Solutions · Fashion & ApparelApparel, footwear & luxury

“Size 12” is not a measurement. It is a marketing decision.

Two garments labelled M can differ by six centimetres across the chest, and both are correct. Every shopper choosing a size without published measurements is guessing — and fit is the leading return reason in apparel.

style_code
GRV-MC-241
name
MERINO CREW NECK SWEATER
colour
Navy / Oat / Moss / Rust
size
XS S M L XL XXL
composition
Merino wool
body_measurements
not supplied
care_symbols
not supplied
fibre_percentages
not supplied
538

Attributes required per variant · knitwear

Vertical
Apparel & footwear
Specimen
GRV-MC-241
Standards
ISO 3758 · EU 1007/2011
Matrix
4 colours × 6 sizes
Completeness
5 / 38
Specimen · one style, two states

"Merino wool" is not a composition

EU textile labelling requires fibre percentages by weight. This garment is 85/15 with polyamide, and the second fibre is the one that changes care and durability.

No body measurements

Only label sizes. A shopper cannot tell whether your M matches the M they own, which is the single most common reason apparel comes back.

Four colours × six sizes

Twenty-four variants inherit every gap above. A missing attribute here is not one error, it is twenty-four.

GRV-MC-241 · AS RECEIVEDstyle-level row

Merino Crew Neck Sweater

style_code
GRV-MC-241
name
MERINO CREW NECK SWEATER
colour
Navy / Oat / Moss / Rust
size
XS S M L XL XXL
composition
Merino wool
body_measurements
— not supplied —
care_symbols
— not supplied —
fibre_percentages
— not supplied —

3 of 8 fields empty · each gap repeats across 24 sellable variants

Label sizes with no measurements · a composition that breaches EU fibre labelling · multiplied by 24

Attributes per variant · knitwear5 of 38 populated

5

Attributes supplied at style level and inherited.

38

Attributes needed per variant for filtering, sizing and compliance.

PL. 01The standards

Three that decide whether a garment sells or comes back.

Measurements against label sizes

A label size is a brand decision, not a dimension. Two garments both labelled M can differ by six centimetres across the chest, and both are correct.

Which means a shopper choosing a size is guessing unless you publish measurements. Fit is consistently the leading return reason in apparel, and it is the one return reason that is purely a data problem rather than a product problem.

Half-chest

Laid flat, underarm to underarm

Body length

High point of shoulder to hem

Sleeve length

Shoulder point to cuff, convention named

Grading

The increment between consecutive sizes

Scale mapping

EU / UK / US / JP, per garment type

Fit descriptor

Relaxed, regular, slim — defined by measurement, not adjective

PL. 02Failure modes

Six ways apparel data becomes a returns line.

01

The fit return

The garment is exactly as described and comes back anyway, because the shopper had no way to compare your M to the M in their wardrobe. You pay shipping twice and lose the sale.

returns
02

Unfilterable colour

"Oat", "Bone" and "Sand" are three of your colours and none of them appear under a beige filter. The shopper who wanted beige never sees any of them.

lost discovery
03

Variant drift

A colourway is created by duplication rather than inheritance. Six months later it has a stale composition and nobody knows which is right.

compliance risk
04

Care as a sentence

"Machine washable" in the description and no ISO symbols. It does not render in market-specific templates and cannot be validated by anyone.

rejections
05

Composition rounding

Percentages that do not sum to 100, or a blend declared as a single fibre. A labelling breach that is trivial to detect and routinely shipped.

regulatory
06

Assortment mismatch

Sizes published to a market you do not carry them in, generating orders you cannot fulfil and cancellations that damage seller metrics.

cancellations
The multiplier

Every failure above happens once at style level and lands twenty-four times in your catalogue. This is why apparel data debt compounds faster than any other vertical, and why fixing it at style level with real inheritance is worth considerably more than fixing twenty-four listings by hand.

PL. 03Rollout

One season, then every season.

Apparel rollouts are best scoped to a season rather than a catalogue, because the intake cadence is the thing that has to work.

01
Week 1

Matrix audit

We measure attribute completeness at style, colourway and size level, and identify variants that were duplicated rather than inherited.

What you provide

A product export including variants · no integration

02
Weeks 2–3

Measurement and composition

Body measurements, fibre percentages and care codes are extracted from tech packs, spec sheets and care labels, with a source per value.

What you provide

Tech packs or supplier specs, in whatever form you hold them

03
Weeks 3–4

Colour and size mapping

Brand colour names are mapped to filterable families, and label sizes to regional scales per garment type, keeping your naming on the page.

What you provide

Your colour card and size charts, if they exist as documents

04
Ongoing

Seasonal intake

Each new season is resolved on the same rules, so a range built in June does not disagree with one built in January.

What you provide

Nothing — new styles inherit the resolved model

Next

Find out which returns were avoidable.

Send one season plus your return reasons. We will resolve measurements, composition and colour families for the range, and show you which fit returns correlate with styles that published no measurements at all.

One season · Return reasons optional · Findings in 10 working days