While self-checkout (SCO) technologies are not a recent development within retailing, in the last ten to fifteen years the pace of development and adoption has quickened considerably, especially in the grocery sector. In addition, there are now far more ways SCO is being realized in a broader range of retail environments. While fixed SCO machines or robots remain the most dominant system in operation, retailers are offering their shoppers alternatives, including:
Scan and Go: store devices provided to consumers for scanning items they wish to purchase.
Mobile SCO: consumers’ handheld devices that allow them to scan and, in some cases, pay for items.
Smart Trollies or Carts: some of which can detect items placed within them.
Whole Store SCO Systems: allow shoppers who have registered and scanned themselves into a store to pick up items and leave without further scanning or paying for a purchase.
Growing Concerns
Unquestionably, self-checkout allows retailers to considerably reduce labor costs and provide some consumers with greater convenience and choice in how they can check out. But recent studies have documented how they can also increase retail losses, primarily due to how they generate opportunities for malicious and non-malicious events to occur.
Examples include:
- Customers accidentally or deliberately not scanning some of the items they wish to purchase.
- Users misrepresenting one weighted product for another (the grapes for carrots “trick”).
- Customers using pre-printed barcodes to undervalue more expensive items.
- Customers carrying out “walk-aways,” where a customer scans all their items but does not pay for some or all of them.
Not only do such incidents generate direct losses for retailers, but they can have a negative effect on in-store stock file accuracy (this often happens without a retailer becoming aware until their next stock audit), which can generate indirect losses through missed sales due to out-of-stocks and increases in product wastage.
A 2018 ECR Retail Loss report estimated that for each percentage of retail sales processed through fixed SCO, a retailer will suffer an increase in an unknown loss of one basis point. This means that if 50 percent of sales value goes through this type of system, then a retailer could see an additional loss of 0.5 percent of their retail sales. With some estimates suggesting that unknown losses (shrinkage) for a grocery retailer might be about 1.5 percent of retail sales per year, this would represent a 30 percent increase in losses. More significantly, estimates for other types of SCO system losses were even more concerning—scan-and-go and mobile rates of losses could be higher, possibly as much as seven to ten basis points of loss for each 1 percent of transaction value.
Despite growing awareness of the scale of losses increasingly associated with SCO, the predominant trend is for retail to introduce more of these systems—bringing them into different retail environments, such as apparel, and extending the options available, such as mobile SCO. The question is not whether retailers will use SCO but how to better manage and control SCOs to ensure they don’t hinder profitability—especially since they account for a growing proportion of sales.
Responding to the Problem
As other ECR research on SCO has identified, retailers are now experimenting with, installing, and adopting a plethora of interventions to manage the SCO environment. These have largely been focused on key points within the shopper journey: registration, entering the store, selecting product in-aisle, the checkout area, and exiting the store. In addition, they have coalesced around four broad themes: the application of various types of technologies, such as video technologies and weight-based interventions; changes in how guardianship is delivered, such as changes to the selection and training of SCO supervisors; adjustments to store processes, such as closing fixed SCO machines in off peak times; and changes to the design and layout of stores and SCO-specific environments, such as the use of corrals and exit control gates.
The Key Role of SCO Supervisor
Often referred to as SCO supervisors or assistants, staff employed to provide oversight and assistance to customers who are checking out via fixed SCO machines and those using current scan-and-go systems are a critical part of many store teams. Typically, they are responsible for several fixed SCO machines and routinely respond to alerts and problems customers generate and experience (such as age verification and security weight-check issues).
In addition, they are expected to act not only as a form of deterrence to SCO misuse by customers (such as not scanning items) but respond when they become aware of customers engaging in malicious and non-malicious behaviors. As such, there is an expectation that they will be reactive (fix problems as they occur) and proactive (stop theft and errors occurring through their presence and oversight) in this space.
In terms of scan-and-go systems, these staff are primarily required to ensure customer problems with the technology and checkout process are quickly resolved and to perform system-generated audits of a select number of customers. For the most part these are partial audits where an algorithm decides which customer should be checked and how many items in a customer’s cart or basket should be scanned for accuracy.
For instance, the system may require a member of staff to scan five items in a customer transaction supposedly containing thirty items. In many respects, this is a difficult task for staff to perform—not only in terms of deciding which products to scan but managing what can be seen as a “challenging” interaction with a customer (in effect, the audit is suggesting the customer may have not scanned all the items, and the customer’s transaction and, by implication, honesty, needs to be checked).
Understanding the SCO Supervisor Experience
Much of the research to date has focused on measuring the impact of SCO on retail profits, or better understanding the approaches and views of retail organizations deploying them. The results presented below seek to broaden this understanding by adding the views and experiences of those tasked to operate these systems and working on the “front line” of SCO.
The research is based on online responses from over 6,000 staff working in nine retail companies—two based in the US, two in Australia, and five in Europe. All were large retailers and predominantly grocers, with a combined annual turnover of $320 billion, a total of nearly 12,000 stores, and together employing approximately 1.4 million staff.
The study focused on gaining insights into their work experiences, including the training they received, what they witnessed, and their views on how the work may be better controlled and managed.
Key Findings
The Self-Checkout Working Environment:
- Most SCO supervisors usually work on their own (59 percent), with the largest proportion having responsibility for seven or more SCO machines (38 percent).
- Nearly two-thirds do not believe they can cope with their allocation of SCO machines or can only cope when they are not busy (63 percent). As the allocation of SCO machines increases, staff feel less able to cope.
