Building a size chart a factory can produce from

The direct answer: a production size chart is a table of points of measure against sizes, with a measurement method, a tolerance and a stated fit intent. It describes the finished garment rather than the wearer. Everything below is how to arrive at one, whether you are starting from an existing chart, from a reference garment, or from nothing.

Three charts, three jobs

The word chart covers three documents that are routinely treated as one, which is why a brand can hand over a size chart and a factory can still say it cannot produce from it.

The gap between the first and the third is where programmes stall. A consumer guide might carry three measurements; a garment specification for the same product carries considerably more, and how many depends entirely on the garment. A plain tee needs relatively few. A lined jacket with pockets, a collar and adjustable cuffs needs many more, because each of those components has dimensions that have to be right.

There is no universal number to aim at. The right count is the one that leaves nothing important undefined, and that is a property of the product rather than a target to hit.

Choosing a base size

The base size is the size the pattern and the fit are developed in, before grading produces the rest of the range. Choosing it is a real decision and it has a few common misconceptions attached.

• It is not automatically M, or the middle of the range. It is the size the brand wants to get exactly right first.
• It is not necessarily the best-selling size, though that is a reasonable argument for choosing it.
• It is not necessarily the same as the sample size a buyer asks to see, although in practice they are often the same for good reasons.
• It should be somewhere the grading can work outwards from sensibly in both directions, which usually argues against picking an extreme.

Once chosen, the base size carries disproportionate weight: fit is approved there and then propagated. That is why fit development concentrates on it, and why proving the range afterwards is a separate step, covered in fit development.

If you already have a chart

Much of the work is auditing what is there rather than rebuilding it.

The tolerance check is the one worth doing first. If a tolerance is wide relative to the grade between sizes, adjacent sizes overlap and a garment can be within specification in two sizes at once. That is a specification fault rather than a production one, and it cannot be sewn around.

If you have a reference garment but no chart

A physical garment is a good starting point and a limited one, and knowing which parts are which saves disappointment.

So a reference garment can be measured into a garment specification for one size directly, and the rest of the range still has to be decided. What working from a reference involves more broadly, including the intellectual property boundary, is covered in manufacturing from a reference garment.

If you have nothing

This is more common than brands expect, particularly on a first product, and there is no universally correct chart a factory can simply apply. What there is, is an order of decisions.

1. 1Define the customer. Which market, which wearer group, and anything already known about them from existing sales or from the category.
2. 2Decide the measurement basis. Whether the range will be built around target body dimensions or developed directly as garment measurements from a fit you already have in mind.
3. 3Choose the base size, and be explicit that everything else will be graded from it.
4. 4State the fit intent in words before any number is written, because the numbers are an expression of it.
5. 5Draft the garment specification for the base size: every point of measure the garment needs, with methods.
6. 6Sample and fit in the base size, adjusting the specification to what was approved rather than leaving the original numbers in place.
7. 7Set the grading across the range, which is the subject of pattern making and grading.
8. 8Validate at the extremes, since a grade that works in the middle can fail at the ends of a range.
9. 9Assign the labels last, once the garments exist as measurements.

Step six is often skipped. If the sample was approved at a measurement different from the specification, the specification should be updated to the approved value. Leaving the original number in the document means the approved sample and the specification disagree, and a later production team has no way of knowing which one to follow.

Tolerance, and where it comes from

A tolerance is the deviation from a target that production may have and still conform. It is not a grade, and it is not a shrinkage allowance.

Tolerances are agreed rather than universal, and sensible ones are not uniform across a garment: a measurement a customer notices deserves a tighter band than one nobody checks. What matters technically is that every tolerance is narrower than the grade between sizes, and that the whole set is agreed before production rather than argued about during inspection. How tolerances function in inspection is covered in quality control.

Where the chart lives

The production size specification belongs in controlled product documentation rather than in an email thread, because it will be referred to by people who were not part of the original conversation, possibly a year later on a reorder.

• Size names, matching what the labels will say
• Every point of measure, with its measurement method or a drawing
• The measurement for each point in each size
• The tolerance for each
• The base size, identified as such
• The grading relationship, or a note of where it is recorded
• The fit intent
• Which sample the specification was approved against

There is no single mandatory template for this and different buyers organise it differently. What matters is that each item is answered somewhere and that it is clear which document governs when two disagree, which is the same principle set out in the tech pack guide.

One further note for uniform and workwear programmes, where the problem is often inverted. Rather than mapping retail labels, the task can be fitting a known group of people, which means collecting wearer measurements, running trial sizing and deciding a size distribution before any range is finalised. The garment specification work is the same; the input is a population you can actually measure, which is a considerable advantage.

Size sets for staff roll-outs and size ratios for merch

For a staff roll-out, the size run comes from the people who will wear the garments. A fitting round, where staff try fit samples in every size, or a headcount survey, where each person picks a size from the garment measurements, gives a count per size for each garment; a small reserve per size covers new starters and wrong picks. Programme planning sits on the corporate clothing page and the workwear page.

Merch ratios follow the audience rather than a retail curve: a concert crowd, a company giveaway and a sports club each buy a different spread, and sell-through from an earlier drop is a strong input. The minimum is 300 pieces per colour and per style and the sizes are split inside it, so the ratio is stated per colour, for example black S 30, M 90, L 100, XL 60, XXL 20. Extended sizes and separate men's and women's fits are quotation-stage decisions, since a different fit is a different pattern; how they count against the 300 is confirmed per project at the quotation stage. See custom merch.