Artificial intelligence’s role in retail continues to evolve as merchants seek to make better merchandising and inventory decisions, deliver better product recommendations, launch chatbots and even identify fashion trends. As the fallout from the holiday season continues amid record return rates, some of the biggest retailers are now leveraging the technology to decide whether handling returns is worth the hassle.
Amazon and Walmart are among retailers deciding to simply refund shoppers for their online purchase without putting them through the rigmarole of returning a product, in a move that ultimately alleviates stress on both the companies’ internal logistics networks and major third-party partners. A recent National Retail Federation (NRF) survey noted that for every $1 billion in sales, the average retailer incurs $106 million in merchandise returns.
Retailers have been experimenting with different ways to minimize returns’ bottom-line impact, such as by having more shoppers return online purchases in stores or even lockers without having to package or mail the item. Third-party service providers such as Happy Returns, Optoro, Newmine, Returnly and Narvar are among the services making this happen. But even with more of these platforms offering assistance, the rapid growth in returns is pushing retailers to take more drastic measures.
While the idea of letting shoppers keep the products seems extreme, there is logic behind the decision for retail giants that would otherwise eat the costs. For many inexpensive items, or larger ones that include hefty shipping fees, it is often cheaper for the retailer to refund the product without taking it back.
Here’s where the artificial intelligence aspect kicks in: a Walmart spokeswoman told The Wall Street Journal that it has a specific plan in place when it deploys this policy, noting that the “keep it” option is designed for merchandise it doesn’t plan to resell. This merchandise is determined by numerous factors that the technology examines, including customers’ purchase history, the price of the products and the cost of processing the returns.
The practice gained attention in 2017 when Amazon introduced “Returnless Refunds” as an option for sellers. At the time, Amazon said the feature was “highly requested” by sellers that sought to avoid the time and cost of managing returns shipping and processing or for items that would be hard to resell.
For all the buzz AI has gotten in recent years, particularly in trying to prevent returns in the first place, this does seem like a useful and realistic application of the technology. While Newmine estimates that 65 percent of returns can still be controlled by the retailer due to necessary improvements in product descriptions or visuals, that still leaves 35 percent that is out of their hands.
If a retailer is putting all the effort into getting the right product to the consumer only for them to find some other reason to send it back, then AI’s value in the return process only becomes more valuable in determining whether the cost is worth it. Given that AI has such a hand in dynamically pricing products and recommending the optimal product, it makes sense that more retailers would use it to calculate the “worthiness” of a return.
To prevent shoppers from discarding these items entirely, which could create a whole new sustainability problem, retailers would instead have to be creative about how they incentivize shoppers to keep the product.
A Target spokeswoman told the Wall Street Journal that in a “small number of cases” the retailer gives customers refunds and encourages them to donate or keep the item. Similarly, online pet products retailer Chewy told a consumer to donate a cat harness to an animal shelter instead of returning the product. The company refunded the shopper and sent her a new harness in a bigger size.
The anticipated returns hype lived up to its billing, with consumers sending back an estimated $428 billion in merchandise to retailers last year, approximately 10.6 percent of total U.S. retail sales in 2020, according to NRF. But this number jumped even more dramatically for online returns—while e-commerce accounted for $565 billion in 2020, $102 billion of merchandise purchased online was returned, or approximately 18 percent of online sales.
Overall, the online returns process adds up, as processing e-commerce returns can cost $10 to $20 per item before freight fees enter the picture, Locus Robotics CEO Rick Faulk told the WSJ.
The number of e-commerce packages that were returned in 2020 jumped 70 percent from 2019, according to reverse logistics platform Narvar. More than half of the increase comes from the e-commerce sales spike, Narvar estimated, while more than a quarter was the result of shoppers’ not wanting to return web orders to physical stores.