Amazon Go: Invisible Interaction Yet Visible Personalization


The IT giant Amazon has been extensively testing its first smart offline store “Amazon Go” in its hometown Seattle, Washington since its exposure at the end of last year (González, 2016). This concept store has gained significant popularity as it claims to have no cashiers or checkout lines thanks to computer vision, deep learning algorithms and sensor fusion.


This indeed opens the door to the “pick up and just walk out” next-generation shopping experience. While the entire process seems to require no intervention, participation or interaction, it is not as simple as what it looks like. A customer will be identified the moment he or she walks into the store by swiping an application. His behaviors before, during and after the process of shopping will be captured and recorded, as well as the final list of groceries he purchases and the bill. For instance, someone may pick up two baguettes at first, but then return one to the shelf and keep the other as it is gluten-free. By implicitly collecting and analyzing all those information, Amazon could learn their preferences and even maybe study the way they think or hesitate when facing several alternatives.


The concept store starts with only groceries, therefore it is currently considered as the offline store of Amazon Fresh, the online grocery service of Amazon, which has been steadily and quickly across the globe in recent year (Deagon, 2017). The trial of Amazon Go could help Amazon Fresh match the recommendations to these testing customers’ preferences, for instance if one product is gluten-free. Then Amazon could see if such upgrades of recommendations can increase the purchase willingness on Amazon Fresh and evaluate the effectiveness of the offline learning process and algorithms. Since Amazon Go seems to plan for ambitious expansions per quantity of both the goods and the stores (Stevens & Safdar, 2016), this learning process will dramatically help to optimize the offline business as well.


Nevertheless, while either Amazon or traditional market chains such as Walmart and Wholefoods would find it hard to walk down this path. First of all, this complicated system in the shop serves as the prerequisite, and it has to cover every corner with sensors and scanners. It would not be surprising if this can cost a pretty penny. Moreover, since it would be a huge workload for computing and might cause delays if such systems works in a wholesale store like Walmart, Amazon Go has to remain tiny or mid-sized, which means customers may have relatively limited shopping choices.


This could be painful for Amazon. The concept store can surely attract tons of customers and learn mountains of their information in reality, but customers can only choose from limited products — while infinitely more can be found available via its familiar website. In other words, Amazon Go may serve as a wonderful input resource and help greatly with the recommendation system, but it can barely contribute an appreciable revenue itself in the short run.


Therefore while Amazon Go is by all means a technological revolution and hopefully will undergo ambitious expansions, it will have to serve as a concept store for a very long period due to the above difficulties. Amazon’s business model will keep focusing on online services — like it always has been. Amazon Go, on the other hand, will mainly help to learn better about the customers’ preferences and thinking process, and eventually increase the efficiency of its online recommendation system, especially for Amazon Fresh.






Deagon, A. (2017) Amazon, Having Conquered E-Tail, Tackles Physical Retail. Investor’s Business Daily. Available from: [Accessed 15th February 2017].


González, A. (2016) Amazon unveils smart convenience store sans checkouts, cashiers. The Seattle Times. Available from: [Accessed 15th February 2017].


Stevens, L. & Safdar, K. (2016) Amazon Working on Several Grocery-Store Formats, Could Open More Than 2,000 Locations. The Wall Street Journal. Available from: [Accessed 15th February 2017].


Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s