The Must-Dos for Successfully Implementing Frictionless Checkout
NATIONAL REPORT — Frictionless checkout, which allows brick-and-mortar customers to purchase items without having to go through a cashier system, is gaining momentum fast.
“Frictionless checkout technology is giving people the opportunity to stay in their bubble and not have to interact with others. They still have the opportunity to ask an employee for help if needed, but they do not have to interact face-to-face,” said Michael Jaszczyk, CEO of software company GK Software USA.
E-commerce giant Amazon is credited as being the first to bring frictionless checkout to the U.S. retail market with the January 2018 opening of its Amazon Go store in Seattle. The store’s “Just Walkout Technology” enables customers to grab items off the shelves and walk out without checking out. Customers merely need to scan their smartphone using the Amazon Go app upon entry, and are charged for the goods picked up during their shopping trip upon exiting.
Frictionless checkout recently hit the convenience channel when Ricker’s announced that it will roll out Skip, a hybrid of mobile pay and the Amazon Go concept, to its 58 c-stores throughout Indiana. Skip's technology reduces the speed of checkout from an average of 60 seconds to an instant pay-and-go action controlled by the customer.
The advent of true autonomous checkout technology is still in its early innings, according to Michael Suswal, co-founder and chief operating officer of Standard Cognition, a startup with its own artificial intelligence (AI) platform that allows buyers to grab what they want without having to go to a cashier.
“Amazon’s store in Seattle does a terrific job of showing what that experience will be like. That said, Amazon’s technology is not appropriate for other retailers and their varying retail environments,” he said. “Amazon has created a new store around technology it has developed, not technology for existing retail stores.”
Tracey Wiedmeyer, chief technology officer for InContext Solutions, a provider of enterprise mixed and virtual reality solutions focused on retail optimization and shopper engagement, agrees that frictionless checkout is a really new technology and it’s too early to say whose approach is best.
“Clearly, Amazon has been the leader here and since they have such a huge following, there’s a lot of trust in their decisions, so people are always excited to see what they’re going to do next. And there are others, like Walmart or Kroger, who are hot on their heels,” Wiedmeyer said. “…But while Amazon Go and some other players in the space are beta testing right now, there’s not been enough data or any true examples to compare yet.”
Requisites for SuccesS
Whatever the approach, Jaszczyk advises that ease of use is the foremost requisite for success.
“I heard of one grocer who implemented a mobile self-scanning solution and the consumer had to put weight-based items on scales in the store, type in a three-digit PLU code, print a sticker, put it on a plastic bag, and then scan it with the mobile app. This is not ease of use,” he said. “In an ideal world, the scale would use image recognition to identify the product and then use NFC to transfer the information to the self-scanning device.”
Those who leverage cutting-edge technologies to “train their systems” are going to be the fastest and most efficient, according to Wiedmeyer. He explained that any sophisticated machine-learning algorithm needs loads of training data to recognize real-world objects and relevant contextual actions with any degree of confidence and accuracy. And when applied to cashierless checkout scenarios, the problem set multiplies rapidly.
“You have to think about the thousands of SKUs in the store, the way they are placed, what it looks like when someone picks it up vs. puts it in their cart, and then what people look like and how they behave. It’s incredibly intricate,” Wiedmeyer noted.
One way to train algorithms is with video feeds or static images of real-life objects. A better way to do so, he believes, is with virtual simulations — that is, 3D digital versions of products and store environments that allow these computers to rapidly run millions of different scenarios.
Standard Cognition’s Suswal has a list of five must-dos he says are requisites to making any approach to autonomous checkout a viable market solution:
1. Privacy: Shoppers deserve a solution that serves their needs without compromising their privacy.
2. Scalability: Any autonomous checkout solution must be scalable without requiring long and complicated installations, including potentially shutting down the store.
3. Experience: Autonomous checkout must allow retailers to delight customers. Two of the biggest complaints most U.S. retailers receive from customers are long wait times in line and poor customer service.
4. Flexibility: Autonomous checkout should give retailers the ability to amplify their brand with their customers. The fewer hardware types used in autonomous checkout, the more flexible a store can be with its merchandising and brand experience. Payment flexibility is also important. A viable solution must accept cash, credit and other popular payment methods.
5. Insights: Retail analytics are valuable, but they must be seen through the lens of privacy. A lot of information can be gathered by retailers, but should have no connection to any specific individual.
Download our full report, “Frictionless Checkout Gains Traction,” by clicking below.