Self-checkout changes more than the way customers pay. It also changes who takes responsibility – and where risks arise. When shoppers scan, bag and pay for their own purchases, the opportunities for mistakes, unintentional losses and deliberate manipulation increase. This makes loss prevention at Self-Checkouts a major challenge for retailers.
More importantly, loss prevention is a fundamental design consideration. It should not begin once an SCO area is operational. It needs to be addressed from the very start of the planning process.
That is why Level 5 of the Multi-Level Approach (MLA) focuses on Fraud Detection & Risk Management. It builds on the previous levels to establish a systematic approach to identifying, assessing and mitigating risks.
Self-checkout creates new behavioural environments
An SCO area is far more than a collection of machines. It is a behavioural environment in which people with different abilities, experiences and intentions make decisions and carry out transactions independently.
Losses can occur in several ways. Customers may accidentally fail to scan items, select the wrong products or leave transactions incomplete. Deliberate manipulation adds another dimension, including intentional non-scanning or barcode switching.
The scale of these risks is illustrated by the Self-Checkout Loss Report 2026 from ECR Retail Loss. Based on research involving 39 retail companies, the report identifies an average 22% increase in merchandise losses during the first year following the introduction of self-checkout. It also highlights the potential of organisational, design-related and technological measures to reduce losses.
The crucial question, therefore, is not simply whether an SCO system has security features. It is how effectively those features are integrated into the overall operation.
Without reliable data, self-checkout loss prevention remains reactive
Before retailers can reduce losses, they need to understand where they occur and what causes them.
This is where MLA Level 5 connects directly with Level 3 – Network & Analytics.
Transaction records, voids, staff interventions, error messages and unusual transaction sequences can all provide valuable insights. Analysed systematically, these data can reveal recurring anomalies at particular terminals, within specific product categories or during certain stages of the checkout process.
However, an anomaly is not necessarily evidence of theft. An unscanned item may just as easily result from confusing on-screen instructions, an operating error or a technical malfunction.
Collecting data alone is therefore not enough. Information must be accessible, reliable and meaningful within its operational context. Only then can it provide a sound basis for informed decisions and targeted preventive measures.
Loss prevention at self-checkouts goes beyond security technology
In practice, loss prevention is often associated primarily with cameras, security scales and other monitoring technologies.
These systems can certainly play an important role. Their effectiveness, however, depends on how well they work alongside SCO zone design, customer processes and staff involvement.
Consider a self-checkout installation without a conventional security scale.
Removing the scale also removes an established means of checking whether scanned items correspond to those placed in the bagging area. If the associated risks are to remain under control, alternative methods must fulfil the necessary monitoring functions.
Possible approaches include camera-based movement analysis, comparing product movements with recorded scan events, and monitoring defined pick-up and deposit areas.
When a system detects a potential irregularity, it might first display a clear, helpful message to the customer. If the anomaly persists, a member of staff can be alerted.
This graduated escalation process allows an appropriate response without treating every deviation as an attempted theft.
Alternative technologies cannot, however, be assumed to replace security scales completely in every situation. Their suitability and effectiveness must be assessed against the specific processes and risks involved.
Better usability also means better loss prevention
One frequently overlooked relationship is the connection between the usability of a self-checkout system and its vulnerability to losses.
Unclear instructions, poorly organised bagging areas and unnecessarily complicated processes increase the likelihood of customer errors. At the same time, they can create situations in which deliberate manipulation becomes harder to detect.
This makes MLA Level 4 – Usability & Customer Experience – an integral part of effective loss prevention. The relationship between operational data and user experience is explored in more detail in our previous article, Data and Usability in Self-Service.
Clear process guidance, unambiguous feedback and ergonomically designed workspaces help customers complete transactions correctly. They also make it easier for staff to distinguish genuine irregularities from everyday operating difficulties.
The objective should therefore never be to generate as many alerts as possible. What matters is identifying relevant events, minimising false alarms and enabling appropriate responses.
People remain an essential part of the security concept
Even the most sophisticated technology cannot replace the experience and judgement of well-trained staff.
SCO attendants help customers, resolve operating difficulties and respond to unusual situations. Their role combines customer service with important monitoring and loss prevention responsibilities.
To perform these tasks effectively, they need well-designed, clearly visible SCO areas, defined responsibilities and systems that provide understandable, context-sensitive alerts.
Overwhelming staff with a constant stream of poorly targeted alarms undermines both service quality and the effectiveness of loss prevention.
Data protection is equally important. Monitoring and analytics solutions must be designed to comply with applicable privacy legislation. Camera-based systems, in particular, require careful consideration of purpose limitation, proportionality and the processing of personal data.
Address risks before roll-out and not afterwards
The central objective of MLA Level 5 is to integrate loss prevention at Self-Checkouts into the overall concept from the outset. Risks should not simply be identified once a system is operational. They need to be systematically assessed during the planning and design stages.
This includes evaluating potential operating errors, deliberate manipulation and vulnerabilities within the checkout process. The findings provide a basis for selecting appropriate technical, organisational and staffing measures.
Their effectiveness should be tested in pilot installations and reviewed regularly throughout ongoing operations.
This approach helps avoid costly retrofits while balancing security requirements with commercial viability, ease of use and customer service.
Loss prevention does not begin with an alarm. It begins with system architecture.
This is precisely where the Multi-Level Approach delivers value. Its five levels are not treated as separate disciplines. Together, they form a coherent planning framework that connects commercial objectives, spatial design, data, usability and risk management.
Because effective loss prevention is not something to add to a self-checkout system once everything else is in place.
It is part of what makes the entire system work.
In the fifth and final article of our MLA series, we will explore a critical question: Who brings all five levels together, resolves conflicting priorities and takes responsibility for the system as a whole?
