The chart and the grade sheet
The chart is the promise - the sizes a style will be offered in - and the grade rules are the machinery that keeps the promise: written instructions for how each dimension of the pattern moves from one size to the next. On intimates the work is unusually consequential, because a bra is a fitted garment with several independent dimensions that do not scale together. Band length, cup volume and wing height each follow their own logic, and a style that fits beautifully at the size the sample was cut in can drift apart at both ends of the ladder unless the grade was engineered rather than extrapolated. This note is about that engineering: what gets graded, where the breakpoints sit, how a size set is sealed, and how grading errors surface later as returns.
What gets graded
Every dimension that affects fit or assembly appears in the grade sheet. The usual entries for a bra program:
| Dimension | What the rule controls | Sensitivity |
|---|---|---|
| Band length and elastic cut | How the anchor lengthens per size | High: the anchor drives the fit |
| Cup pattern and depth | Volume per cup letter within the band | High: volume errors read as fit failure |
| Wing depth and height | Side support at the larger sizes | Rises toward the ladder ends |
| Strap length and width | Reach and load spread per size | Moderate: elastic absorbs some error |
| Hook-and-eye rows and width | Closure spread at the larger bands | Steps, not linear |
| Lace and print placement | Motif alignment as panels grow | Visual: most visible at center front |
Panty programs grade a simpler sheet - waist and leg openings, rise, gusset length - but the same principle holds: the rule is written, not assumed, and the sewn result is checked against it before bulk.
Breakpoints: where linear grading ends
The trap in grading is the smooth average. A rule that adds the same increment per size will hold across the middle of the ladder and fail at the ends, because bodies do not scale linearly and neither does garment mechanics. Real programs write breakpoints into the sheet: defined sizes where construction changes rather than stretches - a deeper band and wing above a stated cup size, an added hook row at a stated band, a wider strap at a stated cup volume. The note on band ride-up describes the classic end-of-ladder failure, and it is a grading failure more often than a cutting error: a band shallow for its load climbs, and the wearer reads it as a size problem. Breakpoints are also where fit sessions happen - the style is refitted at the breakpoint sizes, not assumed to carry.
The size set and what it seals
A size set is the ladder made physical: garments sewn in production fabric and elastic across the ladder - the ends, the breakpoint sizes and a center size at minimum - and fitted before bulk releases. What it seals is the whole specification chain: the chart, the grade rules and the plant's ability to execute both. Once approved, the size set becomes the reference bulk is inspected against, and the measurement sheet in the purchase order is the size set's numbers written down. Cutting a size set costs days; skipping one costs the discovery that the grade breaks at the third-largest cup, made after several thousand pieces have been cut to a broken rule. Of all the steps a program can skip, this is the one whose absence is least forgivable and most common.
Size ratio inside the run
The grade decides whether the ladder works; the ratio decides how the run is spent across it. A marketplace style usually runs a core ratio, weighted to the sizes that sell and light at the ends, while a brand launch may spread the ladder more evenly to gather real fit data from the market. Both are legitimate, and the ratio is a planning decision made together with the quantity per color, not a leftover. The interplay with minimums is direct: the minimum counts pieces across the whole ladder, so a deep ladder at a thin run starves the end sizes first. The mechanics sit in the note on OEM and MOQ explained.
Label conventions across markets
Sizing is also a language problem. Band and cup conventions differ across markets - the same garment reads as different sizes depending on which convention the label wears - and conversion charts are the bridge, with the detail in the notes on the size chart and international conversion. The production consequence is narrower than the shopping consequence but real: the label convention is chosen in writing per program, and the grade sheet speaks the same convention as the label. A program graded in one convention and labeled in another produces a warehouse of garments that are internally consistent and commercially wrong.
When grading goes wrong
The failure signatures are distinctive enough to read from returns data. Bands that ride up at the larger sizes point at band depth breakpoints. Straps that dig at the fuller cups point at strap width held flat across the ladder. Cups that gap at the smallest sizes point at a cup grade written for volume and not for the smaller frame's geometry. Each signature names its dimension, which is why the desk asks for returns reasons by size, not only by style - the buyer-side reading is in the note on why sizing varies across brands. Returns are the grade sheet's report card, arriving a season late unless the size set catches the problem first.
Writing it down
Grade rules live in the tech pack: the chart, the rules per dimension, the breakpoints with their construction changes, and the measurement sheet per size. Verbal grading - a bit deeper at the top, like the reference style - is how programs inherit someone else's fit history without the file. The desk writes the sheet, fit sessions confirm it at the breakpoints and the size set seals it. The consumer-facing side of the same subject sits on the sizing and fit guide page.
Where the desk fits
Programs opened through the quote form carry the intended ladder and any fit history in the brief, so the desk and the plant engineer the grade before sampling rather than after returns. The sequence is the one described in the note on sourcing intimates from China, with the size set in the slot where it belongs: after the sample approves, before bulk releases, and before the minimum is spent.