
By Emeritus Professor Adrian Beck
Introduction
A perennial concern for many retailers is the extent to which they are victims of thefts where stolen items are simply wheeled out of their stores using carts provided to consumers – cart push-out thefts. However, as with many forms of retail loss, moving from anecdotal concern to verifiable evidence is less than easy – the amorphous mass that is ‘shrinkage’ by its very nature offers little in the way of hard data on its underlying causes – fundamentally it is predominately a record of losses where the cause is unknown.
This routinely leaves the loss prevention team having to make rather generalised assumptions about what problems to prioritise: external theft, internal theft, supplier fraud, lack of adherence to store processes? This in turn influences decisions about what interventions to invest in to try and reduce the cost of loss to the business. Not only must loss prevention leaders make a judgement call on what causes are driving the greatest losses, but also then decide what will be the most impactful ways in which they can be addressed – not easy work!
One way to address this is to try and ‘unpick’ the shrinkage number – identify elements of loss that can be verifiably associated with various types of cause – in effect turn some of the unknown into known. There are various ways in which this can be done, one of which is to carry out detailed studies and experiments focussed upon generating robust data on the extent of any given problem. Equally, the same can be done to measure the impact of any given intervention. The key is using data collection methods and analysis that go beyond just the shrinkage number.
Researching Cart Push-outs
This was the approach taken to understand not only the scale and extent of cart push-outs but also to assess the impact of a particular intervention: Gatekeeper’s Purchek wheel locking system. In simple terms, the system is designed to physically stop one of the wheels on a shopping cart turning at the exit of a store if the cart has not been through a predetermined payment area, such as a checkout lane. It also triggers an alarm and in some circumstances video recording of the incident.
This has two potential effects: first, it makes it much easier for store/security staff to identify which customer has triggered the alarm – look for the immobilised cart! Secondly, it significantly amplifies the offender’s sense of risk of apprehension – the cart is now very difficult to move quickly, and staff will be aware of which cart it is and who is associated with it – time to make a run for it!
Working with two of the largest UK retailers I carried out a detailed study to measure the losses they were experiencing from cart push-outs and what impact the Purchek system had on these losses. Using data from 239 stores across a 15-week review period, as well as analysing the context of over 900 incidents, the study provided some of the first independent insights into this issue.
Here are some of the research highlights:
- A Grocery store with annual sales of approximately $54 million, is likely to experience at least 140 cart push-outs a year (almost three a week), costing in the region of $43,000 a year.
- Cart push-outs may account for approximately 4% of all unknown stock loss (shrinkage).
- The average cost of a cart push-out incident is $305, although some incidents were valued at more than $2,750.
- Of those stopped, 21% went on to pay for the products they were trying to steal.
- There was little evidence that cart push-out interventions were a generator of violence and/or verbal abuse (just 3% of recorded incidents) – offenders were more likely to run off, make excuses and leave, or go back and pay.
- The study showed a marked decline in the number of incidents and weekly value of losses associated with cart push-outs after Purchek had been installed for three weeks. The number and value of losses declined by on average 49% and 47% respectively after the initial period of use.
- Return on Investment modelling revealed that for a Grocery store with annual sales in the region of $54 million, the Discounted Payback Period was just over 1 year, generating a Net Present Value of $63,400 and an Internal Rate of Return of 70%.
- The research highlighted the value of Alarm Activation Visibility or Kinetic Crime Prevention – offering security staff a much more unambiguous method of identifying which customer had triggered the exit alarm.
- A key facet of Purchek was its capability to deliver deterrence through an initial period of detection – a ‘golden period’ of potency in the early weeks after installation. For this to be achieved, however, the research found that it was critically important store staff were made available to respond to the alerts.
Conclusions
The research generated a considerable amount of other data on this topic, including insights into the types of offenders involved – further information is available by contacting the team at Gatekeeper. What was key to this research was the development of a methodology to measure the scale and extent of the problem of cart push-out thefts – securing verified evidence to contextualize longstanding industry concerns. This then enabled a detailed assessment of the Purchek system to be undertaken to better understand it effectiveness and its operational context. In many respects, this is a good example of the appliance of science!
Download the full Study

