Neatsy AI, a U.S.-based company, is utilizing the depth-sensing capabilities of the iPhone’s FaceID selfie camera to function as a foot scanner, creating 3D models to accurately predict comfortable sneaker sizes.

The company’s application, presently in a soft launch phase for iOS with an official release scheduled for next month, begins by gathering some preliminary information from the user regarding their sneaker fit preferences. It then guides them through a series of steps to generate a 3D scan of their feet using the iPhone’s front-facing camera. This scan is then employed to provide personalized fit predictions for sneakers available for purchase within the app, displaying a unique fit score (on a scale of five) in green text alongside each sneaker model.

Purchasing footwear online often results in a significant number of returns when customers find the fit is not as expected, due to the lack of standardized sizing across brands. Neatsy aims to address this issue by leveraging artificial intelligence to incorporate a more precise, individualized fit assessment into the shopping experience.

Established in March 2019, the startup has secured $400,000 in pre-seed funding from angel investors to bring its iOS app to market. Currently, the app is accessible in the United States, United Kingdom, Germany, France, Italy, Spain, Netherlands, Canada and Russia. 

Neatsy employs a machine learning model to analyze users’ foot scans, predicting a comfortable fit for a variety of popular sneaker brands—including Puma, Nike, Jordan Air, and Adidas—based on scans of sneaker insoles, according to CEO and founder Artem Semyanov.

He explains that they are also considering the materials used in shoe construction and plan to continuously refine the algorithm based on user feedback regarding fit. (The startup has obtained a US patent for its 3D scanning technology used for shoe recommendations.)

The team evaluated the algorithm’s effectiveness through commercial trials this summer, demonstrating a 2.7x reduction in sneaker returns related to sizing issues and an overall 1.9x decrease in returns among a test group of 140 participants.

Returns represent a substantial expense for online retailers; Neatsy estimates that sneaker returns alone cost ecommerce businesses approximately $30 billion annually, factoring in logistical expenses and issues like damaged packaging and missing items.

“Overall, return rates for online shoe purchases typically range from 30% to 50%, with fit and size discrepancies being the most frequent reasons,” states Semyanov, who previously led the machine learning team at Prism Labs prior to founding Neatsy.

“Zappos reports that customers return roughly 50% of their most expensive footwear purchases. Furthermore, 70% of online shoppers make returns each year. Statista projects that return deliveries will cost businesses $550 billion by 2020,” he shared in response to inquiries via email.

“A 2019 UPS survey revealed that 73% of shoppers’ overall returns experience influences their likelihood of making future purchases from a retailer, and 68% say it impacts their general perception of the retailer. This highlights the significance of the issue!”

“Retailers are compelled to absorb the high costs of returns to maintain customer purchasing behavior. Our approach focuses on addressing the root causes of returns, rather than merely managing the symptoms.”

While major ecommerce platforms like Amazon concentrate on streamlining logistics to reduce delivery and return friction, numerous startups have attempted to resolve size and fit challenges using various digital and traditional methods—ranging from 3D body modeling to ‘smart’ sizing suits or even brand-specific measuring tapes (such as Nudea’s fit tape for bras)—however, a universally effective solution remains elusive, and many of these ventures have either ceased operations or been acquired without substantial success.

Although Neatsy is tackling a problem that many other founders have attempted to solve, it is specifically focused on the sneaker market—a relatively narrow segment that could facilitate the development of a valuable tool.

The company also benefits from the widespread availability of the iPhone’s sensing technology. (In contrast, Solely Original, a custom shoe design company, has offered personalized fit by charging a premium for sending out individual fit kits.)

However, even when focusing on sneaker comfort, Neatsy’s foot scanning process requires users to follow a considerable number of steps (as demonstrated in the accompanying video). Additionally, users need to have a pair of solid-colored socks available (patterned socks are not supported). Therefore, it is not an instantaneous process, although the scan itself only needs to be performed once.

At the time of this writing, we were unable to personally test Neatsy’s scanning process, as it requires an iPhone with a FaceID depth-sensing camera. On a 2nd-generation iPhone SE, the app allowed navigation through the scan instruction steps but then stalled at the point where the scan should have begun—displaying a green outline of a left foot against a black background.

The team indicated that this is a bug they are addressing, and the scanner will be disabled for iPhone models lacking the necessary hardware. (The App Store listing states compatibility with the iPhone SE (2nd generation), but does not specify that the foot scan feature is unavailable.) 

Currently, Neatsy’s app operates as a direct-to-consumer ecommerce platform, offering a selection of sneaker models to tech-savvy Gen Z and Millennial consumers. However, it is also intended to serve as a showcase for retailers interested in the technology.

When asked about this, Semyanov confirmed the company’s long-term goal is for its custom fit model to become a standard component of the ecommerce landscape.

“The Neatsy app is our quickest route to demonstrate our vision for the future of online shopping,” he told TechCrunch. “It attracts users to shops and we receive a revenue share when users purchase sneakers through our platform. The app functions as a new, low-return sales channel for retailers and allows them to directly observe the economic impact of reduced returns.”

“Looking ahead, we believe our future lies in B2B partnerships, with all ecommerce shops eventually integrating fitting technology—and we are confident that ours will be the solution. It will become as essential as credit card payment integration in online stores.”