Between discarded packaging, shipping fees, inventory shortages and damaged merchandise, returns cost retailers a fortune. In 2017, retail returns cost roughly $350 billion, and they were projected to rise to $550 million in 2020. While some returns are inevitable (with reasons ranging from defective products to late deliveries), others are entirely avoidable. In fact, 46% of shoppers surveyed in Narvar’s 2019 The State of Online Returns report said their No. 1 reason for returning products was incorrect size, fit or color. Only 3% intentionally bought multiple items knowing they’d return some. Based on this information, these two reasons account for almost half of online returns.
In our modern “convenience culture,” brands are complicit in this economic (and environmental) waste. Some folks have even gone as far as suggesting death to the free return policy. Given the state of online commerce, it would be hard to imagine a world in which retailers decided to roll back their return policies. Unless done universally, many sellers might find themselves at a huge disadvantage to shops that continued to accommodate consumers with two-day shipping.
Hobby shopping and finicky taste aside, many shoppers genuinely need friendly return policies because it’s still hard to commit to a purchase sight-unseen. That’s because customers might not know which products and sizes will best suit them. But, increasingly, e-tailers are turning to artificial intelligence (AI) and machine learning (ML) to help customers not only discover new products that may pique their interest, but also identify SKUs and specific sizes they’ll happily keep in their wardrobe.
An Idea for the Evolution of Recommendation Engines
Amazon is perhaps the most famous example of big data driving e-commerce. As the world’s largest online store tracking purchasing behavior across its shopping site and tens of thousands of other merchants with its payments platform, Amazon knows a lot about affinity and propensity. The digital mega-mall applies every piece of data it knows about you and other shoppers like you to power its recommendation engine, helping you find and purchase more products than you originally planned. It has even commercialized its recommendations engine into a B2B service.
However, the focus of nearly all personalization tools has been to increase average cart sizes and customer lifetime values. This tunnel vision prioritizes revenue growth but neglects sales preservation.
Using the same technology but going a step further to analyze return-based data points, merchants can place more weighted value on product upsells and cross-sells they know shoppers are more likely to purchase and keep. In turn, brands will likely see more delighted customers and meaningful improvements to their bottom line. While this might not be a novel idea, it’s one worth resurfacing because the retail return problem is only projected to grow over time.
Increasing Interest in Fit Technology
In categories like fashion, not-so-standard sizes add another element of complexity to the shopping experience. An investigation by Racked found discrepancies in sizing from various retailers, revealing that a pair of size 8 jeans from Uniqlo ran more than 5 inches larger at the waist compared to a size 8 jeans from Gap. This illustrates that brands do not agree on standard sizes. Shoppers will instead seek out their ideal fit based on the way clothing is cut.
Beyond that, there’s a matter of consumer preference, which machine learning expert and Bold Metrics COO Morgan Linton suggested to me is best measured on a per-brand basis. (Full disclosure: Author’s company is a client of Bold Metrics.) In fact, major retailers like Adidas, Canada Goose and Levi’s enlist Bold Metrics’ platform. With virtual sizing and try-on technology, customers can ditch the error-prone process that follows after pulling out the measuring tape. I’ve seen firsthand how we were able to reduce our fit-related returns by as much as 73%.
There are other AI/ML players in the fit technology space, such as Fit Analytics, which maps its large dataset to the brand’s sizing information. Using the Google Cloud Platform, Fit Analytics enlists the help of big data to make “250 million size recommendations every month for over 150 of the world’s top apparel and footwear companies.”
Another vendor, True Fit, tries to go beyond fit to power personalized product recommendations based on a number of factors, including fit. In partnership with Google Cloud, True Fit offers retailers the “Fashion Genome,” which is a large connected dataset for the fashion retail industry. This shift toward data-driven personalization speaks to how fit technology is becoming a positive for both retailers and consumers.
Generous return policies are so prevalent because they reduce friction in the purchasing process. But fashion retailers should consider incorporating AI and ML tools to help increase customer satisfaction when it comes to finding a perfect fit because even the process of repackaging a return order is an undesirable outcome for your shoppers and your bottom line.