Talk about being stuck between the Scylla and Charybdis. Retailers love it when shoppers buy more; it’s a central goal of many of their marketing efforts! But consistent data also show that the shoppers who buy the most are also the ones who return the most items, and the costs of returns represent one of the most central negative measures that retailers seek to avoid in their marketing efforts. So what’s a seller to do, stop selling more, or accept more returns? With effective AI, they might have another option that will allow them to set sail into a profitable future.

By carefully tracking what happens when people click on search ads, including what they buy, how many, and what they return, machine-learning–based AI can establish which customers tend to be most likely to return products they buy. Then those algorithms combine those data with insights about the return rates for specific types of products (which different retailers can define however they want). With these combined data, the AI system offers real-time predictions about returns and the profits on each sale.

Once it has enough of these combined data points, it integrates all of them together, which then provides meaningful insights into where to target search ads, namely, to whom and for which products, in a way that encourages sales without increasing the risk of return behaviors. One Dutch fashion retailer that adopted such a system ultimately rethought its overall strategy: Rather than trying to increase sales, it would aim to increase customer lifetime value. Thus, it stopped targeting its advertising to consumers that it knew would buy, but that it also predicted were likely to return. As a result, its returns have decreased, even if just by 5 percent, but its profits have risen by 16 percent.

This direct application thus seems effective, but other retailers appreciate more indirect uses of AI to ensure they can keep appealing to all their customers. Consider Perry Ellis for example: “After identifying items with the highest return rates, they used AI sentiment-analysis tools to determine which phrases in those products’ descriptions might create confusion over key elements, such as size or fit, that most often lead to returns.” If consumers were returning clothing because they were confused about its specific attributes when they placed the order, then clarifying the information could eliminate this risk, at little cost. For example, buyers assumed that the simple phrase “Machine Washable,” that appeared on digital product pages for some of shirts meant they could throw the tops in with their regular laundry. But the fine cotton or linen actually requires more gentle care, so the product pages now specify, “Machine wash according to instructions on care label.” Simply by alerting consumers to the need to take more care in washing the items, the brand lowers the risk that people demand returns after just one wash, when their delicate shirt has unraveled after being washed with jeans and towels.

Such efforts are in line with the relatively longer-standing applications of AI by fast fashion brands that rely on machine learning to provide more accurate recommendations to online shoppers. If they buy the right size or know precisely which accessory goes with an item already in their cart, the reasoning goes, they will have less reason to initiate a return. But nearly every retailer continues to look for effective, efficient ways to build on and extend such insights to reduce returns. Today’s consumers believe that easy, free returns are their right. As a result, following the most recent holiday season, they returned an estimated 16.5 percent of the items that were purchased, accounting for approximately $817 billion in returns. That’s a level that demands dedicated AI attention.

Discussion Questions

1. How often do you return items to retailers, and why? Could better information reduce the amount of returns you initiate?

Sources: Patrick Coffee, “Retailers Enlist AI in Fight Against Returns,” The Wall Street Journal, December 18, 2023; Liz Young, “Retailers Are Bracing for Their Postholiday Returns Hangover,” The Wall Street Journal, December 26, 2023