In a world of busy schedules, endless retail options and new technologies designed to make everything more convenient, today’s consumers have far less patience for lines than they did in the past.
While retailers offer online ordering, in-store pickup and delivery to increase convenience, many are also looking to automated and mobile checkout to reduce lines. For retailers, these technologies offer the opportunity to improve the customer experience and devote staff resources to more value-added activities. For consumers, the ability to shop faster and more efficiently reduces the time commitment of shopping.
New technologies enabled by smartphones, cameras, sensors and artificial intelligence might soon reduce the relevance of the traditional cashier-based checkout system. Amazon’s Just Walk Out technology automates the entire checkout process with a system of cameras, weight sensors and AI to detect merchandise consumers take from or return to shelves. The system automatically tallies items in customers’ baskets, then charges their Amazon account as they leave the store. As of July, the ecommerce giant was operating 17 such stores in Chicago, Seattle, San Francisco and New York City and could open 3,000 Amazon Go outlets by 2021.
Many other retailers are trying cashierless checkout systems powered by apps and consumer mobile devices. In November 2018, Sam’s Club opened a Sam’s Club Now store in Dallas that operates with no cashiers. The 32,000-square-foot facility uses 700 cameras in tandem with an app and electronic price tags that can be updated in real time. Shoppers use the Scan & Go app to scan their products and pay for them on the way out.
Kroger, which operates more than 2,500 supermarkets in 35 states, is piloting a Scan, Bag, Go option that enables consumers to scan their own items and pay on the way out with their device or through a human cashier. Meijer introduced its Shop & Scan capabilities to 23 stores in the Chicago area in March after a successful test run at more than 50 stores in Michigan. The mobile app was downloaded more than 1 million times, and more than 80 percent of customers with the app used the Shop & Scan feature multiple times.
Whether it’s a fully automated system like Amazon’s or a consumer-driven mobile app, cashierless checkout offers several major benefits, says Jerry Sheldon, a retail analyst with IHL. It improves the customer experience and brand perception by increasing convenience while reducing the wait. It also enables retailers to shift human labor from the checkout to more valuable activities like greeting and engaging customers on the floor.
“Most consumers feel the checkout process is a drag and a waste of time,” Sheldon says. “It enables [retailers]to focus more on customer interaction and helping shoppers.”
And consumers generally like anything that increases convenience. According to GPShopper, nearly half of U.S. internet users believe scan-and-go technology would make in-person shopping easier, and 43 percent would rather try scan-and-go than wait in a checkout line. Research from MuleSoft said 60 percent of global internet users would prefer to shop at other retailers if they offered an experience like Amazon.
“People are just growing less patient of inconvenient checkout experiences,” says Ran Peled, vice president of marketing for retail automation platform Trigo Vision. “Checkout remains the biggest pain point for customers.”
‘Scan and go’ or ‘just walk out’
There’s a big difference between mobile scan-and-go technologies and Amazon’s “just walk out” technology. In the scan-and-go application, the consumer serves as the cashier and conducts the transaction using a mobile device. It’s a first-stop experience that’s already gaining adoption because it reduces complexity in the checkout process, and because most consumers already have a smartphone with the capabilities to support it, Sheldon says.
“From the barrier-to-adoption standpoint, the consumer mobile for self-checkout makes a lot of sense,” he says. “It’s where a vast majority of retailers will begin.”
The other end of the spectrum is technology that uses cameras and AI to automate the entire checkout process. In a model like Amazon’s, neither consumers nor retail associates must scan items nor process any payments. But the technology is more complex, requires more equipment and can be more costly to implement, Sheldon says.
New systems entering the market allow retailers to test the waters with flexible approaches. Grabango, a Silicon Valley startup founded in 2016 with a focus on computer vision and machine learning, offers a flexible “checkout-free technology” that enables the retailer to determine the desired level of human involvement.
The lightweight system is easy to install and retrofit to existing stores, says Andrew Radlow, vice president of business development with Grabango. High-quality sensor hardware and high-precision computer algorithms can adapt to the existing store format, acquire the location of every item in the store, then produce a complete real-time planogram. The system keeps a running total of the items shoppers select and can even offer product information or coupons. When finished, the shopper can pay by smartphone or at a cashier.
Many of these new checkout applications are taking off in supermarkets because of the high inventory turnover. It’s also the retail sector consumers visit the most: The typical consumer now makes 1.6 trips each week to a grocery store, according to recent data from Statista, and the time spent driving, in traffic, wandering the aisles and standing in line can easily add up to a few hours per week.
The large inventory volume at checkout, high volume of customers and number of cashiers typically needed makes automated checkout an attractive option in grocery. “It’s low-hanging fruit, and probably more attractive than in any other industry vertical,” Sheldon says.
In July, Giant Eagle, which operates more than 470 stores across the country, announced it was piloting the Grabango “checkout-free platform,” the first for a large grocer in the United States. Computer vision and machine learning automatically tally items as consumers shop; they can then make payment on their mobile device or to a cashier as they leave.
