At restaurants and movie theaters around the world, consumers encounter a handy way to personalize their soft drinks: Coca-Cola’s high-tech vending machines, which allow people to mix and match their preferred flavors and beverages. Cola fans can create whatever odd or conventional drink combinations they prefer, whether that means mixing orange Fanta with Sprite, adding water to the lemonade to make the flavor less intense, or blending caffeinated and non-caffeinated versions to find a happy middle between getting energy and not being awake all night.

Although the machines thus appear dedicated mainly to keeping customers happy with personalized offerings, they provide another key benefit to Coca-Cola, in that the dispensers also are equipped with advanced artificial intelligence (AI) and data analysis capabilities. By tracking what the millions of users are putting in their cups, Coca-Cola gains unique insights into popular trends and possible new product ideas. It also can define its regional strategy, adjusting the recipes of its various beverages to match local tastes.

Notably, in the approximately 200 countries in which it sells Coke, Sprite, and other brands, Coca-Cola adjusts the recipes to reflect consumers’ preferences, such as adding more sugar to some versions or expanding the no-calorie options available to consumers in another geographic region. The vending machines and the local data they gather are invaluable for these efforts, because they signal what local populations are drinking every time they go to the movies. This combination of big data with localized analyses gives Coca-Cola a powerful advantage in terms of ensuring it is crafting products to meet local preferences.

The variety of mixes that people chose from the start also led Coca-Cola to revise its vending machines, to allow consumers to add a quick “shot” of flavor into conventional beverages. Without the careful analysis of buyer behaviors, the company might have simply assumed that its existing products were meeting all the flavor profile needs of its consumers, but when people started combining totally unexpected flavors (vanilla orange cherry Coke, anyone?), it recognized an unmet need in the market.

Some of those creative combinations are probably unique to small market segments, who are sufficiently served by special treats at the vending machines. But Coca-Cola also uses all these data, across regions and sales environments, to inform its new product development activities. So if you enjoy having access to Cherry Sprite in bottles and cans at your local grocery or convenience store, you can thank millions of consumers who mixed cherry-flavored drinks in their Sprites in the past, leading Coca-Cola to predict it would be a successful product introduction.

Source: Sources: Bernard Marr, “How Coca-Cola Is Using AI to Stay at the Top of the Soft Drinks Market,” Artificial Intelligence News, May 7, 2019; Bernard Marr, “The Amazing Ways Coca-Cola Uses Artificial Intelligence and Big Data to Drive Success,” Forbes, September 18, 2017; Anirudh VK, “Coca-Cola Leans on Data Analytics & AI for Deeper Industry Insights,” Analytics India, June 19, 2019.