Canadian Vending

Features Innovation Trends
Cashless in Seattle

How Amazon is venturing into vending

September 29, 2017
By Michelle Brisebois


Amazon gobsmacked the retail sector this year when it piloted small grocery stores without cashiers or checkouts. Amazon’s Seattle-based test store used computer vision, deep learning algorithms and sensor fusion, like that found in self-driving cars. Amazon branded it “Just Walk Out” technology.

“Once you’ve got everything you want, you can just go. When you leave, our “Just Walk Out” technology adds up your virtual cart and charges your Amazon account. Your receipt is sent straight to the app,” promotes a video by Amazon. The media went wild; the retail sector went wild. A world without cashiers where you serve yourself! Who would have thunk it?  Vending, that’s who.

Just file this one under the heading of “what’s old is new again”, but what Amazon has done is to create a micro market with a very robust technological underpinning. The company has figured out how to apply digital measurement to the physical retail space. The operative phrase is “computer vision, deep learning algorithms and sensor fusion.” Let’s unpack that. Computer vision gives technology the ability to mimic what the human eye can do. For the Amazon Go model, that means facial recognition so that when you walk into the store, the app knows it’s you and that your account will be charged as you start shopping. Sensor fusion monitors weights and motion so that as an item is removed from the shelf and placed in the basket it knows to charge the shopper’s account. If the item is removed from the basket and placed back on the shelf, Sensor Fusion knows to refund the item to the customer. Deep learning allows the computers to continuously learn and adjust course by analyzing all of the data collected. As the software learns, it will continuously improve the shopping experience based on customer behaviour. It’s this final area where Amazon truly shines and where so many other industries and companies fall flat. Everyone is swimming in data that they don’t understand or use to improve. Companies also fail to look at their businesses holistically.

Everybody thinks Amazon’s online retailing is the dragon consuming bricks and mortar purchases. Actually, e-commerce in Canada still represents less than 10 per cent of all purchases (CIRA March 2016) and consumers still value going into physical stores. Amazon’s e-retailing isn’t where they make their money. They make their money with the Amazon Prime subscriptions and with their cloud computing arm. Amazon no doubt sees connections between physical retailing and these cash cows. The stores deliver the volume of customers where offline and online worlds are blended. That Amazon Prime subscription will now give customers access to better prices on groceries, subscriptions to meal deliveries that in turn keep the customer very “sticky” to the brand. While very few companies can boast the infrastructure Amazon has to mine, analyze and actualize the customer data they’ve collected, stealing a page from their play book would still be prudent. Make the time to shake hands with your data.


Vending solutions that take electronic payment have a lot more data than many other physical retailing outlets do. Are you looking at transactional trends:  time of day, average sale, peak times, popular products? These are baseline metrics and you should know what they are. According to a Price Waterhouse Coopers and Iron Mountain survey in 2015; only four per cent of businesses are set up for true success in terms of exploiting their data. The survey reports that “43 per cent of companies obtain little tangible benefit from their information, while 23 per cent derive no benefit whatsoever.” Human resources focused on data analysis, solid tools to harvest and summarize the data, a culture defined by evidence-based decision making and open mindedness – these form some of the key pillars for using data effectively. Marrying the behavioural data with the attitudinal research and insight is where the business breakthrough will come. Research need not be expensive. Tweet questions to your followers about products you’re considering or incorporate product reviews into your experience somehow.  

Any food delivery channel needs to start bringing their A game because the rules of engagement are shifting at warp speed. Know the numbers and use them to make your customers’ lives easier and to tell great stories about your business in the process.

Michelle Brisebois is a marketing consultant specializing in digital content strategy and retail/ in-store activation. Michelle has worked in the food, pharmaceutical, financial services and wine industries.  She can be reached at