AI Try-On Tech Tackles Retail's $850B Returns Problem

A wave of AI startups is targeting one of retail’s most expensive habits

Online returns cost U.S. retailers $849.9 billion in 2025. Apparel is the worst offender. And a new class of AI virtual try-on startups — Catches, Genlook, AIUTA among them — is racing to close the gap between “I think this fits” and “I know this fits” before the UPS label gets printed.

If you run a small online retailer anywhere in Appalachia, you have already felt the cost. Free return shipping is no longer optional, reverse logistics eat your margin, and the items that come back rarely sell at full price again. The question is not whether AI fit tools matter for your business. It is which ones are ready for a five-person shop today.

The news

The National Retail Federation reports that 15.8% of all retail sales were returned in 2025. Online returns hit 19.3%. The total bill: nearly $850 billion. CNBC’s coverage of the emerging try-on ecosystem calls returns “silent killers” of retailer margins — the costs do not show up on the sales report, but they quietly vaporize profit on the back end.

Who is fighting the returns problem with AI

A few names worth knowing:

• Catches — Generates a “digital twin” of the shopper. Claims up to 10% lift in conversions and 20-30x ROI for brand partners.
• Genlook — Integrated into Shopify’s ecosystem so storefronts can add try-on without custom development work.
• AIUTA — Powers ASOS’s try-on experiments. PYMNTS reports that ASOS has seen a 160-basis-point reduction in returns through this approach.
• Google — Moved its Doppl try-on tool into Search on April 30, covered in our piece on the Search rollout.
• Amazon and Adobe — Building integrated try-on inside their own stacks for partner brands.

The problem by the numbers

• 15.8% of 2025 retail sales returned — $849.9B total
• Online return rate: 19.3%
• Gen Z averaged nearly 8 online returns per person in 2025
• 82% of consumers consider free returns essential before they buy

Why this matters for small retailers

Big retailers publish return rates and absorb the cost across millions of orders. Small retailers feel every returned jacket. The margin math is brutal: a $60 sweater with a $9 reverse shipping cost, a $6 restocking cost, and a 20% markdown on resale nets less than half the original order value. A 5-10 percentage point drop in returns is the difference between a profitable quarter and a break-even one.

The bigger shift is cultural. Shoppers have been trained by fast fashion to treat their closet like a dressing room — order three sizes, keep one. Guggenheim analyst Simeon Siegel put it plainly to PYMNTS: “Proactively minimizing returns can be a meaningful driver of profitability.” Virtual try-on is the most visible lever retailers have to change that behavior at the moment of purchase.

For Appalachian online retailers — especially the growing number of artisan, outdoor gear, and apparel shops selling nationally through Shopify or Etsy — the stakes are real. You do not have a warehouse operations team optimizing reverse logistics. Every return is handled by the same person who packed the order.

Our take

The technology is promising, but the case for a small retailer adopting try-on today is narrower than the headlines suggest.

What is real. Shopify-native integrations like Genlook genuinely do lower the activation cost for small storefronts. A Shopify shop can enable try-on without engineers. And Google’s Doppl-in-Search rollout means shoppers will experience try-on on your products whether you pay for a vendor or not — provided your Google Merchant Center feed is clean.

What is hype. PYMNTS points out that no vendor has published a definitive return-rate reduction tied solely to virtual try-on. ASOS’s 160-basis-point figure is encouraging but overlaps with other initiatives. Startups quoting “20-30x ROI” are projecting, not measuring.

The bottom line: If you sell apparel or gear online and already use Shopify, turning on a native try-on integration is a low-cost experiment. Everything else — custom deployments, “digital twin” platforms — is premature for a small shop.

What the coverage is missing

Two underreported angles matter for the Appalachian market.

First, product photography is the real gate. Try-on engines need high-quality, consistent, model-agnostic product shots. Most small retailers still shoot on a friend in a field or a mannequin in a basement. The try-on tech fails on inconsistent inputs. Upgrading photography is a bigger lever than picking a vendor.

Second, returns are a customer service problem, not just a fit problem. A shopper who is unsure about sizing will message your support channel before ordering. If no one answers, they either order three sizes or abandon the cart. Automating that pre-purchase conversation with an AI intake widget like Hollr or a Five-Star review and support agent often catches the fit question earlier in the funnel than a try-on widget would. We have seen small retailers cut returns meaningfully just by answering pre-purchase sizing questions within five minutes.

What you should do this month

If you sell apparel on Shopify:

1. Clean up your product imagery — consistent lighting, plain backgrounds, full garment visibility
2. Enable a native try-on integration (Genlook, Shopify’s own tools, or your theme’s built-in options)
3. Make sure your Google Merchant Center feed is up to date so the Doppl-in-Search rollout picks up your products automatically

If you sell other categories (outdoor, home, craft):

1. Skip the try-on conversation — it does not apply
2. Focus on returns reduction through better sizing/spec information and automated pre-purchase support
3. Track your return rate monthly so you can tell whether anything you change is working

If you are deciding whether to invest:

1. Pull your last 90 days of returns data — count items returned, reason codes, and net margin lost
2. If fit is the top reason code, try-on is worth a pilot
3. If “changed mind” or “quality not as expected” lead, try-on will not fix it — the issue is upstream in product or expectations

Watch for

• Native Shopify and WooCommerce try-on modules becoming standard by Q3 2026
• A shakeout among try-on startups — expect consolidation as Amazon and Google absorb the category
• New NRF return data later this year showing whether 2026 moved the needle

Where this lands

AI virtual try-on is going to matter for small online apparel retailers. It is not going to matter immediately, and it is not going to matter for every category. The vendors most likely to win the small-business market are the ones that plug into Shopify or Google Search without requiring an IT budget.

The returns problem is not going anywhere. But for most Appalachian retailers, the highest-ROI fix for 2026 is still the basic one: answer pre-purchase questions quickly, photograph products well, and be specific about sizing. Try-on is a nice addition once those are in place.

Need help figuring out which AI tools are worth your time and which are noise? Get in touch — we help small businesses in the Appalachian region adopt the AI that actually moves the numbers.