How Retailers are Using AI to Create Dynamic Return Policies
Walmart’s announcement that it will offer curbside returns and return pickup at home for Walmart+ members is causing other retailers to take stock of their own return policies.
Return policies that are frictionless, transparent, and consumer-friendly have been shown to lead to dramatic increases in profitability and brand loyalty, but bridging the gap between online and offline product returns has been a longstanding challenge for multi-location retailers.
Walmart’s new “No Concerns” campaign promotes easier return policies during the holiday shopping period. Members of the Walmart+ membership program will soon be able to schedule the pickup of eligible returns in the Walmart app, and they’ll be able to hand off returns to delivery drivers without needing a box or label. Another component of the program is curbside returns, which will allow shoppers to remain in their vehicles while returning Walmart purchases.
In addition to using technology to make online and in-store returns easier, retailers like Walmart are also relying on tools like artificial intelligence to decrease overall return rates.
Somewhere between 20% and 30% of e-commerce sales come back to retailers as product returns, compared to 9% of brick-and-mortar sales. Retailers have deployed all sorts of strategies, from tightening policies to charging fees for return shipping. According to a report by Appriss Retail and Incisiv, 73% of returns are due to factors that retailers can control. For example, retailers can track consumer habits like overbuying and returning unwanted items and address shoppers personally through messaging and dynamic policies.
Intentional losses, which have become another major issue for retailers this year, can also be isolated and prevented using AI to understand in real-time where certain behaviors are taking place.
“Accepting product returns comes with a cost and a risk. These costs include the revenue lost from the product itself, money spent on return packaging and processing, as well as the money lost when a product is damaged or becomes out-of-season during the return process,” says David Speights, Appriss Retail’s chief data scientist.
In an annual survey conducted with the National Retail Federation, Appriss found that 10.3% of returns were identified as fraudulent. Despite this, Speights says it’s rarely in the retailer’s best interest to enact more restrictive returns policies.
“Not accepting returns also comes at a cost. Restrictive return policies can scare away potentially loyal customers and reduce sales,” Speights says. “Retailers must find a way to detect and deter fraudulent consumers without impacting any others.”
At Appriss, Speights works with retailers like Home Depot and Dick’s Sporting Goods to create consumer-friendly return policies and combat internal fraud. He says technology can streamline returns, so the vast majority of consumers see fewer restrictions, while still protecting retail brands from losses due to fraud.
When the fast fashion retailer Zara announced a new policy to charge customers for returns, and even block returns altogether in some cases, the backlash was swift. Consumers questioned why the retailer would seemingly punish all customers, including the company’s most loyal shoppers, for fraudulent actions committed by a few.
Speights says AI could have been used to mitigate Zara’s losses without alienating the brand’s most loyal shoppers. He says many retailers are using AI to improve the customer experience and provide loyal shoppers with less-restrictive return policies by utilizing their full shopping and refund histories and separating the good shoppers from the bad.
“AI can recommend dynamic policies and create behavior-based returns recommendations that give high-value customers exciting perks like longer return periods or can provide targeted incentives that can recapture revenue and modify shopper behavior, such as if a retailer wants a consumer to make a return in-store rather than ship it back,” Speights says. “And this is occurring all while the system is identifying and recommending restrictive policies for shoppers whose returns disproportionally cost the retailer.”
However, implementing an AI-based approach to returns does require a shift in thinking for retail associates, from a policy-driven system to a consumer-tailored system. Speights says most associates are used to the “customer is always right” approach. While this response leads to positive customer service reviews, it may cost some retailers in the long-run.
“Traditionally, retailers have always looked at returns as just a cost of doing business. One of the biggest challenges in implementing an AI-driven returns approach is a lack of awareness around the capabilities in this space and in the actual risks that exist,” Speights says. “Beyond realizing how AI can mitigate returns abuse and fraud, many aren’t aware of the opportunities available to improve the consumer experience.”