The largest study of self-checkout loss ever conducted was published in June 2026, and its headline finding changes how retailers should think about the entire self-checkout question. Led by Professor Matt Hopkins of the University of Leicester and commissioned by ECR Retail Loss, the study analyzed millions of transactions across 39 retailers with a combined annual turnover exceeding one trillion euros. It included seven of the top 30 retailers globally. In stores where self-checkout is present, it now handles an average of 54 percent of transactions. Self-checkout has passed the staffed lane as the preferred way customers pay.
The study also confirmed the concern that has followed self-checkout since it launched. Stores with self-checkout record losses about 0.42 percentage points higher than comparable stores without it. In the first 12 months after installation, average store losses rose by 0.26 percentage points. Loss through theft increased by an average of 22 percent in the year after self-checkout was installed. Those numbers are real, and they are the reason some major retailers pulled back on self-checkout expansion.
But the most important finding in the entire study is not the size of the loss. It is the source of it. Colin Peacock, group strategic coordinator at ECR Retail Loss, put it in one sentence: most self-checkout loss is generated by ordinary people making everyday mistakes. That is good news, because it means retailers can engineer their way out of the problem. The shrink is not the technology. It is the deployment. And deployment is something a retailer can fix.
What the Study Found About Where the Loss Actually Comes From
The instinct when shrink rises at self-checkout is to assume theft. The data tells a more precise story. ECR research found that 52 percent of retail shrinkage is accidental, compared to 48 percent attributed to malicious intent. At self-checkout specifically, the losses come from a mix of causes that are mostly not deliberate. The study named them directly: honest mistakes with awkward items, unfamiliar user interfaces, and fiddly barcodes. Customers get frustrated when items will not scan, and staff can be slow to help.
The specific numbers make the point sharper. Mis-scanning was the most frequent source of loss, reported in between 1 and 4.8 percent of transactions. Losses tied to unpackaged products, the produce and bulk items a customer has to find in a list, occur in 20 out of every 1,000 self-checkout transactions when customers must select those items manually. That is not a security problem. It is an interface problem. The customer cannot find the code for loose apples, so the apples do not get scanned. The loss is real, but the cause is friction, not fraud.
Walkaways generate the highest loss by value, averaging 88 euros per event. A walkaway is what happens when a customer gets far enough into a frustrating transaction, hits a wall, and simply leaves without completing it. That is the most expensive form of self-checkout loss, and it is entirely a design and experience failure. Nobody walks away from a transaction that is working.
Why This Reframes the Whole Self-Checkout Decision
Theft You Manage. Friction You Engineer.
The distinction between accidental and malicious loss matters because the two require completely different responses. Malicious loss is a security and enforcement problem. It is hard to eliminate and expensive to police. Accidental loss is a design problem, and design problems have engineering solutions. When the ECR study found that most self-checkout loss is accidental, it was really saying that most self-checkout loss is addressable through better deployment rather than more surveillance.
This is the argument I have made throughout my work in self-service, and it is the reason I wrote a book called How to Implement Self-Service Without Failing. The retailers who struggle with self-checkout are almost never the ones with a technology problem. They are the ones who deployed the hardware without designing the experience around the specific ways real customers actually use it. The technology works. The deployment is where success or failure is decided.
The Nudge That Proves the Point
The single most convincing evidence that self-checkout loss is a deployment problem comes from the interventions that work. The ECR study found that nudges, simple signals to the customer that the system is paying attention, were hugely effective at reducing shrinkage. When a customer receives a gentle on-screen prompt to rescan a missed item, between 80 and 97 percent do so immediately. The item was missed by accident. The prompt gave the customer a chance to fix it, and they took it.
Think about what that number means. If a simple, non-confrontational on-screen prompt recovers 80 to 97 percent of missed scans, then the overwhelming majority of those missed scans were never theft in the first place. A thief does not rescan the item when politely asked. An honest customer who made a mistake does. The effectiveness of the nudge is the proof that the loss was accidental. As I described in the self-checkout bagging and loss prevention analysis, the retailers who manage shrink well are the ones who design for the honest customer’s mistake rather than treating every shopper as a suspect.
