Blog: Rethinking Loss Prevention in Retail

That was the focus of my recent podcast conversation with John Harmon, Managing Director of Technology Research at Coresight Research. Together, we dug into how shrink is evolving—and how AI, when applied thoughtfully, can help retailers respond in ways that are effective, sustainable, and human-centered.

“Shrink breaks down into three major categories,” Harmon explained. “About a third is customer theft, a third is employee theft, and the last third is operational or supply chain loss. That’s what makes it so difficult—there’s no one-size-fits-all solution.”

In my own work, I’ve seen this complexity firsthand. Technology can absolutely help, but only when it’s part of a broader strategy. Tools like computer vision and AI-driven video analytics can flag errors, identify patterns, and even detect theft in real time—but the real breakthrough comes when we think about how to act on those insights without harming the customer experience or placing staff at risk.

“The idea of ‘nudging’ is powerful,” Harmon said. “You’re not confronting a customer—you’re giving them a chance to self-correct. It’s safer for employees, and it still addresses the issue.”

Nudging, as Harmon described, is a subtle, non-confrontational prompt—like an on-screen reminder or a gentle audio cue at the self-checkout—that alerts a shopper to a potential scanning error or missed item. Instead of involving a store associate right away, it gives the customer an opportunity to recognize and correct the mistake themselves, maintaining the flow of the transaction while reducing friction and potential conflict.

That balance is key. AI can help identify loss, but we still need human judgment and operational nuance to address it properly. At Diebold Nixdorf, we’ve worked hard to build systems that respond to different types of shrink, from non-malicious scanning errors to high-value theft in aisles—always giving retailers the flexibility to decide how and when to intervene.

As John pointed out, the next few years won’t just be about smarter tech—they’ll be about comprehensive, connected solutions. “We’re close to unified systems that can track what’s happening at self-checkout, in the aisles, at the back door—everywhere. The data is there, we just have to bring it together.”

AI won’t eliminate shrink overnight. But with the right data, the right intent, and a more thoughtful approach, we can begin to design retail environments that are smarter, safer, and more resilient.

To learn more about how Diebold Nixdorf helps retailers, contact us here or to check out the full podcast episode here.