Omnichannel

Reducing shrink without dismantling customer experience and trust

Shrink isn’t going away. Economic pressure, and everyday opportunistic theft, will continue to challenge retailers. But heavy-handed enforcement or experiences that make honest customers feel policed will just push them further from brick and mortar retail.

Generated by AI/Adobe Stock

May 29, 2026 by Daniel Gabay — CEO, Trigo

Summarize with

Retail loss prevention technologies for many years have treated retail theft as largely a self checkout problem. Cameras watch scanners and associates intervene (if necessary) at the point of payment. And yet, despite significant resources invested in these measures, shrink continues to rise to the tune of over $100 billion dollars.

That's because roughly 80% of retail theft, targeting highly stolen goods, that comprises so much of shoplifting is concealed at the shelf, hidden under clothing, slipped into bags, or nested inside other products long before a customer ever reaches checkout counters. By the time a transaction begins, most of the loss has already happened and traditional systems are simply looking in the wrong place.

The challenge for retailers today isn't whether to use technology like AI to address shrink, but how to do so in a way that meaningfully reduces loss without creating friction, mistrust, or a hostile in-store experience for honest customers.

Loss prevention needs a bigger scope

Most modern loss prevention strategies only monitor one slice of shopper behavior. That's because many physical AI tools still focus almost exclusively on checkout actions like scanning motions, barcode visibility, or item flow across a register. These approaches can catch visible mistakes, like an item left on the counter or a barcode accidentally covered, but miss all the products that never make it to the checkout in the first place.

When theft is treated as a checkout anomaly, retailers often tighten controls with receipt checks and guards, increasing friction for everyone, or accept rising losses as the cost of doing business. Neither are sustainable. The most important shift retailers can make is capturing the entire product journey.

Preventable loss doesn't occur at a single place in a store. It unfolds across a sequence of actions like when someone picks up an item, moving through the store, interacting with other products, and ultimately exiting. Systems that only observe the end of that sequence are operating without context.

By contrast, when technology is designed to understand the full product journey from shelf to exit, it becomes possible to distinguish theft methods and engage in more proportionate responses. Reducing shrink can't require treating every customer like a suspect. It requires knowing when to not intervene and how to intervene if necessary. Securing more information on loss also empowers retailers to better allocate resources and adjust store formatting.

One size fits all is the enemy of customer experience

A common failure in loss prevention is uniform response. When an AI system can't tell the difference between mistake and theft, every incident triggers the same alert, the same escalation and creates the same tension. This is how retailers end up embarrassing honest customers over minor mistakes or overwhelming staff with noise until critical alerts are ignored.

The alternative is to respond based on risk. When systems understand the energy drink taken on aisle 12 did not get scanned, retailers can align interventions with their values and policies. The proper response might be a gentle automated nudge at the self checkout or a pause for verification at the retailer's discretion. The key is that AI informs, but retailers decide how to respond.

It's tempting to just accept some customer friction as the unavoidable price of security. But friction is often a symptom of low situational awareness. When retailers lack confidence in what's happening, they overcompensate with more checks and more confrontations such as locking up high-value items or scrutinizing each receipt at the exit. The irony is that better detection enables gentler intervention.

Customer experience means privacy is a design advantage

Retailers face another difficult decision about whether to use AI technology for loss prevention with or without facial recognition and identity-based tracking. Facial recognition can improve re‑identification, but it also introduces significant privacy, regulatory and reputational risks — so many retailers are evaluating privacy‑first alternatives. Beyond regulatory and ethical considerations, customers simply don't want to feel like their privacy needs to be surrendered or their data are being stored in a system.

Systems that prioritize anonymous tracking and avoid biometric identity are still highly effective at reducing shrink while sidestepping the reputational and legal risks that come with storing sensitive personal data. More importantly, retailers reinforce a simple message to customers: the store only wants to make sure an item is paid for and is not concerned about who the shopper is. In an era where trust is a competitive differentiator, that distinction matters.

Shrink isn't going away. Economic pressure, and everyday opportunistic theft will continue to challenge retailers. But heavy-handed enforcement or experiences that make honest customers feel policed will just push them further from brick and mortar retail. Addressing this challenge starts by simply asking whether the technology actually sees what matters. When most high-risk theft is concealed long before checkout, systems built to watch only the transaction will always fall short.

When that vision improves, shrink goes down and the customer experience gets better. The most effective loss prevention strategies aren't those that catch more people, but those that intervene less often, with greater precision and confidence.

About Daniel Gabay

Daniel Gabay is Co-Founder and Chief Executive Officer of Trigo, a computer vision leader offering retailers full-store, full product journey loss prevention to ensure items picked up from store shelves are scanned at checkout. Under his leadership, Trigo has evolved from autonomous store innovation into a multi-layer, real-time loss prevention platform.

Connect with Daniel: