Notes |
|---|---|---|
| Chest / bust (circumference) | 4 cm | Usually the anchor dimension |
| Waist (circumference) | 4 cm | Can need a split grade in extended ranges |
| Across-shoulder | 1–1.5 cm | Grades far more slowly than the body |
| Body length | 1–2 cm | Often held constant for a cropped silhouette |
| Sleeve length | 1–1.5 cm | Grades with arm length, not with torso |
| Neckline depth / width | 0.5–1 cm | Small steps; large steps change the design |
These are working figures for a women's top range, not universal rules. The value of the table is that it makes the decisions explicit and reviewable, so that a disagreement about a specific size can be traced to a specific rule rather than argued as a matter of taste.
3. Running the fit sample round so that it converges
A fit sample that comes back and generates twenty comments is not progress; it usually means the brief was under-specified and the sample room filled the gaps with judgement. Three practices keep a sample round productive.
Send the reference with the brief. A garment with the drape and fit you want is worth more than a page of description. Where the reference is a competitor garment, note which specific properties you want carried over — the neckline drop, the armhole depth, the hem weight — because no reference garment is right in every respect, and the sample room will otherwise copy it wholesale.
Comment on fit, not on preference, in round one. The first round should resolve structure: does the armhole sit correctly, does the neckline hold its shape, does the garment stay in place when the wearer moves. Colour, trim, label placement and finishing details are real decisions but they belong in the later rounds, and mixing them into round one lengthens the cycle without improving the fit.
Number the comments and require a numbered response. This sounds administrative and is in fact the single most effective practice in the whole process. When comments are numbered and responses are numbered against them, it becomes immediately visible when a point was not addressed — rather than being discovered at the bulk inspection.
| Round | Purpose | What is decided |
|---|---|---|
| Proto sample | Prove the pattern and the silhouette | Construction, proportion, overall fit direction |
| Fit sample round 1 | Resolve structure on a body | Armhole, neckline, length, where it sits |
| Fit sample round 2 | Correct the residual fit points | Measurement adjustments within the grade rule |
| Pre-production (PP) sample | Prove the production fabric and trims | Fabric behaviour, decoration, labels, packaging |
| Sealed sample | Fix the bulk reference | The physical standard the bulk will be measured against |
The move from fit sample to PP sample is where fabric enters the picture formally, and it regularly resets fit. A garment that fits correctly in a 160 GSM sample jersey can sit differently in the 180 GSM production cloth, and a change of fibre blend can alter recovery enough to move a neckline. For that reason we run the PP sample on production fabric and production trims, not on a substitute — the point of the round is to catch exactly this class of change.
4. The sealed sample: what it must contain and who holds it
The sealed sample is the most important document in the programme and the least discussed. It is the physical object against which the bulk will be judged, and if it is incomplete or held by only one party, the approval it represents is not enforceable.
A workable sealed sample set contains:
- The garment itself, in the production fabric, with production trims, labels and packaging.
- The approved measurement chart for the size the sample represents, with the tolerances that apply.
- The colour reference, including the light source under which the colour was approved.
- The decoration specification — print or embroidery placement, size and technique.
- The packing specification — folding, hangtag position, polybag and carton configuration.
Two copies should exist: one held by the brand, one held by the manufacturer. When both parties hold the same object with the same accompanying documents, a bulk inspection becomes a comparison rather than a negotiation. When only one party holds it, every inspection becomes a discussion about what was actually agreed.
5. Where the approval cycle most often stalls
| Symptom | Usually caused by | Correction |
|---|---|---|
| Six sample rounds, still not converging | Comments not numbered; points re-raised after being closed | Number the comments and require numbered responses |
| Fit correct in sample, wrong in bulk | PP sample run on substitute fabric | Run the PP sample on production fabric and trims |
| One size in the range fits, others do not | Single grading rule applied across the full range | Split the grade at the mid-size where the body relationship changes |
| Bulk rejected against the sample | Sealed sample incomplete, or held by one party only | Complete the sealed sample set and duplicate it |
| Measurements disputed at inspection | Method of measurement not stated in the tech pack | State the method for every measurement point |
| Colour approved, bulk rejected | Light source not recorded at approval | Record the light source with the approved swatch |
Read as a list, the corrections are all documentation. That is the real finding of this article: fit approval does not fail because the pattern was difficult, it fails because the four documents that describe the garment were allowed to disagree.
Frequently asked questions
How many sample rounds should a women's top need?
Typically three to four working rounds: a proto sample to prove the pattern and silhouette, one or two fit sample rounds to resolve structure on a body, and a pre-production sample on production fabric and trims. If a programme is routinely needing six rounds, the cause is usually not the pattern — it is comments that were closed and then re-raised, or a brief that left the sample room to fill gaps with judgement. Numbering the comments and requiring numbered responses against them fixes most of this, because an unaddressed point becomes visible instead of being discovered at bulk inspection.
Why does my garment fit differently in production fabric from the sample?
Because fabric drives fit more than most people expect. Moving from a 160 GSM sample jersey to a 180 GSM production cloth changes the drape and the way the garment hangs; changing the fibre blend can change recovery enough to move a neckline or an armhole. This is precisely why the pre-production sample should be made in production fabric with production trims, rather than in an available substitute. A PP sample on substitute fabric proves the pattern and says nothing about the garment that will actually ship.
What should be in a sealed sample?
The finished garment in production fabric with production trims, labels and packaging; the approved measurement chart for the sample's size including tolerances; the colour reference with the light source under which colour was approved; the decoration specification with print or embroidery placement, size and technique; and the packing specification covering folding, hangtag position, polybag and carton. Two identical copies should exist, one held by each party, so that a bulk inspection is a comparison against a shared object rather than a negotiation about what was agreed.
Can one grading rule cover the whole size range?
Often not. The relationship between bust, waist and hip changes across a woman's size range, so a single uniform step applied from the smallest to the largest size tends to produce a size at one end of the range that fits almost nobody. Many ranges need a split grade — a different rule above and below a mid-size. Whether a range needs a split depends on how far it extends and who it is designed for, and the decision belongs in the tech pack rather than with the pattern maker, because it is a decision about the customer rather than about arithmetic.
How should measurements be written so they are not disputed at inspection?
Every measurement point should carry four things: the name of the point, the method of measurement, the nominal value with its tolerance, and the grading rule across sizes. The method is the one most often missing and the one that causes the most argument, because a chest width measured flat across the front and the same nominal value measured on a form can differ by more than the tolerance being applied. Once the method is written down, an inspection becomes a check rather than a debate.
Do you produce sample sets before a bulk order?
Yes. Sample making is a defined stage in our process, sitting between fabric selection and mass production, and it exists specifically to fix the measurements, construction and finish before bulk begins. For customers developing a new style we work through proto and fit samples and then a pre-production sample in production fabric, and we duplicate the sealed sample so that both the customer and our factory hold the same reference. Low minimum order quantities mean a brand can develop and test a style before committing to a large volume.
Send a tech pack and we will return a fit and grading review
Guangzhou Maodi Clothing Co., Ltd. is a women's apparel OEM and ODM manufacturer specialising in camisoles, tank tops, sleeveless tops, built-in bra tops and knitted garments, with over 17 years of apparel manufacturing experience across a 12,000 m² facility and a monthly output capacity of up to 1,000,000 pieces. Our process runs from design development and fabric selection through sample making, mass production and final inspection, with a low minimum order quantity for brand trials and the capacity to scale into large-volume repeat programmes. Send a tech pack, a reference garment or a size chart, and our team will return a fit review, a grading recommendation and a sample plan before production is scheduled.
