What happened to virtual fitting rooms?

If you’ve been in fashion e-commerce for less than, say, six years, you might just be wondering what sort of question this is. You mean they’re not a new thing? Well, no – not exactly, anyway.

First of all, I’m going to keep this to fitting rooms – solutions associated with fit. There are plentiful solutions that help shoppers to create outfits, and even augmented reality-type solutions that aim to help the shopper visualise what they’ll look like in a specific garment or outfit. For the purposes of this look in the rear-view mirror, we’ll concern ourselves with fit.

In fact, according to Wikipedia, virtual fitting rooms trace their origins back to 2005, when Levi’s developed its Intellifit system, based on a bodyscanning technology. Whether this counts as a virtual fitting room is debatable: shoppers needed to physically be in one of the stores that had one the body scanners to use it. I certainly have never seen one, so it’s safe to say that it didn’t take off like Levi’s might have hoped.

By the early 2010s, an Estonian start-up named Fits.me had come up with an ingenious idea: create shape-shifting, robotic mannequins and photograph them, hundreds of times, each time in a different shape representing a different combination of measurements. Repeat for all available sizes of garment. In ecommerce stores, when a shopper entered their measurements they would be shown the appropriate photographs, so the shopper could see exactly how the garment would fit them. Ingenious and highly visual – but wholly unscalable, relying on the physical shipment of garments to a studio in large quantities.

A simpler approach – Fit Advisor – also from Fits.me, was based on garment measurements. This solution, too, had its drawbacks: retailers and brands multi-source their products, and garment measurements simply weren’t available in any consistent format; extracting and verifying them was time-consuming if not impossible. This remains a problem for virtual fitting rooms dependent on this approach… but more on this particular issue in a future post.

Like some other virtual fitting room solutions, Fit Advisor also relied on shoppers knowing their measurements – and this was another snag. Shoppers typically don’t know their measurements. When they don’t know, they tend to guess. And when they guess, they tend to get it wrong. The inevitable upshot of this is inaccurate recommendations; solutions that rely on shoppers knowing their measurements remain plagued by the same problem today. Again, more on this in a future post.

Some solutions attempted to get round this by asking shoppers about similar items they already owned that fitted them well, and what brand that was. This information was to be used to create knowledgebases that would enable them to know that a shirt in size A from brand B would fit similarly to shirt C from brand D (where A was the item under consideration by the shopper now, and C was the one they already owned and liked). This technique is still used now – but, of course, it has the significant drawback that it reminds the shopper of a garment they already like from another brand and… hey, why don’t I just go back to their store and buy a shirt from them, as I know that’ll fit nicely..?

Just about all virtual fitting rooms suffered from one other near-fatal flaw: a lack of evidence of value. In ecommerce, certain metrics are standard – conversion rates, return rates, average basket value – and virtual fitting rooms promised to increase, decrease and grow these metrics respectively. But it proved very difficult to attribute consumer behaviour accurately: not all garments had the option of the fitting room, what if a customer used the fitting room but ignored the result, is it still a conversion if they send the item back, and what was the impact on returns really..?

The lack of data, slightly clunky shopper experiences, the nightmare levels of IT and operational resources needed to deploy and maintain a virtual fitting room, the lack of customisation possible and – perhaps worst of all – a nagging sense of whether they worked and whether they were worth the effort… all of these factors combined to rather take the wind out of the sails of the fledgling virtual fitting room sector by the mid-2010s, although some providers still soldier on with differing degrees of success. Nonetheless, many of the ‘original’ virtual fitting room companies are either gone, or have pivoted to solve a different challenge. No virtual fitting room dominates the market – and no-one talks about them. The most successful fitting room solution in 2023 is still the humble, much-despised size chart!

Yet the challenge of high levels of returns still plagues retailers. Worse, a new factor has emerged in recent years: shoppers increasingly care about sustainability – and returns, especially for the largely unnecessary reason of “wrong fit”, contribute negatively to sustainability. Shoppers know it.

Here at Swan, we hope to change how retailers look at virtual fitting rooms. A decade after they first properly emerged, technology has caught up with the initial concept. Explore our site now to find out how (spoiler alert: it’s Artificial Intelligence), and we’ll answer more common questions about virtual fitting rooms in future posts on this blog.