From dresses that don’t fit right to shoes that just aren’t comfortable, sizing issues have plagued the fashion industry for decades. With e-commerce return rates hitting 16.9% in 2024, per National Retail Federation and Happy Returns, this problem is growing as online shopping booms. Could AI be the solution we’ve been waiting for?
Can AI Resolve Sizing Issues In Fashion Retail?
The fashion industry faces a significant challenge with a high volume of returns, particularly in e-commerce. A study by McKinsey estimates that each return costs between $21 and $46 as many customers order multiple sizes and return those that don’t fit. This impacts profit margins and contributes to increased costs in logistics, restocking, and handling.
In an era where the consumer wants everything at their fingertips, retailers have one shot to get it right or risk losing to a competitor. This is where AI comes in. ‘One of the struggles for consumers is that all of the brands are doing what they want to do to fit their niche in the market… part of what we do is serve as a translation engine between all of the different sizes brands are putting out there and just making a 1-to-1 recommendation for the consumer’, Christopher Moore, Chief Analytics Officer, True Fit tells Wall Street Journal.
Retailers are addressing this issue with AI-powered sizing apps that use data-driven algorithms to create personalised sizing recommendations based on the body measurements and past purchases. E-retailer Zalando is investing in sizing tech solutions to recommend the best size based on customer measurements based on two photographs. Retailers can download the True Fit application to integrate the machine learning dataset, Fashion Genome. The platform reimagines the conventional size guide by leveraging AI-powered fit recommendations, a service built on the collective insights of millions of shoppers and AI enhancements.
Choosing the right size is a major hurdle in fashion e-commerce and AI-powered sizing solutions have the potential to fix fashion’s inconsistent sizing, and reduce return rates. Let’s take a look at the current state of AI in sizing, examine its success, limitations, and the future of personalised shopping.
Why Sizing Is a Long-Standing Problem
Sizing is one of the most persistent issues in the fashion industry. Despite decades of advancements in garment production and a growing demand for inclusivity, sizing remains inconsistent across fashion brands, causing confusion and frustration for consumers. A ‘Medium’ in one brand may fit like a ‘Small’ in another. This lack of standardisation means that a person may wear different sizes depending on the brand, style, and the country of manufacture. What’s meant to be a simple process of finding the right fit has become an exercise in trial and error, making the retail experience time-consuming.
The fashion industry’s historical emphasis on ideal body types has also contributed to the sizing problem. Designers traditionally cater to limited body shapes, which has resulted in a system that doesn’t account for the diverse spectrum of body types and sizes. While brands are now making strides toward offering plus sizes, the inclusivity of these sizes is inconsistent. Physical stores let customers try and compare sizes. Online, shoppers rely on photographs, generic charts, and descriptors, which do not convey how a garment will fit on their specific body. While there are promising movements toward more inclusive sizing and innovative sizing solutions, such as body scanning technology and personalised sizing algorithms, the fashion industry must make inclusivity the norm rather than the exception.
How AI and Data Technologies Aim to Solve Sizing
AI size recommendation apps are designed to recommend the best-fitting size for a specific product, based on individual user data and predictive algorithms. These apps can be integrated directly into a storefront to guide customers during the product selection process. Sustainable activewear label Girlfriend Collective collaborated with AI sizing company Bold Metrics to deploy ‘Find My Size’. Walmart acquired startup Zeekit to power an AI virtual fitting room. The technology uses real-time image processing, computer vision and deep learning to simulate how clothes would look on the user’s body. Let’s take a look at the widely used AI-powered sizing solutions in fashion e-commerce:
Amazon is using body scanning and other technologies
Body Scanning
Body scanning technology is revolutionising how the fashion industry addresses the sizing issue by providing precise, 3D measurements of an individual’s body. For example, 3DLOOK’s mobile app enables users to take a photograph and receive tailored sizing recommendations based on their unique body shape. Brands like Uniform American Worker have successfully integrated 3DLook into their customer experience, improving fit satisfaction and reducing returns.
Body scanning also benefits the fashion production process. Designers can access accurate, diverse body data to create more inclusive sizing and better-fitting garments. For example, companies like Unspun use body scans to produce custom-fit denim, reducing mass production and minimizing waste. As body scanning becomes more integrated into fashion, it could lead to more sustainable, inclusive, and efficient clothing production that cater to a wider range of body types.
Amazon is using body‑scanning and AI‑powered sizing tools to reduce online returns by improving fit accuracy. One of its initiatives involves gathering 3D body‑shape data via scans or images to build statistical models of customer body types, which helps match items more precisely to a shopper’s dimensions — a move aimed at reducing cases where items don’t fit and are returned. More recently, Amazon’s Fit Insights tool uses brand size charts, review feedback and return history to give size recommendations and flag items with poor return health — helping shoppers pick better sizes and lowering the chance of returns.
AI Fit Recommendation Engines
AI-powered fit recommendation engines leverage advanced algorithms to predict the best size for customers with high precision, using data from past purchases, returns, and detailed customer profiles. These systems analyze factors like material stretch, garment cut, and individual preferences to refine recommendations. Fit Analytics, acquired by Snap in 2021, is a leader in this space. It partners with global retailers like ASOS, North Face, and Calvin Klein to tailor size recommendations. By analysing customers’ previous purchases and each brand’s unique sizing, the platform helps reduce returns by providing accurate fit predictions and improving the shopping experience.
