Size and Fit Guidance

Competing size systems

The first obstacle is that different markets label sizes differently. A shoe or garment carries one number in the EU, another in the UK and another in the US, and shoppers crossing borders — increasingly common in cross-border ecommerce — face a translation problem before they even consider fit. The table below is illustrative only, to show that the systems diverge, not to assert exact conversions.

Brands break their own rules

Even within a single system, size is not consistent, because a labelled size is a promise each brand keeps differently. One brand's medium is another's large; one label runs small, another generous. Shoppers cope by learning their size brand by brand, which is exactly the knowledge that does not transfer when they try something new.

This inconsistency defeats naive guidance. A system that maps every brand's "size 9" to a single canonical 9 will mislead, because those nines are not the same shoe. Useful guidance has to model each brand's real behaviour, not just its labels.

Fit is more than a number

Size and fit are not synonyms. Two garments of identical size can fit completely differently depending on cut, silhouette and intended shape — slim versus relaxed, high-rise versus low, structured versus draped. A number captures a rough scale; it says little about how a piece sits on an actual body.

Fabric compounds this. Stretch, weight and construction all change how a size wears, so the same measurements feel different in rigid and elastic materials. This is why sizing resists a single lookup table and why it overlaps with — but is not solved by — the visual preview of try-on.

The returns cost

Sizing failures are expensive. Fit and size are among the most common reasons fashion items come back, and every return carries shipping both ways, processing, and often a markdown or write-off on the returned item. High return rates can erode the margin on a category entirely.

Returns also carry an environmental cost that shoppers and regulators increasingly notice, and a customer-experience cost, since a bad fit sours the whole purchase. McKinsey and Business of Fashion have both identified returns as a structural drag on fashion-ecommerce profitability. Better sizing is one of the few levers that improves margin and experience at once.

Structured size data as the fix

The path forward is to treat size as structured data rather than a display string. That means capturing the size system, actual garment measurements, fit characteristics and fabric behaviour as discrete, machine-readable fields — the difference between knowing a label says "M" and knowing the chest measurement, cut and stretch behind it.

With structured size data, systems can genuinely help: translating across markets, comparing a new brand to what a shopper already owns, and flagging when something runs small. Without it, guidance is guesswork dressed up as advice. This is the same data-quality discipline that runs through size availability and the wider Swibblr thesis that discovery rests on clean data.

Localization and the cross-border case

Sizing is where localization stops being a translation exercise and becomes a data one. A shopper in one country buying from a retailer in another needs sizes presented in a system they understand, converted reliably, with fit expectations set for that market. Getting this wrong is a direct cause of cross-border returns.

As international fashion ecommerce grows, size localization becomes a competitive edge rather than a nicety — a theme we develop in localization in fashion search and the broader fashion-ecommerce trends. Sizing done well quietly removes one of the biggest reasons a cross-border purchase fails.