Should I Keep One Size Chart or Localise per Market?

The Answer Changes With the Channel

The decision is not really about countries. It is about who controls the size information your customer sees, and how many places that information has to be maintained.

Read that grid as a cost ladder. The further right you move, the more the localisation is driven by a commercial counterparty rather than by the customer's body. A brand that localises because a chain demanded it is making a calculated investment. A brand that localises for eight markets because it seemed thorough is usually paying for inventory it cannot sell.

How to Build One Chart That Serves Three Markets

This is the sequence we use with buyers before a linen range goes to sampling. It takes about a week and it removes most of the arguments that surface later at the label stage.

Once that structure exists, "localise or not" stops being a philosophical question. The chart is global, the mapping is local, and both are cheap to maintain. A practical reference for the market-by-market conversion itself is our linen sizing guide for global markets.

Cost and Inventory Impact

Localising grading is a decision with a per-style, per-colour cost. The numbers below are the ranges we see on a linen program with four sizes and three colours, at a 60-piece minimum per style and colour.

The inventory line is the one that decides the question. A localised grading programme does not just cost more to develop; it splits your buy. Three markets at four sizes and three colours means thirty-six positions to forecast instead of twelve, and the minimum order per style and colour applies to each of them. On a first season with modest sell-through data, that much spread usually produces stockouts in one market and markdowns in another at the same time.

Label localisation, by contrast, splits almost nothing. The garment is identical; only the printed tape differs. That is why the mapping approach wins for most brands: it captures nearly all of the commercial benefit of localisation at a fraction of the operational cost.

Label minimums deserve a line of their own. A printed satin label typically carries a minimum of 1,000 to 5,000 pieces per variant, and the unit price falls sharply toward the top of that range. Two variants ordered at the low end can still land close to one variant at the high end once the second plate and the second proof are counted, so the quotation is rarely as lopsided as it first looks. What scales badly is the third and fourth variant, where the label minimum starts to outlive the style it was ordered for.

What to Localise and What to Keep Global

The split is cleaner than it first appears. Anything that describes the body is global. Anything that describes the transaction is local.

The two lists have different owners, which is why mixing them causes trouble. The global list belongs to the technical team and changes only when the pattern changes. The local list belongs to compliance and marketing and can change season by season without a single stitch moving. If both lists live in the same document, every label update becomes a grading review.

The print method sits on the global side of that line, even though it feels like a local decision. Woven tape, printed satin and heat-transfer print behave differently on linen: a heat transfer on a coarse weave can crack along the fold, and a woven tape adds bulk inside a garment-washed seam. Decide the method once for the whole program and let only the language block change inside it. Placement, durability and print methods are covered in our guide to care labels and branding for linen garments.

What Each Market Actually Expects

The differences are real, but they are differences in notation and expectation rather than in body shape.

Notice that all four rows end in the same place: a centimetre table decides the fit, and the letters are a translation. That is the whole argument for one chart. The moment you accept that the chart is the product and the letters are the interface, localisation becomes a display decision instead of a development decision.

Three Mistakes That Cost Real Money

Each of these appears regularly in first-season linen programs, and each is avoidable before cutting.

The common thread is that all three trade a small saving now for a large cost later. The mapping approach has the opposite profile: a small amount of work up front, and a chart that stays correct while markets, channels and labels change around it.

Applying It to a Linen Range

Two linen-specific details decide whether the chart survives contact with customers.

First, shrinkage. A 3-7% first-wash shrinkage on a 116 cm dress length is 3.5 to 8 cm of length, which is a full size step. The chart has to be built on the washed measurement, and the care label has to state a care method that keeps the result stable. A chart measured before washing and a label promising tumble drying will produce complaints that look like a sizing failure but are actually a care contradiction.

Second, drape. Linen garments are cut looser than the body because the fabric has no stretch and hangs from the shoulder. Customers comparing a flat garment measurement to a body chart will always read the garment as too big. Say explicitly, on the site and on the label, that the numbers are body measurements. Our own programs treat this as part of the size block rather than as a footnote, because it removes a large share of the "runs large" feedback that linen collects by default.

Both details live in the global layer, not the local one. They apply identically in Chicago, Lyon and Manchester, which is another reason to keep the technical chart in one place and let only the presentation localise.

Petite and tall ranges are the one case where a separate chart is usually justified, because the difference is proportion rather than a letter translation: a shorter rise, an adjusted waist position and a shorter sleeve cannot be produced by relabelling. The economics still apply. At 60 pieces per style and colour, a petite range on one best-selling linen dress is a sensible test; a petite range across six styles in three colours is a forecast problem dressed up as a sizing decision. Launch the depth on the proven style, measure the sell-through, and extend only when the data supports it.

Next Step