Appriss Retail estimates that fashion and other retailers worldwide lost a combined $796bn in 2025 due to the twin pressures of returns and shrink, according to its 2026 Total Retail Loss Benchmark Report, which consolidates multiple forms of leakage into a single enterprise-wide view.
The findings draw on survey feedback from more than 1,000 consumers and a large-scale analysis of returns behaviour tied to 250 million unique customer identifiers. The report’s “total retail loss” framework integrates shrink, fraud, returns abuse and operational leakage to show how losses accumulate across functions and channels, rather than treating each category as an isolated problem.
Appriss attributes the majority of 2025 losses to merchandise returns, estimating returns-related losses at $706bn. It says $100bn of that figure was preventable and linked to fraud and abuse, equal to 14.2% of all returns. Within returns-related preventable loss, returns abuse accounted for 12%, while fraud represented 2%.
Shrink contributed another $90bn. The report says preventable shrink was driven mainly by employee theft ($26bn), inventory errors ($19bn), operational errors ($12bn), and organised retail crime ($9bn).
Appriss Retail CEO Michael Osborne said retailers need to address returns, fraud and shrink together instead of tackling each in separate silos. “Returns overwhelmingly power the majority of financial loss that retailers endure. Every dollar lost to returns is a dollar straight off the bottom line. To stop the bleeding, leaders must look at returns, fraud, and shrink through the lens of Total Retail Loss, build a system of collaboration, and implement cross-functional muscle. Retailers that continue to work in silos will continue to erode profits.”
The 2026 Total Retail Loss Benchmark Report also points to the cost of fragmented data across channels, highlighting cross-channel fraud linked to buy online, return in-store (BORIS) activity, which the report estimates generated a $4bn loss.
Returns volumes differed by channel. Buy in-store, return in-store was the largest component at $367bn, representing 52% of all returns. Buy online, return in-store accounted for $208bn (29%), while buy online, return online reached $131bn (19%).
Beyond the headline loss figures, the report examines how customers prefer to return goods, what influences loyalty and how shoppers perceive AI-assisted returns decisions. It argues that analytics and AI used precisely can identify abuse patterns and trigger warnings without automatically pushing customers away. According to Appriss’ data, 90% of consumers would purchase again after receiving a warning, helping protect $75bn in retained revenue.
However, consumer trust in automation remains cautious: 80% want transparency in how AI reaches decisions, 71% trust human associates more than AI for approvals, and only 10% trust AI outright. Appriss concludes that retailers can deploy AI most effectively when they centralise omnichannel data and communicate clearly so abusive behaviour is sanctioned while legitimate returns are processed smoothly, protecting loyal shoppers and improving profitability.