What the Data Actually Says

The conversion metrics are real. According to Shopify's 2024 merchant survey, fashion brands using AR try-on features reported an average 27 percent increase in conversion rates compared to static image catalogs. Some implementations achieve even higher add-to-cart lifts. Internal data from early adopters of Genlook shows that shoppers who engage with the VTO widget and generate an image see a 35 percent higher add-to-cart conversion rate compared to those who simply open the widget but don't use it. But here is where the story diverges. According to 2025 Forrester Research analyzing retail technology ROI, retailers implementing comprehensive solutions observed return rate reductions averaging 23 percent across apparel categories, with some achieving up to 40 percent reductions for specific product types. That 23 percent average is meaningful, but it is not the 27 to 35 percent lift you see in add-to-cart metrics. The gap is not a rounding error. It suggests something important: a customer who adds to cart after using VTO is not the same as a customer who confidently buys what they added.

The Difference Between Confidence to Order and Confidence in Fit

Add-to-cart metrics measure decision closure. A shopper saw the product visualized on a body, felt more certain about the decision, and put it in their cart. That is a real outcome. But what the metric does not capture is whether that customer felt confident about fit itself, or just confident enough to try it. Apparel return rates tell the real story. The ICSC's 2024 survey reveals that apparel purchased online in the US has a return rate of 22 percent, compared to just 6.2 percent for in-store purchases. The gap between online and offline is not about whether the customer was willing to buy. It is about what happens after the item arrives. Physical stores allow customers to try on. Online shopping forces prediction. VTO is prediction technology. It shows what a garment might look like on a customer's body. But showing is not the same as guaranteeing. Fabric behaves differently when worn versus when visualized. Seam placement, stretch, and drape cannot be fully captured in any two-dimensional or static representation. A customer who sees a garment fit well on a visualization might still experience surprise when the physical item arrives.

Why Bracketing Complicates the Picture

The real obstacle to lower return rates is not just visualization. It is behavior. Gen Z leads the way in bracketing behavior, with 51 percent of them having done so — and they also spend more and return more, with an average of 7.3 returns per shopper and an average order volume of $174. Bracketing is the practice of intentionally ordering multiple sizes or colors of the same item with the plan to return all but one. VTO changes the add-to-cart moment — a customer can see the garment fit before adding to cart. But VTO does not change the calculation behind bracketing. A shopper with a 10 percent size variance between brands still might order two sizes to be certain. VTO might make them slightly more confident, but it does not eliminate the variance. That customer still generates two orders and one return, even with VTO enabled. The add-to-cart lift looks good in the metrics. The return rate barely budges.

What Actually Reduces Returns at Scale

The data reveals a pattern. Return rate reductions happen when fit prediction moves from what a garment might look like to what it will actually measure on a customer's body. Sizing accuracy is the operative phrase. VTO that shows how a garment will look does something different than VTO that predicts how a garment will fit. One is visual. One is dimensional. The 23 percent average return reduction that Forrester tracked likely reflects implementations that moved beyond pure visualization. Implementation quality is code for accuracy of the underlying fit model. A solution that maps a customer's actual body dimensions to how fabric will sit on those dimensions produces different outcomes than a solution that simply shows the garment on a realistic body.

The Categories Where VTO Moves the Needle

Not all apparel returns equally. Swimwear and lingerie return rates hover between 30 to 35 percent — some of the highest in fashion. These items require exact fit and comfort, which is hard to judge without trying on. Structured garments like blazers and tailored pants also see outsized benefits from VTO. In these categories, fit is measurable. Chest width, sleeve length, and waist gap are discrete variables. A VTO that accurately maps those dimensions produces meaningful return reduction. In categories where fit is subjective or where fabric behavior dominates the experience, the return reduction is smaller. Categories presenting particular challenges include highly draping fabrics, stretch materials, and layering pieces. The 23 percent average return reduction is not universal. It is the average across a distribution where some products and categories see 40 percent reduction and others see near zero. Brands that see the higher reductions are those where VTO answers a specific, measurable question: will this garment fit these dimensions.

The Bracket Buying Question VTO Cannot Fully Solve

Here is what neither add-to-cart metrics nor even 40 percent return reductions reveal. Some of that add-to-cart lift is not prevented returns. It is prevented bracket buying. If a customer was uncertain and planning to bracket, VTO might make them confident enough to order one size. That is a prevented return. But if a customer was always going to try one item in multiple sizes because their uncertainty crossed multiple size categories, VTO solves only part of the problem. The customer still might order two sizes, knowing one will fit better than the other. Close to two-thirds of consumers admit to participating in at least one costly returns behavior, from wardrobing and bracketing to sending back different items or empty boxes. Not all of that behavior is driven by sizing uncertainty. Some is driven by style preference, by the desire to compare, or by the low cost of return options.

The Real Gap: More Orders, Not Fewer Returns

The reconciliation between the 27 to 35 percent add-to-cart lift and the 23 percent average return reduction is this: VTO increases orders. It does not reduce returns at the same rate because some customers who would not have ordered without VTO now order and then return. That is not a bad outcome. It is a different outcome. More orders at the same return rate is a net positive for most retailers. It is margin positive. But it is not the same as confidently saying VTO solves the returns problem. It solves the conviction problem. It moves customers from uncertain browsers to confident buyers. Some of those buyers then return for reasons VTO cannot address.

For a customer buying a structured garment where fit is the only question, VTO that accurately maps body dimensions to garment dimensions solves the returns problem. For a customer buying a draped garment where fabric behavior dominates, or where style preference is the real uncertain variable, VTO reduces returns but does not eliminate them.

What This Means for Implementation

If a brand is seeking to reduce returns, the VTO metric to care about is not add-to-cart lift. It is the return rate on orders placed after VTO engagement versus orders placed without. Even that metric has a built-in limitation. Customers who use VTO are different from customers who do not. They are customers who felt uncertain enough to try visualization. That makes them a higher-risk cohort to begin with. A 20 percent return reduction for that cohort is significant. A 40 percent reduction in a structured category is exceptional. But expecting VTO alone to move a 22 percent overall online return rate to single digits assumes that fit was the only driver of returns. It was not. Fit is one driver. Style preference, color interpretation, and sizing variance across brands are others. VTO addresses fit. It addresses visualization. It does not address subjective preference for how a garment looks or feels when worn.

The Honest Number

Virtual try-on increases the likelihood that a customer who uses it will complete the purchase. Deloitte highlighted that retailers using AR or AI see a 40 percent increase in conversion rates and a 20 percent increase in average order value compared to those not using these technologies. That is proven and real. Virtual try-on also reduces the return rate on those orders, but by a smaller margin than the conversion increase. When you increase your customer pool through higher conversion, your return rate reflects a broader distribution of customer preferences and fit needs.

The answer to the headline question is both — but at different magnitudes. The add-to-cart lift is real and immediate. The return reduction is real but smaller, because it is fighting against bracketing behavior, subjective preference, and the gap between prediction and physical experience. For apparel brands, the ROI case for VTO is not about returns alone. It is about conversion lift net of return impacts. VTO is a solution to the confidence crisis. The returns crisis is driven by sizing variance, bracketing, and the subjective nature of fit. Those are harder problems.