Online fashion returns seek sustainable solution

A new AR size fitting startup examines the problems that returns pose for online shoppers and what innovative solutions exist to help

Online fashion is booming, but one issue prevents it from reaching its full potential: what size should customers buy? 

With inconsistent sizing between brands, confusing fit guides, and no way to “try before you buy,” the industry is losing billions in returns, frustrated customers, and missed sales.

SizeSutra, a UK-based fashion tech startup, is working to address these challenges with a frictionless, artificial intelligence (AI) powered solution that one of its co-founders, Priyanka Srivastava, said works like a global positioning system (GPS) for clothes.

"Just like a GPS pinpoints a location to give accurate directions, SizeSutra is being designed to read a customer's body data to guide them to the right size — fast, securely, and with zero guesswork," Srivastava told RetailTechnology.co.uk.

A £7 billion problem no one has solved

In the UK alone, returns as a result of poor fit and sizing inconsistency account for over £7 billion in lost revenue and more than 300,000 tonnes of waste annually. From return shipping to overproduction, sizing is a key driver of environmental damage and financial loss. But most “solutions” — like size charts or questionnaires — still expect the customer to figure it out for themselves.

"What makes SizeSutra different is its radically simple user experience," explained Srivastava. "There’s no need for manual input, body measurements, or trying on multiple outfits."

Here’s how SizeSutra works:

1. The customer opens their camera.
2. They stand in front of it for a few seconds.
3. Our AI detects body landmarks and makes an instant recommendation.

"That’s it," she added. "No typing. No measuring. No confusion."

The startup founders said it offers a one-click experience that integrates seamlessly into any brand’s existing online store, from Shopify to custom e-commerce platforms, without disrupting the journey. It is also designed to work across all PC and mobile devices and be intuitive and easy-to-use by all demographics.

Brand-specific fit intelligence

Most size recommendation tools use generic body standards or crowd-sourced sizing data. But SizeSutra's founders contend that is not good enough.

"SizeSutra maps the customer’s body directly to the retailer’s actual sizing chart, meaning recommendations are made per brand, per garment, not based on some industry average," Srivastava said.

For example, if a brand’s size medium is cut smaller or larger than others, SizeSutra will know this and adapt accordingly. This level of brand-specific accuracy boosts customer trust and reduces return rates dramatically.

Another feature SizeSutra thinks will be game-changing is fit flexibility. "We don’t just tell customers, 'you’re a size medium,'" she continued. Instead, SizeSutra offers three intelligent fit options:

• Right Fit – The standard, most accurate match based on the brand’s cut
• Tight Fit – A snug look for those who prefer body-hugging styles
• Loose Fit – A relaxed recommendation for casual or layered outfits

This allows customers to choose the fit they want, not just what the brand assumes. It’s true personalisation, powered by computer vision.

Virtual fitting rooms reimagined

At its core, SizeSutra's founders stress that its innovation is about restoring choice, confidence, and convenience to the online shopping experience.

"Much like a GPS gets you from A to B without needing to know the entire map, SizeSutra gets shoppers to the perfect fit — without needing them to understand sizing charts or garment measurements," added Srivastava.

"Retailers benefit too, from fewer returns, which lower logistics costs and emissions; higher conversion rates as a result of more confident buyers; and, increased customer loyalty by offering personalised experiences from the first click."

Built using advanced computer vision models such as OpenPose and MediaPipe, SizeSutra’s system can identify key body landmarks in seconds. All data is processed securely and ethically, with privacy by design. No sensitive data is stored without user consent, and the tool is compliant with the Euopean Commisson's General Data Protection Regulation (GDPR).

The backend then matches the user’s proportions with brand-specific garment data, delivering recommendations in real time.

Retail and AI expertise

Co-founders Priyanka Srivastava (CEO) and Abhishek Srivastava (CTO) hope their new tool combines fashion industry insight with cutting-edge AI capability.

The startup has just shortlisted as one of AI Forge Accelerator final 25 and NVIDIA's Inception Programme entrants. It also won the Corporate LiveWire Innovation New Retail Management Software of the Year Award, 2024/25.

Srivastava concluded: "If GPS transformed how we travel, SizeSutra is here to transform how we dress.

"The fashion industry has spent too long relying on outdated sizing systems and expecting customers to make them work. It’s time for technology to meet people where they are — literally, which starts with SizeSutra."