Insights · Apparel & Fashion
We read real apparel listings, checking one thing: could an AI shopping assistant confidently tell a buyer whether a piece would fit them?
Apparel brands are, on the whole, good at describing what a garment is. Fabric composition, weight, cut and silhouette were almost always stated clearly — an assistant reading these listings can usually explain what a piece is made of and how it's meant to sit.
But the one detail that decides "will it fit me?" was almost never in the text a buyer — or an AI — actually reads.
This wasn't usually a case of the information being missing altogether. In most listings we checked, the fit answer existed somewhere on the page — a size chart image, a fit-finder widget, a size-guide tag. It just wasn't in the readable description. An assistant reading the listing text sees the fabric and the silhouette, but can't confirm the fit, so it hedges rather than confidently matching the piece to a buyer's body.
Why this happens
A size chart or fit-finder works well for a human shopper who can click through and compare. But an AI assistant reading a product description doesn't see that chart — it sees the words on the page. When the measurements and fit notes live only in an image or an interactive widget, the assistant is left with the size name (S/M/L) and nothing about how that size actually runs.
This holds across most of general apparel retail — even brands that write rich, detailed copy about fabric and construction still routinely leave fit out of the text. It doesn't hold everywhere: brands where fit-critical categories like bras are the whole product tend to publish full sizing in text, because the fit question can't be deferred to a chart. And it works differently again for made-to-measure pieces, where there's no standard size to chart in the first place — the fit question is answered by the process, not a measurement.
What to do about it
State garment measurements in the description text — chest, waist, length per size. Not just in the size chart image.
Add a plain fit note — "true to size," "runs small, size up," "relaxed, oversized fit." It's often cheaper to write than a measurement and answers the buyer's actual question.
State the model's height and the size they're wearing, if you photograph on a model. It's a cheap, partial fit cue that's rarely stated.
For fit-critical pieces — bras, fitted footwear — publish full sizing in the text itself. Where fit is the whole product, a chart alone isn't enough.
Keep stating fabric composition and weight — you already do this well. It's the strong foundation the fit detail builds on.
Keep describing the cut and silhouette — also done well. It tells an AI what the piece looks like; fit tells it who it's actually right for.
Squiggle reads your entire Shopify catalogue and shows you which products state the fabric and cut but leave fit for the buyer to guess — and what to add first.
Get my free Store Audit →No software to install. No account required. No obligation.
This finding is based on Squiggle's own reading of public apparel listings, scoped to new and standard-sized retail. It describes a pattern observed across the listings we checked, not a claim about the apparel industry as a whole, and no individual store is named. The finding does not hold the same way for vintage or one-of-one resale (where measurements are often the primary fit signal in text), made-to-measure apparel (where fit is answered by the process, not a published measurement), or intimates (where fit is the product). Public catalogue data only — no account access, sales data, or private information was used.