Digital Twins And Virtual Try-Ons
Digital twins, which are virtual avatars of customers based on their body scans, try on clothes and footwear virtually, enabling shoppers to see how items might fit and look before making a purchase. This technology is gaining traction in footwear, where fit is often more complicated due to foot shape and gait. Nike has developed a solution called Nike Fit within their app, which uses AI-powered body scans to measure the user’s feet and recommend the most accurate shoe size. By using this technology, Nike provides customers with better-fitting shoes, reducing the number of size-related returns. The virtual try-on experience also makes online shopping more reliable and enjoyable, as customers can visualize how shoes will fit before making a purchase.
Automated Sizing Technology
As far back as the early 2000s, RFID showed the potential to transform retail by giving businesses the ability to quickly and accurately track all items, whether it was in the store or warehouse. But few retailers integrated this technology, citing the high cost and implementation issues. However, Zara implemented item‑level RFID tagging, enabling real‑time tracking from manufacture to checkout and returns. Combined with machine‑learning analysis of customer size, fit and style behaviour [via online and in‑store interactions], Zara can forecast which sizes are likely to sell in each market and ensure those sizes are available. This reduces the probability of customers selecting the wrong size. Furthermore, by capturing size‑related return data and feeding it back into stocking, design and distribution decisions, Zara progressively reduces fit‑mismatch returns and optimises inventory turnover.
AI-Powered Foot Scanners
AI-powered foot scanners are advanced technologies that capture precise measurements of a person’s foot, including its shape, pressure points, and gait. A combination of 3D scanning, pressure-sensing plates and AI algorithms can analyse a foot’s unique characteristics and match them to an ideal shoe model. The data collected helps in selecting the right shoe size, enabling brands to offer personalized recommendations. This technology is aimed at providing a more accurate, comfortable, and customised fit for each customer.
Adidas, for example, has partnered with companies like Volumental to incorporate AI-powered foot scanning technology into their flagship stores. The brand offers tailored shoe recommendations based on real-time measurements of foot shape and pressure distribution. Nike uses an AI-powered augmented reality foot scanner in its mobile app. The Nike Fit feature lets users scan their feet with a smartphone camera; the app creates a 3D model of the foot and recommends the ideal shoe size. By providing exact size suggestions, Nike’s AI tool helps customers get a better fit in footwear, minimizing return rates.
How Will Retailers Benefit with AI Sizing Technologies?
Can AI Resolve Sizing Issues In Fashion Retail?
- AI recommends a single, most-likely-to-fit size, discouraging multi-size orders and reducing returns.
- AI analyzes offers personalized recommendations based on fit, cut, and material rather than generic sizes.
- AI adjusts future recommendations based on historical return data, helping avoid sizing issues.
- AI tools convert global size variations, reducing the risk of incorrect size choices for international buyers.
Challenges That Are Still Holding AI Sizing Back
AI sizing solutions face several hurdles such as the inconsistency of size standards across different brands and regions. Despite its ability to analyze historical customer data and suggest the most likely size, there is a lack of universal sizing standards. This means that AI algorithms must account for brand-specific fits, which can otherwise confuse customers and lead to inaccuracies in the size recommendations. Additionally, the cultural and regional differences in body types complicate the predictive power of AI, particularly when dealing with international customers. Let’s take a look at some of the challenges —:
- Quality and volume of data used to train AI: AI algorithms rely on customer feedback, return rates, and previous purchases to make accurate predictions. However, if a fashion brand, particularly smaller or newer brands, does not have enough data, the AI struggles to generate accurate results.
- Data Privacy: With the introduction of body scanning technologies, concerns around biometric data privacy have emerged. Shoppers may be hesitant to share detailed body scans with companies, fearing misuse or data breaches.
- High Implementation Costs: While large corporations can afford the high costs of AI-powered sizing technologies, smaller businesses may struggle to implement them.
- Consumer Trust: Many customers remain sceptical of AI-driven recommendations. They may trust their own knowledge of their bodies over an algorithm, which could affect adoption.
- Cultural and Body Diversity: No dataset will ever account for the diversity in human bodies, so even the most advanced AI systems may fail to provide a perfect fit for everyone.
The Future: Will AI Standardise Fashion Sizing?
The future of fashion sizing is shifting from a one-size-fits-all approach to a personalised, made-to-measure model. Rather than creating a universal sizing standard, AI is set to revolutionise how clothing and footwear fit by tailoring them to individual body types. As AI technology advances, we can expect widespread availability of custom-fit garments and footwear at scale, powered by AI-driven systems that recommend the perfect fit and also create garments designed specifically for each customer. This approach promises to enhance satisfaction and sustainability by reducing returns and waste. The real question isn’t whether AI will redefine sizing, but how soon it will become a mainstream solution accessible to every shopper, regardless of their location or preferred brand.
Jasmeen Dugal is Associate Editor at FashionABC, contributing her insights on fashion, technology, and sustainability. She brings with herself more than two decades of editorial experience, working for national newspapers and luxury magazines in India.
Jasmeen Dugal has worked with exchange4media as a senior writer contributing articles on the country’s advertising and marketing movements, and then with Condenast India as Net Editor where she helmed Vogue India’s official website in terms of design, layout and daily content. Besides this, she is also an entrepreneur running her own luxury portal, Explosivefashion, which highlights the latest in luxury fashion and hospitality.
