Amazon Go's got the right mix to reverse self-service's disappointments
First installed in 1992 at a Price Chopper Supermarket in New York, the self-checkout concept is certainly not new.
However, system and human errors have led to disappointment and a fundamental “ask” from customers to do all of the work has led to a lack of overall acceptance.
Amazon wants to change this with Amazon Go, blending technology, machine learning, customer adoption of system-essential applications, and a vast customer knowledge and earned trust.
First announced in late 2016, the first Amazon Go, in Seattle, is welcoming all shoppers, with additional locations rumored for 2018.
The timing is right, both for the technology and for Amazon, which has already demonstrated excellence on required customer journey points like a widely adopted Amazon app; retailer and customer trust, which especially important given perceived privacy breaches affecting companies like Facebook and Cambridge Analytica; fair and clear pricing; historical customer data (online search browsing, wish lists, orders, returns and reviews); a customer-embraced one-click checkout; and integrated and frictionless payment (potentially to be expanded by an Amazon consumer banking offering with a transaction fee saving debit card?)
However, the real glue to Amazon Go is its use of AI, cameras and supporting technology to truly enhance the user experience, essentially asking customers to exert nothing more than “grab and go."
As with all new use of technology there is an adoption curve: Kinks need to be ironed out, enhancements need to be made, and broader adoption by both consumers and retailers need to be had. And this is happening rapidly.
Beyond Amazon Go we find companies like AiFi retrofitting existing CCTV to underpin no-checkout stores; MediaMark Austria through its Saturn Express mobile app and the U.K.’s Co-op testing “scan barcode and go” technology; and Faimdata using existing CCTV to facially recognize and then track customers’ physical journeys.
The views and opinions expressed in this article are those of the author and do not necessarily reflect the opinions, position or policy of Berkeley Research Group LLC or its other employees and affiliates.