- Most respondents believe between one and six machines per member of staff is the optimum ratio (84 percent). While SCO machine ratio is a useful indicator, allocating labor to SCO spaces should also take account of transaction volume and velocity, the alerting context (age-checking requirements), use of control interventions (security weight checking), and the crime-risk rating for any store.
Self-Checkout Training:
- Most staff had received training prior to working on SCO (74 percent), although those who worked only on SCO when it was busy were more likely to have not received training.
- Those who had received training were more likely to feel they could cope with their workload.
- One in three staff receive training for only an hour or so prior to working on SCO (35 percent). Those who received longer periods of training were more likely to say they could cope with their workload and deal with non-scanning misuse.
- Most staff who receive training find it useful. The length of training was associated with perceived usefulness; the longer the training, the more useful it was viewed.
- Most training programs include aspects of LP, although one-third stated it did not (36 percent). When staff receive this type of training, they almost unanimously think it is useful (92 percent).
- Longer courses typically cover more SCO-related topics; The shortest programs (just an hour or so) cover 30 percent fewer topics than those lasting the longest (two days or more). Those who received more topics typically felt better prepared to cope with their SCO environment.
- Respondents thought training could be improved by including the following: more on LP (greater awareness of, and capabilities to deal with, theft in general); dealing with technical issues when SCO machines had faults or glitches; and customer service issues, including dealing with negative or difficult customers, how to better multitask and be more vigilant, and how to effectively engage with customers.
The Self-Checkout Operating Experience:
- On average, respondents believe 51 percent of all SCO losses are caused by malicious customer behavior.
- Respondents stated the most frequent process-related issues or alerts they experienced are age-related checks, having to remove security tags, and dealing with security weight alerts—all occurring many times an hour. They also noted bar codes not scanning, problems with loose item looks, and payment issues were also regular events.
- When it came to the frequency of customer misuse, non-scanning, scanning but not paying, and misusing the product lookup function were the top three, with at least one-third saying they happen at least hourly. The data points clearly to staff believing SCO misuse by customers is common.
- One-third of staff (32 percent) stated incidents of violence and verbal abuse occur very often or often in their SCO space (at least weekly). The survey also found a correlation between how frequently violent and verbal abuse incidents were thought to be happening and the number of SCO machines respondents had to supervise. Those with more SCO machines were more likely to say incidents happen more often.
- Respondents were almost unanimous in their view that having additional staff to supervise the SCO area would be an effective or very effective intervention to deal with SCO problems (96 percent).
- SCO staff provided recommendations for how their working environment could be improved, particularly in three areas:
- Guardianship: More staff, better trained, and greater management awareness of the issues affecting SCO.
- Technology: More robust, up-to-date, and
better-designed SCO machines, combined with alerting to help identify problems; more reliable security weight checking; the ability to monitor and respond remotely; and some form of active exit control. - Design: More space to enable staff to have easier access and improved lines of sight of SCO machines.
Improving Retail Self-Checkout
The results from this study have provided valuable insights not only into the working experiences of SCO supervisors but also how these types of programs can be improved. In particular:
- The Value of Guardianship: The importance of SCO supervision and how retailers need to move from a “minimal possible” staff operating model to one more focused on creating an “optimum control’” operating environment.
- The Importance of Training: Recognizing the value of training and why it matters. Staff are keen to see their training include more on LP-related issues, how to deal with problematic SCO technologies, and improving their customer service skills.
- Providing Technological Support: Respondents were almost unanimous in expressing frustration with the quality and reliability of SCO technologies. They want more reliable, up-to-date, and better-designed systems. In addition, they are keen to have a range of alerting options that will help them do their work more effectively.
- Designing Zones of Control: SCO spaces need sufficient space to allow staff and customers to move freely and easily, encompassing clear and unambiguous ingress and exit controls. Above all, the design of SCO areas needs to ensure it amplifies a sense of risk for the would-be offender and enables staff to effectively play a reactive and proactive role.
- Keeping Guardians Safe: Retailers have a commitment to protect their staff and ensure they are creating a safe working environment for them. The survey revealed worrying levels of violence and abuse, and respondents believe, in part, this is due to frustrations generated by current staffing levels and inadequate, glitchy, or poorly designed SCO technologies.
- Better Measuring of the SCO Environment: More needs to be done to better measure factors that affect the management and control of SCO spaces, including the frequency of SCO alerts; average SCO alert wait times; average SCO queue times; the number of voided, abandoned, or walkaway incidents; and the number of incidents of violence and verbal abuse.
Learning From Front-End Teams
This research has provided a wealth of insights into how SCO teams try to manage and control the SCO environment. Above all, the research has highlighted how important people are in this process, but if they are to be successful, they must be given a manageable workload; receive high-quality training; be supported with technologies that help them focus on the customers, transactions, and events important for their role; and operate in a space that is well designed and, above all, enables a zone of control to be maintained.
Download the full report here.
After graduating from the University of Leicester in 1988, Professor Adrian Beck worked with John Benyon and others to establish the Centre for the Study of Public Order, which became the Scarman Centre in the mid-1990s and subsequently the Department of Criminology in the early 2000s. He was head of the Department of Criminology from 2009 until 2015 and in 2017 became emeritus professor at the University of Leicester. His research sponsored by the Retail Industry Leaders Association led to the development of the Total Retail Loss concept. Beck is currently academic advisor and researcher with the ECR Retail Loss Group (ecrloss.com). He can be reached at beckbrc@gmail.co.uk.