Retail returns can be a profit lever

Retailers are now treating returns less like a customer-service afterthought and more like an inventory problem with a clock running on it. In 2024, the National Retail Federation and Happy Returns said U.S. retailers expected $890 billion in merchandise returns, equal to 16.9% of annual sales. (nrf.com) That clock matters because processing a return is expensive even before an item goes back on sale. Happy Returns said the average cost of processing a return now exceeds 21% of the order’s value, which means a $100 order can lose more than $21 before markdowns, shipping, or fraud are counted. (happyreturns.com) McKinsey’s argument is that artificial intelligence can change that math by making faster decisions at each step of the return. The firm’s reverse-logistics report says U.S. consumers returned nearly $1 trillion of merchandise in 2024 and retailers spend about $200 billion a year processing returns and recovering value from them. (fibre2fashion.com) The first fix is deciding earlier whether a return should happen at all. Deloitte says artificial intelligence tools can guide shoppers to exchanges, store credit, or the nearest drop-off option before a box starts moving, which cuts transport cost and keeps some sales from turning into refunds. (deloitte.com) The second fix is routing each returned item to the place that can recover the most value from it. Blue Yonder says newer systems can send a returned product to a store or region where demand is still strong instead of defaulting to a central warehouse, giving that item a better shot at selling again at full price. (blueyonder.com) The third fix is inspection, which is where warehouses often lose hours and margin. Deloitte says artificial intelligence can classify condition, spot likely fraud, and recommend whether an item should be restocked, repaired, liquidated, or recycled instead of waiting for a manual decision at a returns table. (deloitte.com) The fourth fix is speed back into inventory. Tecsys says real-time returns processing lets a sellable item go straight back into stock, and that faster reshelving reduces the need for markdowns because the product gets back in front of buyers while demand is still there. (tecsys.com) The fifth fix is labor planning, because returns arrive in waves after holidays and promotions. Deloitte says artificial intelligence can forecast return volumes and help retailers schedule warehouse labor and capacity ahead of those spikes instead of paying for bottlenecks after pallets have already piled up. (deloitte.com) The sixth fix is customer communication, which sounds softer than warehouse software but hits the same profit line. The National Retail Federation said 76% of shoppers consider free returns when deciding where to shop, so retailers are trying to keep the return experience easy while using better data to steer customers toward lower-cost options such as consolidated drop-off points. (nrf.com) There is a second number sitting behind all of this: fraud. Appriss Retail and Deloitte said fraudulent returns and claims cost retailers $103 billion in 2024, with 15.14% of returns deemed fraudulent, so any system that gets better at spotting suspicious patterns can protect margin before a refund is issued. (businesswire.com) The retailers that win here are not the ones with the nicest return label email. They are the ones that can tell, within minutes, whether a black size-8 dress should go back to a store shelf in Dallas, an outlet in Phoenix, a repair bench, or a fraud review queue. (blueyonder.com)