Big Data, Artificial Intelligence and Vending Machines: How Coca-Cola Continues to Assert Dominance

Prior to the coronavirus pandemic, the world was largely headed in the direction of automation and robotics. From dating apps, to robotic cooks in kitchens, to mechanised production lines in factories, human interactions were already becoming increasingly less important – but Covid-19 has only propelled us further into the world of automation. In the wake of Covid-19, the new normal’ will involve continued social distancing and using technology instead of humans wherever possible, and so it is only natural to expect that big data applications will come to be increasingly relied upon. As strange as it may be, removing forms of person-to-person contact will be the priority in a post-Covid world.

Big data application on smart vending machines is one particular area where strong growth is expected over coming years, not only because of its impact in terms of efficiency and optimal product replenishment, but because it requires zero human to human contact or interaction – something people will be seeking for some time in the aftermath of this deadly pandemic.

Coca Cola was – not surprisingly, given its scale and capacity for innovation – the first to begin experimenting with artificial intelligence and big data to create smart vending machines that require less human management. The beverage supergiant hired Hivery, a data-driven platform that works with companies to optimize their retail strategies using AI and big data, to revolutionise the way it marketed and sold its products in over 200 countries. Starting with 60-odd vending machines in Newcastle, a town 160km outside of Sydney, the tech startup began experimenting with AI to see whether it could stack vending machines better than humans could to optimise sales and profits. The company developed an AI algorithm specifically for Coca Cola, and that algorithm resulted in an 18 percent reduction in re-stock visits, plus a 15 percent increase in sales, proving that big data really does make a difference when it comes to vending machine optimisation.

Hivery co-founder Jason Hosking said the gamble paid off and today they work with 20 major clients including those in the USA, Japan and China to develop AI-focused solutions to retail challenges.

The race for optimal replenishment or shelf space is the most challenging problem in the global consumer retail industry, costing suppliers and retailers $400 bln in lost sales per year, he said. Vending machines are data-rich and operationally companies have full control of the supply chain. This means you have good datasets to train your algorithms and have the power and control to make changes in the market while at a retail outlet the decision is ultimately controlled by the store owner.

Today, big data is at the forefront of everything Coca Cola does. From choosing the location of its vending machines, to persuading loyal customers to purchase new products, to showcasing the different product options.  This kind of marketing insight is key. Many companies, for example, want expand their marketing reach using big data. There is no doubt that Coca Cola invests a lot of time in this facet. The company has an astonishing market share of drinks sold through vending machines in the world. Their marketing budget exceeding Apple and Microsoft combined, in 2018 the company spent over $4bn dollars in advertising. The data that informs their marketing and business playbooks has been garnered from consumer choices and its clear they’ve invested a lot of time and analysis in specialised schema markup for their products and services. It is easy to think that the behemoth has just one or two drinks in its product catalogue. In actuality, they own over 500 brands worldwide. Each being supported by vendors and served by vending machines.

Artificial intelligence is the foundation for everything we do. We create intelligent experiences. Artificial intelligence is the kernel that powers that experience, said Coca Cola’s global director of digital innovation, Greg Chambers. And with over 2 billion sales transactions a day in over 200 countries, that means a lot. The vast amount of data aggregated by Coca Cola’s retail outlets (including vending machines) around the world requires huge amounts of analysis across disparate sources to determine which products, packages and flavours are received well by consumers. The relationship with buying patterns is complex, and is one of many reasons why Coca Cola uses AI.

But as Covid-19 continues to impact us in bizarre and unexpected ways, and we scramble to figure out how to live in an overly hygienic and contactless society, big data can help vending machines evolve and adapt in other ways too.

By informing a vendor’s product offering, for example helping them recognise the demand for hand sanitiser, vitamins and face masks, or even shifting to provide fresh ready-to-eat meals prepared by robots, big data is upending the vending machine industry. Big data is even helping vendor companies re-organise the routes by which teams travel to maintain, repair or deliver vending machines while roads are cut-off due to restrictions or lockdown, and helping operators optimise the space allocation of products so that they last longer, meaning less visits and less contact with the surfaces of the products.

As cities around the world go into lockdown and shopping centres, townships and retail outlets are closed as a result, vending operators need to quickly shift to cost-saving initiatives if they hope to ensure their own survival over the coming months. With the help of big data and AI, they might be able to achieve exactly that, meaning that they at least might be able to avoid the Covid-19 induced economic fallout that looms for the rest of the business world.