In the United States, 2.6 million tons of returned garments landed in landfills in 2020, according to Earth.org. Returns cost retailers upwards of $800 billion last year, the National Retail Federation (NRF) found. The deluge of fast fashion hauls and the trend to bracket shop online indicate that return rates and waste will keep climbing.
This crisis has drawn the attention of fashion technology firm Unspun, which has been investigating how it can alleviate the pain points that are causing high such levels of waste. Though the B Corp’s 3D weaving technology, Vega, addresses associated challenges like on-demand production, nearshoring and overconsumption, more could be done.
Related Stories
“There’s only so much reach that we at Unspun can have,” Dan Robichaud-Carew, Unspun’s head of software product, told Sourcing Journal. “We have cool tech; they have a problem.”
On Thursday, Unspun launched FitOS, a data-driven platform that guarantees consumers a perfect fit and empowers brands to deliver on that guarantee through advanced body scanning and AI algorithms.
The suite builds on the lessons learned since Unspun’s inception nearly a decade ago. Described as a new benchmark in personalized sizing recommendations, the software uses Unspun’s proprietary technology to recommend the best-fitting sizes to shoppers whilst helping brands rework their inventory accordingly.
“We know clothing very intimately,” Robichaud-Carew said. “We’ve sold it ourselves, we’ve designed it ourselves, we’ve done the pattern making, we understand stretch factors—for every pair of pants we make someone, we do a heat map to see if this is going to be tight or loose in the right areas.”
Unspun is confident FitOS will reduce return rates by up to 50 percent.
Emphasizing a collaborative approach, FitOS utilizes the actual pattern data sourced directly from a brand, ensuring accurate size recommendations. The long-term goal is to improve or standardize fit throughout the industry at large.
Consumers can access FitOS by completing an online survey or performing a 30-second body scan using a smartphone to create unique 3D models with thousands of data points. Unspun’s proprietary algorithms “synthesize” the data, examining static measurements as well as body shape, proportions and posture. From there, the Unspun suggests which size would look and feel best.
“We think [size recommendation] can really have a positive impact on reducing returns and getting customers to feel more confident in what they’re buying,” Robichaud-Carew said. “They know it’s the right fit for them and so they’re more likely to hit that button to complete the purchase.”
FitOS insights go beyond consumer recommendations, too.
The platform empowers brands to leverage data-driven insights to optimize their design and production processes. Rather than relying on generic fit models—often based on figures that haven’t been updated since the 1960s—FitOS enables brands to create and refine fit models grounded in real customer data, ensuring a tailored, accurate fit that reflects the brand’s demographic.
In turn, the system suggests to brands when to carry fill-in size inventory and recommends pattern grading updates. It can also auto-generate patterns based on customer data.
Simply put, Unspun aggregates the data into tailored takeaways for brands to implement.
“We’ve got this library of anonymized data that we can take and say to a brand this is what your customer looks like; their body shape looks like this and your fit looks like this—if those two are not in alignment, we could recommend changes for that customer,” Robichaud-Carew said.
Inventory optimization provides size curves and inventory recommendations, allowing brands to adjust their fits and stock proactively. With a more informed buying strategy, brands can capture more sales by ensuring the correct sizes are available and match the right customer to the right size.
Additionally, FitOS integrates with Unspun’s custom patterning technology for customized, on-demand production. Vega integration makes low inventory and rapid prototyping possible. The platform works with existing PLM software options as well.
“The goal of this is to reduce returns in the short term by making a size recommendation,” Robichaud-Carew said. “Long term, you’re just seeing much better fit across your entire business.”
