Virtual fitting rooms are more than an image: what fashion brands can actually measure

For most fashion brands, the value of a virtual fitting room is still framed in simple terms: does it help sell more clothes?

It is a reasonable question. Online fashion remains one of the most difficult categories to convert because customers are asked to make a highly personal decision without the information they would normally get in a physical store. They can see the garment. They can read the description. But they still cannot answer the question that often matters most: What will this actually look like on me?

Virtual try-on is designed to close that gap. But focusing only on its visual output misses a potentially more valuable part of the technology. Every time a shopper decides to try a product virtually, they reveal something about their purchase journey. They signal interest. In some cases, they signal hesitation. And what happens after the interaction — whether they continue browsing, add the item to their cart or leave the site — gives the brand a clearer picture of how visual uncertainty affects buying decisions. The image is what the customer sees. The interaction behind it can become a source of data.

A new way to measure uncertainty

Fashion e-commerce has plenty of customer data, but it does not always explain why a shopper failed to buy.

A product page may have strong traffic and low conversion. Is the product simply unappealing? Is the price too high? Does the customer need more information about the fit? Or were they interested enough to consider the purchase but not confident enough to complete it? Traditional analytics can show the drop-off. They are less effective at identifying what happened immediately before it.

Virtual try-on introduces another measurable action into that journey. A high try-on rate, for example, should not automatically be interpreted as a sign of purchase intent. It may also indicate that a particular category creates more uncertainty than others. Structured jackets, fitted dresses or products with unusual proportions may generate more try-on interactions precisely because shoppers find them harder to evaluate from standard photography.

That distinction matters. It allows brands to move beyond asking which products are popular and start asking where customers need the most reassurance. Research increasingly supports the idea that the value of virtual try-on is closely tied to how it affects the decision-making process. A 2024 systematic review of 69 studies on virtual try-on in fashion found that consumers' responses are shaped by factors including perceived usefulness, emotional and utilitarian value, technological gratification and attitudes toward the technology. In other words, virtual try-on does not simply add another image to the shopping journey. Its impact depends on whether it makes the product easier to evaluate and the decision easier to make.

The metric that matters is what happens next

The most useful measurement begins after the virtual try-on interaction. Did the customer add the product to their cart? Did they purchase it? Did they continue exploring similar items? How does their behaviour compare with that of shoppers who viewed the same product but never used the tool?

This is where virtual try-on can be evaluated as a business feature rather than a novelty. In one recent example published by Business of Fashion, DressX reported that users who engaged with its virtual try-on were around three times more likely to add an item to their cart and had a 50% higher purchase conversion rate than users who did not engage with the feature. The same users also viewed seven times more product listings.

The precise numbers will vary between retailers and technologies. More importantly, correlation should not automatically be confused with causation: customers who choose to use a try-on tool may already be more engaged shoppers. That is why brands should test the technology properly. Controlled experiments, cohort comparisons and product-level analysis are far more useful than simply reporting how many people clicked the feature.

Fashion retailer DIDI, for example, tested virtual try-on in an A/B experiment. Shoppers with access to the tool returned items 13.1% less often than the control group, while conversion was 3.5% higher. This is the difference between knowing that customers like a feature and understanding whether it changes commercial outcomes.

Virtual try-on can reveal where product pages are failing

The commercial impact can also be measured more directly. In looksy.tech projects shoppers who interact with the technology have shown up to a 21% increase in purchase conversion and a 17% increase in average order value. AI-powered size recommendations can also help reduce size-related returns by up to 10%.

These metrics matter because they show that the value of virtual try-on does not end with engagement. Brands can track how the technology affects the entire path from product evaluation to purchase — and identify where it has the strongest commercial impact.

The most interesting insights, however, may come from products where try-on usage is high but conversion remains low. That pattern suggests the visual experience alone is not solving the customer's problem. Perhaps the shopper can see how the garment looks on them but still lacks confidence about sizing. Perhaps they like the product but find the price difficult to justify. Or perhaps delivery terms, returns policies or product availability become the final barrier.

In this sense, virtual try-on can help brands distinguish between different forms of purchase friction. The technology cannot tell a brand exactly what a customer is thinking. But it can provide an important behavioural signal that traditional page analytics do not capture: the customer was interested enough to actively personalise the product experience.

If they still leave, the next question becomes more specific. The same logic can be applied at the category level. If virtual try-on has a significant effect on conversion for dresses but almost no effect for knitwear, a brand learns where the confidence gap is most commercially significant. That insight can inform where to invest in better product photography, more detailed size guidance, styling content or other interactive tools.

Returns are another part of the picture

For fashion retailers, the commercial impact of uncertainty does not end at checkout. According to McKinsey's survey of apparel companies, around 70% of returns were linked to poor fit or style. The firm argues that shopping tools that help customers make better choices can therefore play a role in return management, even though the impact of more advanced technologies has historically varied by category and implementation.

That creates an important measurement opportunity for virtual try-on. Brands should not ask only whether users of the tool convert more often. They should also track what happens after the purchase. Do these customers return items less frequently? Are certain categories more affected than others? Does virtual try-on reduce returns caused by unmet expectations, while having little effect on returns related to quality or delivery?

A recent systematic review of virtual try-on research also highlights the technology's potential to improve product evaluation and purchase decisions, while noting that results depend heavily on the quality and realism of the experience. Virtual try-on is not a universal solution to fashion returns: clothing with complex fit requirements remains particularly difficult to replicate digitally. For brands, that means the return rate should be treated as part of a broader measurement framework, not as a guaranteed outcome.

The data can be more valuable than the feature itself

The long-term opportunity may be to use virtual try-on data not simply to evaluate the tool, but to better understand the customer. Imagine a retailer discovers that shoppers repeatedly use virtual try-on on higher-priced outerwear but rarely on basics. Or that first-time visitors are much more likely to engage with the feature than returning customers. Or that try-on users are more likely to compare several products before making a purchase.

These patterns can help answer questions that extend beyond the fitting room itself. Which customers need more reassurance? Which products are hardest to evaluate online? At what point in the journey does visualisation have the greatest impact? And where does the customer still need something else?

For fashion brands trying to improve their own e-commerce economics, this matters because the cost of uncertainty is spread across the entire funnel. It can mean abandoned product pages, lower conversion, multiple-size orders and expensive returns. Virtual try-on creates a measurable interaction at the point where much of that uncertainty is concentrated.

From feature to measurement layer

The strongest use case for virtual try-on may therefore be broader than creating a more engaging product page.

It gives brands an additional layer of first-party behavioural data at a moment that has traditionally been difficult to observe: the moment when a customer tries to imagine the product on themselves.

That does not mean every virtual fitting room will improve every KPI. Research on the technology is clear that perceived usefulness, realism, trust and ease of use all affect consumer response. A poorly implemented tool can become another layer of friction rather than reducing it.

But when measured properly, virtual try-on can help brands see more than whether a customer clicked a button. It can show where interest turns into hesitation, which products require the most reassurance and what happens when shoppers are given more confidence before they buy.

For an industry that has spent years trying to recreate the experience of the fitting room online, that may be the more important shift. The future value of virtual try-on may not lie only in making the digital shopping experience look more personal. It may lie in finally making uncertainty measurable.