Other systems can be deployed in existing stores. Trigo Vision installs off-the-shelf cameras in the ceiling and uses commodity hardware to process complex AI algorithms. It calibrates the cameras and uses innovative sensing models to provide 3D reports that can identify products by customer movements and the interaction between hands and shelves, Peled says. Last year, Trigo Vision worked with Shufersal, Israel’s largest supermarket chain, to install its platform in 272 of its stores.
“We don’t rely on a planogram,” Peled says. “Our system is good enough to understand the product you took, regardless of where you took it from, and then delete it from your shopping basket regardless of where you put it back.”
Combatting shrinkage
While self-checkout and automated solutions offer many benefits, there’s a perception they can also lead to greater theft. One survey of 13 major retailers operating in the United States and Europe found levels of stock loss were 1 percent in stores with self-checkout, compared with 0.67 percent of those with staffed checkouts.
Although some loss can be attributed to mistakes on the part of the consumer, self-checkout can expand opportunities for theft. Many common self-checkout theft tactics, such as bagging items without scanning them or scanning lower-cost items while bagging more-expensive ones, can be replicated with scan-and-go technologies.
After expanding it to more than 100 stores, Walmart ended its Scan & Go pilot after only a few months in May; a former head of checkout innovation told Business Insider theft was a major problem. At the time the program ended, Walmart said the program had low adoption in addition to some of its own friction points “such as receipt checks, weighted produce and un-bagged merchandise resulting from using the program.”
Despite the concerns, the potential for shrink can be overcome with the right optimization of staff and technology, Sheldon says. Spot checks by associates combined with new camera technology and AI can also help retailers identify behaviors related to nefarious activities.
But in contrast to basic self-checkout, many of these new technologies can even reduce theft and error, says Michael Suswal, COO and co-founder of Standard Cognition. For example, Standard Cognition’s AI-powered autonomous checkout platform can detect when someone is not logged into the system and it appears they are about to leave the store with items they have not scanned. The system can then alert associates about the issue.
When armed with the right information, associates can remedy the problem or prevent the theft with non-confrontational questions, such as asking the person if they need help downloading the app or if they were able to find everything they were looking for.
“It’s not a loss reaction system, it’s a real loss prevention system,” Suswal says. “And the best part is they don’t need to worry about accusing anyone of theft because they’ve done nothing wrong.”
Grabango can even charge potential shoplifters for all items on their person, even if hidden from view at checkout or while leaving the store. As the system does the work, store personnel don’t even need to confront suspected shoplifters.
The future of checkout
As retailers expand mobile point of sale in the hands of associates, and as consumers increasingly make purchases from their mobile devices, the industry will inevitably move to more frictionless checkout experiences. Boston Retail Partners noted in a recent report that retailers’ use of consumer-facing mobile apps for POS will continue to increase as a means of self-checkout. The number of retailers offering customer-device POS grew from 8 percent in 2017 to 22 percent in 2018.
Retailers using scan-and-go largely report positive experiences. Eddie Garcia, vice president of product and member experience at Sam’s Club, said in a blog post that since launching the Scan & Go app two years ago, the popularity has continued to increase. This spring, the retailer began testing new computer vision and machine learning technology to make that service even faster and easier.
“Rather than having to locate the barcode and scan it just right, the camera in the app will identify the product with a simple hover and add it to the member’s shopping list,” Garcia said. “It’s as simple as that.”
Retailers that aren’t yet piloting such applications are hindered by IT bandwidth issues and other priorities such as optimizing buy online, pick up in store options. “Many grocers are really optimizing the buy online, pick up in store experience,” Sheldon says. “I think they will continue to look at some of these self-checkout technologies, but they’re busy addressing click-and-collect challenges.”
As cashierless technology providers move to the future, they’ll have to “at least match or beat the accuracy of human cashiers,” Suswal says. Standard Cognition’s research shows that to be roughly 97 percent, a rate the company can already exceed. Continuous learning enables the system to better understand item classification and behavioral analysis with reinforcement about its results.
“We do this as an iterative process repeatedly,” Suswal says. “Whenever we see confidence dips in the system, we’ll go back a day or two later, gather them all up and start retraining those areas.”
Whether it’s a scan-and-go application or a just-walk-out system, retailers will also want flexibility to adopt such technologies to fit their stores and business models. The best technologies will be those that seamlessly blend into the store and user experience.
“It should look like a regular store. We don’t want machines everywhere and robots. Our vision is to bring back the world feel of a market but with all the technology that allows your shop assistants to be a little more personal,” Suswal says.
As even retailers that use scan-and-go continually try to make their experiences even more seamless, much of the retail sector will move to a completely frictionless model, Sheldon says. Improvements in the technology, declining prices due to economies of scale and consumer demand could force the scale to tip quickly
Peled believes consumers can expect to see more of the technologies within the next couple of years. “It will start slowly but once some technology breakthroughs are achieved, and a few retailers are trusting that this is going the right way, it will [happen quickly],” he says.
Craig Guillot is based in New Orleans and writes about retail, real estate, business and personal finance. Read more of his work at www.craigdguillot.com.