New research guide: The Confidence Economy, how AI is changing fashion commerce.

Read it →

Free Tool

What are returns actually costing you?

Enter four numbers. See the true annual cost of your return rate — and what reducing it by 5% is worth to your bottom line.

Your store
500
Completed orders per month
$75
Per completed order
28%
Fashion average: 25–40% · NRF Returns Landscape 2025
$12
Reverse logistics + inspection + restocking per item
Average: $8–18 per item · Richpanel 2026
70%
% of order value recovered when returned item is resold
Typical range: 60–85% · Richpanel 2026
Return rate data from NRF Returns Landscape 2025 and Richpanel 2026. All estimates are directional. Results are not guaranteed.
$57,960
Annual cost of returns
1,680 returns per year
Cost breakdown
$126,000
Revenue lost to returns
1,680 returns × $75
+$88,200
Resale recovery
70% of returned order value recovered
$20,160
Return processing costs
1,680 returns × $12
$57,960
Net annual cost
Lost margin + processing, minus recovery
If your return rate dropped 5 points
$6,750
Revenue recovered
Orders that stay converted
$3,600
Processing costs saved
Fewer returns to handle
$10,350
Total annual improvement
Revenue + processing combined
$863
Monthly improvement
Running monthly margin gain

What causes fashion returns (NRF, 2025)

~50%

Fit and sizing

Buyers ordered the wrong size, or the fit description didn’t match how the garment actually fits.

~25%

Visual expectation gap

The garment looked different in practice — colour, texture, drape, or proportion — than in the photography.

~25%

Occasion mismatch

The shopper bought for a specific occasion and the piece didn’t work for that context.

Source: NRF Retail Returns Landscape 2025. Approximate percentages. Vary by price point and category.

Up to 4×

conversion rate on styled sessions

25%

average AOV uplift in styled sessions

30%

fewer returns with virtual try-on

See your real return numbers in 14 days

Elara’s holdout report shows your actual lift — conversion, AOV, and return rate — versus a matched control group. No self-reported data.