Amazon Go: a bluff or a real revolution in commerce?

The promotional video is dreamy. The setting: a small supermarket like there is on every street corner in our cities. The customer enters, badges his smartphone, chooses the products he wants from the shelves, puts them in his shopping bag and leaves the store. No checkout, no line, no salesmen. It's as simple as that. His shopping basket is automatically analyzed by a myriad of electronic eyes, deep learning algorithms grind it all together and it's done. His account is automatically debited. A revolution. A revolution, really? After revolutionizing e-commerce, would Amazon want to revolutionize physical commerce? Decryption and prospects.

 

Ahe first Amazon Go store opened in Seattle; it is believed to be Business InsiderThis is the first of a network of 2,000 sales outlets that are being progressively installed. Amazon is not the first to come along. It's one of the founding fathers of e-commerce, first focusing on cultural goods and then on a catalogue of more than 500,000 products of all kinds. You can buy everything on Amazon, or almost everything. Over the years, the iconic brand has improved its knowledge of the intricacies of e-commerce to the extreme; it has relied more than rightly on logistics, the key to success and confidence in the field. It has invented online marketing, how to get to know the consumer and the best way to sell him the product he is looking for or simply coveted. Amazon aggregates a considerable marketplace. In short, a web giant that hundreds of millions of customers regularly use on all continents.

So, with its position on the web, what does Amazon do in the physical retail trade, the neighbourhood shop, in "brick and mortar" as they used to say in the early days of the web? Is it a step backwards, a back-pedal? A nostalgic return to the roots? Not quite. Amazon knows that consumers, although they use the web more and more for shopping, are still numerous, attached, at least as far as food is concerned, to physical commerce, and go shopping in stores near their homes or on the outskirts of cities. Hypermarkets have certainly reinvented their trade, but they are still a machine for producing a turnover that can be counted in billions of euros for a single store. Amazon knows this perfectly well and carefully observes that food represents only 5 % of its turnover. On the other hand, what it has also detected is one of the major drawbacks of this traditional trade: you have to queue at the checkouts and wait, sometimes for long minutes, to pay. As with Amazon, there is a solution to every problem, so why not get rid of this constraint? Advances in visual recognition, image flow analysis, sensory sensors, algorithms... make all dreams possible today. So it's only a short step from there to imagining a supermarket without checkouts, where the customer could help himself and leave in complete freedom.

 

The Amazon Go store in Seattle is a prototype intended first to test the concept with Amazon employees. A small area of 170 m2 installed on the ground floor of the company's headquarters. If all goes well, Amazon promises that a network of points of sale will be established, everywhere. But we're not there yet. There are many difficulties that need to be solved first, not least of all.

 

 

First of all, there are technical questions. Amazon is very stingy about explaining the technology used. They talk about visual recognition, algorithms, deep learning, sensory sensors, so many swear words that don't mean much as they embrace a heterogeneous field of realities. We are reduced to conjecture. To identify the product that the housewife is going to put in her shopping basket, we can certainly use visual recognition techniques. These have made enormous progress in identifying complex shapes. But from there to working in a real situation, with a crowd of consumers helping themselves on the shelves, on a sales area of sometimes more than 10,000 m2, there is a margin of technical difficulties that engineers have not yet been able to overcome with flawless efficiency. However, mass distribution is not short of innovations. Most of the major retailers are already practising the "innovation" approach. self scanning ", are adopting the mobile cash registers, as is the case at Sephora or Apple Stores. But considering a system of "just walk out technology As Amazon claims, "it's a challenge when it comes to outlets that carry as many and as diverse items as you would find in a large grocery store. However, let's assume that Amazon will be able to meet this challenge in the near future. So what is the point of this announcement with its technically premature "wow" effects, and above all, what is the strategic stake of such an innovation?

By announcing its concept, Amazon takes all the actors of the mass distribution sector by surprise. They have invested heavily in the last few years in online commerce, they invented the drive but have not managed to profoundly change the concept of the supermarket. Amazon tells them that it is possible and announces with fanfare the emergence of a disruptive model: a model that is neither totally physical nor totally dematerialized. A physical model in which technology is predominant. A model where the customer no longer waits, no longer even realizes that he is buying. A model where the purchase is instantaneous, irreversible and totally painless. A model where the slightest buying behavior, the slightest hesitation is analyzed, scrutinized and dissected by the merchant who has become panoptic. A model where humans, those damned humans who cost a lot, who grumble and sometimes go on strike, can be eliminated; no need for cashiers in Amazon supermarkets.

 

By launching Amazon Go, the Seattle-based firm is kicking the anthill and shuffling the cards by allowing everything possible. The company is drawing on its experience in automation and on the groundswell of robotics and artificial intelligence. Amazon has made this a speciality with its logistics centres. The business magazine Quartz notes: "Amazon already uses robots in its warehouses to transport products to humans, who then ship them. They are working on robots that will soon be able to track products and ship them themselves. Reverse this process, and those same robots will theoretically be able to restock the shelves of an Amazon Go. In the near future, such a store could potentially operate almost without humans. »

The Oxford economists had already predicted that 47 % of the jobs could be replaced by robots in the next few years. President Obama, in a lengthy interview at Wired that we had reported last October expressed concern that "people in low-skilled, low-income jobs are becoming increasingly redundant. With this in mind, he called for the implementation of the Universal Basic Income initiative.

 

Amazon Go, as you can see, is not just a commercial type ad. It's the warning shot that foreshadows major upheavals. It must be taken seriously.