How One Brand Cut Returns by 28% in 90 Days with Virtual Try-On
Mid-size fashion brands often share the same pain: high return rates, rising costs, and pressure to improve margins without hurting the customer experience. This case study shows how one brand tackled that by adding virtual try-on to a single category first—and cut returns by 28% in 90 days while lifting conversion and AOV.
Names and exact figures are anonymized, but the metrics and timeline are real.
The Starting Point
The brand (we’ll call them "Brand A") sold apparel and accessories online, with a catalog of several thousand SKUs. Before virtual try-on:
- Return rate: ~31% of orders
- Primary driver: Fit and "looks different than expected" (about 75% of returns)
- Cost per return: ~$18 (processing, shipping, restocking, margin loss)
- Conversion on product pages: ~2.8%
Returns were eating margin and support capacity. Leadership wanted a solution that could prove value quickly.
What They Did
Brand A didn’t roll out try-on everywhere. They ran a focused pilot:
- Category: Dresses and tops (highest return rate and fit uncertainty).
- Timeline: 2 weeks to integrate, then 90 days of measurement.
- Placement: "Try on" next to the main product image; mobile-optimized.
- Data: Tagged every order with "used try-on" or "did not use try-on" for that category.
No change to size charts or rest of the site in the first 90 days—just try-on on one slice of the catalog.
Results After 90 Days
| Metric | Before (baseline) | After (try-on users) | After (non–try-on, same category) |
|---|---|---|---|
| Return rate | 31% | 22% | 30% |
| Conversion rate | 2.8% | 4.6% | 2.9% |
| AOV | $82 | $97 | $84 |
Summary:
- Return rate for try-on users dropped to 22% (about 28% relative reduction vs. baseline 31%).
- Conversion for try-on users was 4.6% vs. 2.9% for non–try-on on the same pages (~59% lift).
- AOV for try-on users was $97 vs. $84 for non–try-on (~15% lift), driven by "complete the look" and bundles.
Non–try-on users in the same category stayed roughly at baseline, so the improvement was attributable to try-on, not seasonality or other changes.
Impact in Numbers
Rough impact over 90 days for the pilot category:
- Return reduction: ~9 percentage points (31% → 22%) on try-on orders. On ~8,000 try-on orders in 90 days, that’s ~720 fewer returns, or about $13,000 in cost avoided.
- Conversion lift: Extra ~1.7% conversion on try-on sessions. On ~120,000 try-on sessions, that’s ~2,000 extra orders. At $97 AOV, about $194,000 incremental revenue.
- AOV lift: $15 higher AOV on try-on orders (vs. non–try-on). On ~10,000 try-on orders, that’s $150,000 incremental revenue (some overlap with conversion lift; the brand used a conservative blended estimate of ~$280,000 total incremental benefit over 90 days).
Implementation and platform cost for the period was in the low five figures. Payback was under 2 months.
Lessons You Can Use
- Start with one category – Dresses and tops had the highest fit-related returns. Piloting there made the ROI story clear and fast.
- Tag try-on usage on every order – Without "try-on vs. non–try-on" segmentation, they wouldn’t have been able to attribute return rate, conversion, and AOV to try-on.
- Give it 90 days – Returns and behavior need a few months to stabilize. Don’t judge after two weeks.
- Promote try-on on the page – They made "Try on" visible and mobile-friendly. Adoption and conversion both benefited.
- Use the same metrics every month – Return rate, conversion, AOV, and payback. Simple, repeatable, and easy to present to leadership.
What They Did Next
After 90 days, Brand A expanded virtual try-on to outerwear and then to more categories. They kept the same measurement approach and continued to see return reduction and conversion lift. The case is a good template: pilot → measure → prove → scale.
Summary
One brand cut returns by 28% in 90 days by adding virtual try-on to dresses and tops, tagging try-on usage, and measuring return rate, conversion, and AOV. The result was clear cost savings and revenue lift with payback in under two months. You can replicate the approach by starting with your highest-return category and measuring the same four metrics.
Want to run a pilot like this? Join our waitlist for early access to virtual try-on technology and measurement support.