We Start with Amazon
Of course, everyone knows that Amazon has almost single handedly caused the demise of the retail sector. Whether it be free shipping, Amazon Prime or same day delivery, Amazon has improved the experience for everyone. There is now a “Death by Amazon” index, where you can track how 30 different companies, which all compete with Amazon, are doing. But it’s no coincidence that the largest player is ecommerce is also the leader in artificial intelligence (AI) and Amazon is certainly deploying AI everywhere.
How did they get so good?
Amazon has such an advantage because of the vast amount of data it collects. Each time you come to the site, browse products, put them in your cart or leave them in your cart, Amazon collects all of this data and uses it to drive more conversions. For example, it’s “recommender” technology, built on machine learning, is only possible because Amazon can create “look-a-likes” from its own customer base. When you compare this to the amount of data retailers collect, it’s not even close. Retailers are lucky enough to know if you even went to their store, as opposed to interacting with their products, picking them up, putting them back, or even if you were on the site beforehand.
Retailers Bridge the Gap
How do retailers even the data collection game? Retailers are now turning to AI to help them. Several startups are using cameras with image recognition and installing them into stores to generate data like how many people are coming in, what are the demographics of the shoppers and how people are interacting with products. One interesting theory is retailers could record how people are interacting with products, what their facial reactions are when they pick it up, how long they spend with an item, and with this data, retailers could drive more conversions.
Retailers could also use a combination of sensors and cameras to develop real time inventory to help better manage purchases. By using image recognition and RFID tags, retailers could maintain a real-time system of what it’s inventory is. Goldman Sachs estimates retailers lose $253 bn a year from over/under stocking items, a problem that could be reduced by this new technology.
AI Enhances the “Experience”
Retailers will certainly have to rethink their store experience and AI can play a large role in this. One company call Oak Labs has developed technology including a “smart mirror”, which utilizes RFID tags and recommender algorithms to drive engagement. The system works as once someone brings a piece of clothing into a dressing room, RFID data is collected so retailers get an idea of which products are being tried on. Then, as someone tries the product on, the mirror can give recommendations on other styles and even offer payment, speeding up the customer experience. Nike is also innovating with their new Soho store, which once customers enter the store, if they choose to self-identify themselves through the app, the company can enhance their experience upon their return.
But these retailers may be too late
While some of this technology is being tested, Amazon is heading full on by beta launching its fully autonomous store, Amazon Go. The introduction video has been watched almost 10mn times on YouTube, and reflects a truly autonomous store, with almost no employees. No retailers have come even close to matching this.
Amazon Deploys AI Everywhere
The reality is, AI is at the heart of everything Amazon does. While Alexa is a cool device that helps you at home, it also allows people to order products through Amazon with a simple voice command and sometimes even gives discounts if it is ordered through an Alexa .It even gives you fashion advice by taking a picture of what you are wearing. Even five years ago, Amazon patented “anticipatory shipping”, which would try and ship you a package, before you even bought it. It’s also no surprise that Amazon has started hiring people to work on autonomous driving, as billions of dollars could be saved by reducing shipping costs.
Finally, while just this week Amazon announced an initiative to hire 50,000 new workers, this should be met with a healthy dose of skepticism. Amazon already employees people with deep learning experience to build algorithms that recognize shop lifting in its warehouses, and the next logical step would be having these algorithms learn how employees move in their warehouses and use that knowledge to automate them with robots.
Amazon started to kill retail with AI, and it looks like it’s trying to finish it with AI