The questions apparel shoppers ask that your product page must answer
Apparel shoppers ask the same handful of questions before every purchase. Answer them on the page, and returns and lost sales both drop.
Apparel has the highest return rate of any ecommerce category, and it isn't because shoppers are fickle. It's because product pages don't answer the questions a shopper actually has before checkout. Here's what those questions are, why the gaps cost you twice, and a checklist to close them, using a men's dress shirt as the working example.
Why this category bleeds the most
Apparel returns run around 25% industry-wide, with fast fashion near 29% and shoes over 31%, according to Eightx's 2026 return rate benchmarks. Fit is the dominant reason, not defects or damage. Multiple industry analyses put fit and sizing issues at roughly half to two-thirds of all apparel returns, and the pattern by gender is telling: men's returns skew toward "too small," women's skew toward "too large," which means the same vague size chart fails both audiences in opposite directions.
There's a second cost that doesn't show up in a returns dashboard: the shopper who never adds to cart because the page couldn't answer their question. A return at least converted once. A silent bounce never did. Both trace back to the same root cause: a product page that describes the item instead of specifying it.
The questions a shopper is actually asking
Before someone buys a shirt, a jacket, or a pair of pants online, they're running through a short mental checklist, whether or not the page gives them the answers.
| Shopper question | What they need to see | What most PIMs actually store |
|---|---|---|
| Will this fit my body type? | Fit name tied to a real measurement (chest, waist, sleeve) | A marketing label like "Modern Fit" with no numbers |
| Does their sizing run true, small, or large? | A brand-specific size chart, not a generic one | A linked-out PDF or nothing at all |
| What does this feel like against skin? | Fabric composition and weight (e.g., cotton percentage, GSM) | "Premium fabric" |
| Can I wash it or does it need dry cleaning? | Explicit care instructions | Missing, or buried in an image |
| Will it wrinkle after one wear? | Wrinkle-resistance / iron-free claim, verified | Absent |
| Does the color on screen match reality? | Named color plus a second reference (e.g., "French Blue, similar to a light chambray") | Hex-adjacent swatch name only |
| Can I dress it up and down? | Occasion tags (business, business casual, weekend) | Not attributed at all |
| What's the return policy if it doesn't fit? | Stated on the PDP, not three clicks away | Buried in a footer link |
None of these are exotic asks. They're the same questions a good in-store associate answers in fifteen seconds. The product page has to do that job with data instead of a person.
Case study: a men's dress shirt
Take a standard cotton dress shirt. Here's a typical raw feed record versus what an enriched, AI-agent-readable page actually needs.
Raw feed (as it arrives from most suppliers):
| Field | Value |
|---|---|
| Title | Men's Dress Shirt, Blue |
| Fit | Slim |
| Size | M |
| Material | Cotton |
| Care | See label |
Enriched attribute set:
| Attribute | Value |
|---|---|
| Fit | Slim (tapered waist, follows chest line; sizes down from Classic by roughly 2 inches at the waist) |
| Neck / sleeve | 15.5 / 34-35 (dual-number sizing, collar/sleeve format) |
| Chest measurement (flat, laid) | 21.5 in at size M |
| Fabric | 100% cotton, 120s two-ply, 115 GSM |
| Weave | Dobby |
| Collar style | Spread |
| Wrinkle performance | Non-iron finish, verified |
| Care | Machine wash cold, tumble dry low, or dry clean |
| Color | French Blue (light blue with subtle dobby texture) |
| Occasion | Business, business casual |
| Runs | True to size in collar; runs slim in body versus Classic fit of the same brand |
The difference isn't cosmetic. "Slim" alone tells a shopper nothing; "Slim, tapered waist, sizes down 2 inches at the waist from Classic" tells them whether to size up. This is exactly the kind of attribute Proper Cloth's own sizing reference treats as foundational: collar, sleeve, and fit type together, not fit type alone.
The "ask an AI" test
Try this: ask an AI shopping assistant to "recommend a slim-fit non-iron dress shirt in a 15.5/34-35 that won't need dry cleaning." A page that only lists "Men's Dress Shirt, Blue, Slim, M" gives the model nothing to match against. A page with dual-number sizing, a stated non-iron finish, and explicit care instructions gets surfaced, because the model can confirm every constraint in the query against a structured attribute rather than guessing from a title.
That's the same test a shopper runs manually when they scan a PDP for thirty seconds before either adding to cart or clicking back. If the AI can't confirm the fit, neither can the person.
Checklist: what every apparel PDP needs
- Fit name paired with a real measurement delta, not just a label
- Brand-specific size chart embedded on the page, not linked out
- Fabric composition and weight, not "premium" or "quality"
- Explicit care instructions in plain text
- Wrinkle, stretch, or shrink behavior stated when it matters to the category
- A named color plus a real-world reference point
- Occasion or use-case tags
- "Runs small / true to size / runs large" guidance, ideally sourced from actual return or review data
- Return policy visible on the page itself
Most catalogs are missing at least half of this list on any given SKU, and the gaps aren't evenly distributed. A shirt might have great fabric data and no fit guidance; a jacket might have the opposite. That inconsistency is what erodes shopper trust across a catalog, one SKU at a time.
Anglera scores every product page against gaps like these automatically, then fills them from your existing supplier feeds, spec sheets, and past returns data rather than requiring a rip-and-replace of your PIM. It plugs into whatever system already stores your catalog and keeps attributes like fit, fabric, and care current as SKUs turn over, so both shoppers and AI shopping agents get an answer instead of a guess.