What does the future of self-checkout in retail look like?

Why are retail stores switching to self-checkout?

Self-Checkouts are not new to the retail sector, and decades of innovation and implementation of technology has seen the usage increase year on year. In 2021 the industry saw a significant 11% growth, and yet this is low compared to the forecasted 13.3% compound annual growth rate predicted for self-checkouts through to 2030.

Manufacturers brought over 200,000 self-checkouts to the market in 2021, and by 2027 global annual shipments are expected to surpass 300,000 units (see report by RBR). Customers are increasingly expecting that stores will adopt self-checkout, if they have not already, with some customers making grocery shopping decisions based on stores with self-checkout (SCO) available. The installation of SCO has become a competitive requirement for retailers needing to secure their customer base.

The industry, therefore, is accelerating rapidly, sweeping retailers and customers along in its wake. Self-checkout technology across retail is evolving, with aims to reduce staff interventions, improve shopper experience, reduce shrinkage, and improve throughput. Customer reaction to this technology can be polarised – some choose it by default, but others refuse on the basis of perceived complexity, and instead queue at a traditional checkout. Most will tend to use self-checkout but avoid them when they are buying any produce that has no barcode.

What are the challenges with self-checkout for retailers?

There is anecdotal evidence to suggest that as much as 80% of customer transactions in supermarkets could be processed via some form of SCO technology, and yet 3 in 10 customers will actively avoid the use of SCO if they have fresh produce or other non-barcoded items in their basket.

AI powered self-checkout solutions can alleviate customer frustrations, reducing the unexpected item in the bagging area blocker and reducing staff interventions by up to 45%. Speed to exit therefore, is the key to unlocking further potential in the SCO technology base – once a customer has spent time getting their items, they want to get through the checkout process quickly. The mindset shifts from ensuring they have everything they need to how quickly can they exit the store. A recent study showed that up to a third of customers will leave a long queue in search of a better checkout option, with 11% saying they would abandon basket entirely if the queue is too long, clearly an issue to be taken seriously.

There are, therefore, a number of challenges facing retailers, from the need to install or add to the number of self-checkouts in store, to encouraging more customers to utilise the equipment in place.

However, once the SCO is in place there are a number of additional challenges retailers need to take into consideration that may impact on their operational efficiency:

• Length of time it takes for a customer to select a non-barcoded item from the catalogue
• The consequences of inventory inaccuracies should the wrong item be selected
• The likelihood of a customer needing employee intervention before the transaction completes
• The increased risk of intentional or unintentional shrink and fraud.

For customers, the use of the self-checkout can seem complicated, they may fear choosing the wrong item, they may find increasingly long catalogue pick lists frustrating, and opt instead for the staffed checkouts, or they may take their frustrations out on the self-checkout attendants.

Conversely, it could be said that if a customer feels a store offers the best mix of products, a frictionless journey through the store, and a speedy exit, then these are areas that will strengthen customer brand loyalty.

What is the future of self-checkout?

When we take a step back from all this analysis, we see at the heart of the self-checkout evolution a customer demand for faster checkout options, coupled with a retailer demand for efficient infrastructure that frees employees up to perform higher value tasks, and for inventory inaccuracies to be reduced. Retailers need to maintain a balance of good customer experience and high retail efficiency. Shrink costs need to be recovered, and this often results in price increases on items, whereas enhanced retail efficiency reduces costs allowing retailers to retain their competitive edge.

When we approached the challenge of fresh produce recognition, or how to identify non-barcoded items at self-checkout, we began with an analysis of the customer journey through the checkouts. Some of the common issues that either lead to customers avoiding the use of the SCO or experiencing frustration whilst checking out are:

• Having to navigate a complex catalogue
• Remembering precisely which type of produce they had in the basket
• Choosing the wrong item from the catalogue
• Needing to get a store assistant to help with the transaction

For retailers, these challenges can have knock-on effects such as interventions from store assistants to progress the transaction, or longer queue times for customers, as transactions take longer to complete and customers choose to wait for staffed checkouts instead of utilising the SCO. In addition, there are inventory inaccuracies and shrink due to the wrong items being selected.

How can Computer Vision AI (Artificial Intelligence) benefit self-checkouts?

When we begin to think about what computer vision AI can do, the scope can seem vast. We believe the key is to focus on the required results and methodically match solutions to each challenge.

For example, if we take the issues listed for fresh produce recognition above and put it into a table, we can see how the challenges are mapped to parts of the solution, all of which make for an enhanced customer, staff, and retailer experience.

How much time does self-checkout save? Fresh Produce Recognition solution saves between 5-15 seconds per type of fresh produce at the self-checkout, saving customer time, shortening queues and improving customer experience

By understanding the challenges, the solution is built to address each issue and, in turn this addresses the frustrations for both customers and retailers. By removing the need to search huge catalogues this saves, on average, between 5 and 15 seconds per transaction meaning transaction speeds can be increased by up to 67%. In speeding up transactions, queue lengths are reduced improving the customer experience in store.

What does this mean for customer-retailer loyalty?

In a recent study of customer brand loyalty, customers listed their most important factors for grocery stores as value for money, quality of products, convenient shopping experience, ease of making purchases and a positive shopping experience.

At a time when retailers are competing for customer loyalty, a deep understanding of the optimal customer journey and customer experience is essential. Rising costs such as supply chain issues, price of raw materials and inflation put pressure on retailers, which can lead to cost of goods increasing and adds to the risks of customers taking their grocery shopping to a competing supermarket.

Opportunities to make the best use of existing infrastructure provide a valuable and simple action that retailers can take to improve inventory accuracy, reduce friction at self-checkout, and enhance customer experience.