![logo](https://images.prismic.io/ecr-group-staging/3f03199b-b4d4-4a37-89e8-795fe3296228_home-depot-checkout.jpg?ixlib=gatsbyFP&auto=compress%2Cformat&fit=max&w=640&h=400)
The ECR Self-Checkout (SCO) working group met on the 27th of October to discuss the ways retailers were measuring the performance of their Fixed SCO systems. The session began with Professor Adrian Beck, who opened with an overview of the types of metrics that could be tracked, broken into four categories:
1. Utilisation indicators (number of transactions, etc)
2. Service indicators (customer satisfaction audits, etc)
3. Intervention indicators (number of interventions, etc)
4. Loss Indicators (number of items not scanned, etc)
While for the first three categories, the data to monitor, target and act can be acquired with varying degrees of ease, the loss indicators and attempts to record the absence of something happening, such an item not being scanned for instance, is inherently difficult. It was no surprise then that retailers had more metrics on the easier to measure KPI's than the more difficult to measure loss indicators.
This blog summarises the insights for each KPI category based on the group discussion
1. Utilisation Indicators
All retailers were tracking how many and what percentage of their shoppers are using the Fixed Self-Checkout terminals by measuring the number of transactions. Numbers ranged from 50%-60% of all transactions. The average basket size and average transaction value was also usually being tracked. The data from these metrics help support the design of labour planning schedules for SCO hosts, to set and monitor [say] participation targets and lost productivity values. Retailers were also looking at utilisation by payment method, with one UK retailer encouraging their shoppers to use contactless payments, with >80% as the goal.
2. Service Indicators
Most retailers were using some form of customer survey, often employing multiple methods at once, to determine customer satisfaction levels. Some used live-feedback, built into the screen of the terminal or a stand-alone podium featuring a four button face emoji rating system located at the exit of the Self-Checkout Corral. Many retailers reported that they would target a ratio of operators to terminals with numbers ranging from 1:6 to 1:12, the latter currently being tested by one retailer. It was not clear how compliance could be easily tracked, however there were some who tracked response time to interventions as a proxy for SCO Host compliance.
3 Intervention Indicators
Retailers were tracking interventions, the products that cause the most interventions, the speed of response and reason codes. One retailer indicated that they are now tracking interventions down to the SCO host level.
Intervention rates vary. In a grocery store that sells spirits, there is likely to be a high number of interventions than a store that does not sell spirits. As a consequence, the range of interventions per one hundred transactions spread from one retailer where only 3% of all transactions required an intervention, through to 30% plus for others.
Age verification and weight variance were the two most common causes of the need for a SCO host intervention. One retailer shared that these two reasons accounted for 80% of all retailer generated interventions.
Retailers recognised the opportunity to reduce the weight-based interventions by the updating of the weight database and the tolerance for each product on a regular basis. Which is why many also track the number of interventions needed per item.
Participants had little hope for the age verification issue. A few AI-based projects had to be scrapped due to GDPR issues, however in the UK, a government based initiative has encouraged retailers to trial new age verification technology. The results of these trials will be published in June 2022. Click to learn more about the trials
4. Loss Indicators
As the Professor noted, measuring an event that did not happen, such as a non-scan is profoundly difficult. To date, the retailers have relied upon control Vs test, or before Vs after SCO deployment studies to determine the losses. One European retailer in this session shared that their loss numbers went up from 1.9% to 2.3% after the implementation of their fixed SCO terminals in a selected number of stores.
A body of evidence is available in the 2018 ECR report (click here) on the nature and scale of loss at the Self-Checkout, one of the most used data points is that for every one % of store sales that are processed by Fixed Self-Checkout, there is a one basis point increase in unknown loss (also known as shrink).
Increasingly, retailers are looking to technology, especially Computer Vision, to understand, and then prevent losses through scan avoidance; at least five of the retailers in the call had ongoing experiments in the use of such technology, with some on previous ECR working group meetings reporting very positive results.
In addition, retailers on the call shared that they were investigating “risk” transactions, with one retailer sharing that they had identified 16,000 risk transactions with the worst thirty offenders each accounting for over $100,000 of loss. .
To detect these high-risk transactions, several retailers shared how they are beginning to explore Exception Based Reporting (EBR) systems that can flag up those transactions where the customers are simply scanning one or more low-value items, like simple shopping bags or cut-price T shirts in the case of one fashion retailer. Said fashion retailer is not using scales to identify Miss-Scans.
Wrapping Up
When at the end of the session and in follow up emails, the participants were asked as to their key takeaway, overwhelmingly the answer was the insight on the need for the better tracking of abandoned baskets, and reason codes for whole basket voids.
Currently, the reason codes for these voids were not being recorded by the majority of retailers on the call. In a grocery retailer, the three most common reasons for a whole basket void were believed to be;
1) The Customer has switched terminals - for example because they did not realise that the terminal did not accept cash.
2) The Customer had to leave the store and the whole basket was simply left at the Self-Checkout terminal
3) The Customer either left the store with the basket knowing that the SCO host would not make an intervention, or had simply walked away not realising that their contactless payment card needed a chip and pin verification.
Some of the retailers on the call were able to share that they have or are planning to have reason codes now included on the terminal screen, with the SCO host requested to allocate a reason code for each whole basket void. The increase in use of contactless payment is creating a sense of urgency, as the limits are increasing as the banks push for more digital transactions.
In conclusion, retailers track a lot, perhaps too much, possibly because they can and it’s easy to do but if there is no clear action that follows, is there a real point in tracking everything? It was also clear that Loss Prevention specialists are working hard to find ways to measure losses. An example would be walkaways. Potentially, the metrics emerging from new scan avoidance technologies may be another, however there is still a long way to go.
This SCO working group is for retailers, producers and academics only, If you would like to join this working group please send an email to Colin Peacock, at colin@ecrloss.com
Nov 8, 2021
Main office
ECR Community a.s.b.l
Upcoming Meetings
Our Latest news
Join Our Mailing List
Subscribe© 2023 ECR Retails Loss. All Rights Reserved|Privacy Policy