Guide
Sources re-checked on July 26, 2026.
Probably, but nobody has proven a number - and you should distrust anyone who gives you one. What is well established is the cause: size and fit is the single largest reason for online apparel returns, cited by 53% of respondents in Coresight Research's survey of US apparel retailers, well ahead of colour at 16%. So a tool that helps a shopper judge size before checkout is aimed squarely at the dominant cause.
What is notestablished is how much any specific virtual try-on app moves any specific store's return rate. The precise percentages circulating in this category are overwhelmingly vendor-reported and not independently audited. The only figure that will ever be true for your store is the one you measure on it.
Three findings we opened at the source, each with what it does not prove.
US shoppers were projected to return $849.9 billion of merchandise during 2025.
National Retail Federation with Happy Returns, published October 2025
What it does not prove: This is all US retail, not fashion e-commerce, and it is a projection rather than a settled figure.
NRF returns research, via Digital Commerce 360Size and fit is the single largest reason for online apparel returns, cited by 53% of respondents - ahead of colour at 16% and damage at 10%. The same research put the average US online apparel return rate at 24.4%.
Coresight Research, survey of decision-makers at US apparel brands and retailers
What it does not prove: Self-reported by retailers rather than measured from return logs, and US-only. Your own returns data is the better number if you have it.
Coresight Research on apparel returnsAmong shoppers who used augmented-reality try-on, 80% said they felt more confident in their purchase and around two-thirds said they were less likely to return it.
Snap and Publicis Media, research conducted by Alter Agents among 4,028 shoppers in the US, UK, France and Saudi Arabia, published June 2022
What it does not prove: This is stated intent, not measured returns, it is four years old, and AR camera lenses are a different technology from AI garment try-on. Treat it as directional only.
Snap and Publicis Media study, via Retail DiveA shopper returning a dress because it did not suit them and a shopper returning it because it was two sizes off are two different problems, and only one of them is solved by seeing a picture.
Most virtual try-on apps generate an image from a shopper photo. That answers "what does this look like on me?" and it genuinely helps with the styling half of returns. But it does not collect measurements, so it cannot answer "which size do I order?" - and per Coresight, that is the half causing most of the damage.
This is the practical filter when you are evaluating apps. Ask whether the tool collects body measurements, and whether it outputs a size rather than only a picture. Our roundup of Shopify try-on apps lists which ones document a size recommendation - of the six we verified, one does.
The related trap is bracketing: shoppers who deliberately order two or three sizes intending to return the rest. Bracketing is a sizing-confidence failure that shows up as a conversion win and a returns disaster. If you are measuring try-on impact, watch units per order alongside return rate, or a drop in bracketing will look like a drop in sales.
Four steps that separate a real effect from seasonality and self-selection.
Pull your fit-related return rate per category for the last two full quarters, before you install anything. Category matters: a dress and a beanie do not return alike.
Compare orders where the shopper generated and viewed a try-on against orders where they did not, on the same products in the same period. Comparing this quarter to last quarter will just measure your seasonality.
Shoppers who try on are more engaged to begin with, so some of any gap is selection rather than the tool. If you can, hold the widget back from a random slice of traffic for a few weeks and compare like for like.
A 30-day return policy means return data from the last 30 days is incomplete. Read the cohort only after its window has closed.
All sources re-checked July 26, 2026
DressApp's Pay Per Sale plan has no monthly fee and no try-on cap, so you can run a proper before-and-after on your real products without committing budget to an unproven channel. Every try-on carries a recommended size and a fit description per size, so you are testing the sizing mechanism and not just the picture.