• D.C.
  • BXL
  • Lagos
  • Dubai
  • Beijing
  • SG
rotating globe
  • D.C.
  • BXL
  • Lagos
Semafor Logo
  • Dubai
  • Beijing
  • SG


Updated Dec 27, 2023, 2:09pm EST
net zero

AI is helping cut the carbon footprint of online shopping returns

MIKE SEGAR/Reuters
TweetEmailWhatsapp

Sign up for Semafor Net Zero: The nexus of politics, tech, and energy. Read it now.

Title icon

The News

Online retailers are beginning to grapple with a major source of waste and greenhouse gas emissions from their operations that has largely flown under the radar: Returned items.

The pandemic launched an online shopping boom, and a tidal wave of returned items followed. This year, the retail value of product returns in the U.S. alone will approach $1 trillion, according to Hitendra Chaturvedi, a professor of supply chain economics at Arizona State University. Yet costs and inefficiencies in “reverse logistics” mean that for most returned items — up to 9.5 billion pounds (4.3 billion kilograms) in the U.S. — it’s cheaper for the retailer to send the product to a landfill than to bother reselling it: Chaturvedi estimates that 80% of returned products end up in the trash. But a growing number of retailers are turning to artificial intelligence to help.

“We’ve just begun to scratch the surface of where AI can come in to reduce the waste of transporting and processing returned goods,” said Amena Ali, CEO of the reverse logistics management software company Optoro.

AD
Title icon

Tim’s view

While retailers are beginning to take modest steps to reduce emissions from online sales — streamlining packaging or using electric delivery trucks, for example — so-called “reverse logistics” has been largely ignored. That’s ultimately a climate problem, because the emissions associated with transporting and wasting returned goods add up to about the same footprint as 5 million cars driven for a year, according to Optoro.

“The supply chains we have are optimized for the forward movement of goods,” Ali said. “The big challenge with returns is that every return is a different story.”

That’s where AI can help. About a third of U.S. online shoppers engage in “bracketing” shopping, Ali said, in which they buy a range of items with the intention of testing them out and returning some. AI can help cut back on that practice by processing customer feedback on the reasons for returns and generating guidance for the next customers — this shirt style runs big in the shoulders, for example. For items that do get returned, some retailers are using AI-driven software that aggregates returns headed to the same warehouse, cutting back on transportation space. And once items return to a warehouse, AI image processing can help workers identify and sort them, to increase the odds that they can be put back on the shelf or sent on to be resold by a liquidator.

AD
Title icon

Room for Disagreement

Unfortunately, the most effective means of reducing the environmental impact of returns — limiting customers’ return options so they shop more carefully in the first place — runs counter to a key imperatives of online retail, which is to make it as easy as possible. Any barriers to returns will just send customers elsewhere.

“At this moment, there is no significant swell in the e-commerce space to reduce returns,” Chaturvedi said. “If you create a bad return policy, 80% to 90% of your consumers are not going to come back to you.”

Title icon

The View From London

In its most recent sustainability report, Amazon claimed that a service through which customers can troubleshoot problematic items prevented 7.5 million returns, and that for products that did get returned in 2022, “most were eligible to be resold as new.” For those that didn’t qualify, 30% could be relisted as used on Amazon Warehouse. And in November, the company opened its second brick-and-mortar “second chance store” in London to resell returned products.

Semafor Logo
AD