Fashion has a return problem unlike any other e-commerce category. Industry data consistently puts apparel return rates far above the e-commerce average, and the overwhelming majority of those returns come down to one thing: fit. "Too small," "too big," "not what I expected on." Every one of those is a question the customer had before buying — they just couldn't get it answered.
That makes sizing a double tax on a fashion store: shoppers who can't get a fit answer either abandon (lost sale) or buy-to-try (a return you pay for). This guide is about closing that gap with AI chat that answers fit, fabric and styling questions from your WooCommerce catalog before the order is placed.
Why sizing hurts twice
| Customer behaviour | What it costs you |
|---|---|
| Unsure on size → leaves | Lost conversion, no signal why |
| Unsure on size → buys two sizes to try | Guaranteed return + reverse shipping |
| Unsure on fabric/fit → buys → disappointed | Return + a refund + maybe a bad review |
The thread through all three is an unanswered question at the moment of decision. Solve that and you lift conversion and cut returns at the same time — rare in e-commerce, where most levers trade one against the other.
What AI chat does with a fit question
Customer: is the linen blazer true to size? i'm usually a M but between sizes on top
AI: The Linen Blazer runs slightly relaxed, so most customers stay with their usual size. Since you're between sizes and it's a layering piece, an M will give you room to wear it open over a shirt; size down to S only if you want a fitted look. The fabric is 100% linen — breathable but it does crease, which is part of the look. Want me to show it next to the structured blazer, which holds shape better?
Customer: show me both
AI: [compares the two] …
In one exchange the AI answered fit, fabric, and surfaced an alternative — the things a good in-store associate does. It pulls from your product data and size guides, and because the search is semantic, it understands "between sizes," "true to size," and styling intent rather than just matching keywords.
This is also cart recovery
Sizing doubt is one of the top reasons a full cart never checks out. Answering it in-session is exactly the in-the-moment recovery mechanism we cover in recovering abandoned carts with AI chat — and pre-purchase confidence is the highest-quality way to do it, because a sale made with the right size in the bag is a sale that stays sold.
Setting it up on WooCommerce
- Sync your catalog and size-guide content so the AI can reason about fit, fabric and care — setup takes about 15 minutes.
- Add custom instructions capturing your brand's fit notes, e.g. "Our outerwear runs relaxed; recommend usual size unless the customer wants a fitted look. Always mention fabric and care for linen and wool."
- Enable human handoff for styling-heavy or VIP queries you'd rather a person took.
The bigger picture
For apparel, the return rate is the margin conversation — and most of it is preventable, because most of it is a fit question that went unanswered. A catalog-aware AI assistant answers those questions at the exact moment they'd otherwise cost you, converting hesitant shoppers and keeping the ones who buy.
See the full breakdown for apparel stores on the fashion & apparel use-case page →, or try ChatAxon free for 14 days — no credit card required.