The Deployment Decisions That Separate High-Loss From Low-Loss Stores
The Produce and Bulk Item Problem
The finding that unpackaged products generate loss in 20 of every 1,000 transactions points directly to a fixable design decision. When a customer has to scroll through a list to find the code for loose tomatoes, the friction produces both abandoned scans and frustration. Retailers who have addressed this with better item-lookup interfaces, visual product recognition, or simplified produce workflows see materially lower loss in exactly this category. The problem is not that customers are stealing tomatoes. It is that the interface made scanning the tomatoes harder than it needed to be.
The Staff Availability Problem
The study specifically noted that customers become frustrated when items will not read and staff are slow to help. That is a staffing and floor-design decision, not a technology limitation. The retailer who places self-checkout banks where a single attendant can actually see and reach every lane resolves the friction before it becomes a walkaway. The retailer who installs twelve lanes monitored by one attendant standing thirty feet away has engineered the frustration into the layout. The hardware is identical in both stores. The loss is not.
The First 12 Months Problem
The study found that losses rose most sharply in the first 12 months after installation. That timing is a signal. The first year is when customers are least familiar with the system, when staff are still learning to manage the new floor layout, and when the deployment is least optimized. Retailers who invest in customer guidance, staff training, and interface refinement during that first year compress the loss curve. The ones who install the lanes and walk away absorb the full first-year loss as if it were an unavoidable cost of the technology. It is not. It is the cost of an unmanaged deployment.
What This Means for LatAm Retailers
Self-checkout adoption is accelerating across LatAm markets, and the deployment lessons from the ECR study are directly transferable, with one added consideration. In markets where self-checkout is newer, the first-12-months loss curve is the entire current experience for many retailers. They are living through the highest-loss phase right now. That makes the deployment decisions more urgent, not less, because the retailer who designs the experience well from the start avoids the loss curve that the study documents rather than absorbing it and trying to recover later.
Furthermore, the produce and bulk item challenge is often more pronounced in LatAm grocery formats, where fresh and loose items make up a larger share of the basket than in many U.S. or European stores. That means the interface design for unpackaged products is not a minor detail in a LatAm deployment. It is one of the highest-leverage decisions a retailer can make, because it addresses the single category where the ECR study found the most accidental loss.
The Real Lesson of the Largest Self-Checkout Study Ever Conducted
The Technology Already Won
At 54 percent of transactions, the debate about whether customers want self-checkout is over. They do. They value the speed and the control, and the adoption numbers confirm it across 39 retailers and eleven countries. The question is no longer whether to deploy self-checkout. It is how to deploy it so that the loss stays low while the participation stays high. Those two goals are not in tension. They are both outcomes of good deployment design.
The Problem Is Solvable, Which Means It Is a Choice
The most important implication of the ECR finding is also the most uncomfortable one for retailers. If most self-checkout loss is accidental, and if accidental loss can be engineered down through better interface design, staff placement, nudges, and first-year management, then a store with high self-checkout loss is a store that has chosen not to engineer it down. The loss is not a tax the technology imposes. It is the result of deployment decisions the retailer controls.
Self-checkout now handles the majority of transactions in the stores that offer it. The loss that comes with it is real, but it is mostly the honest mistakes of ordinary customers, not theft. That is the finding that should change how every retailer approaches self-checkout. The shrink is not an argument against the technology. It is an argument for deploying the technology properly. The retailers who engineer the friction out of the experience keep both the participation and the margin. The ones who blame the hardware will keep absorbing a loss they could have designed away.
If you are evaluating a self-checkout deployment or trying to bring down loss in an existing one, connect with me here or reach me on LinkedIn. I am happy to walk through the deployment framework we use across the U.S. and Latin America.
Adriana Rivas is a retail technology executive and AI strategist. She is the recipient of the Gold Stevie® Award, Thought Leader of the Year 2026, recognized by Thinkers360 as the #7 Global Thought Leader in Retail, and named to the RTIH Top 100 Retail Technology Influencers 2026. She is the author of How to Implement Self-Service Without Failing, now available in a Revised and Expanded Edition.