The development of artificial intelligence skills is one of the main pillars of Starbucks' digital strategy and vision of the future. Kevin Johnson, CEO since 2017, also says that they aspire to be a reference in the subject, to the point where they can compare their own performance to that of the technology giants.
Unlike McDonalds, which has been making acquisitions, Starbucks chose to develop these skills internally, and in partnership with Microsoft. The company's AI (Artificial Intelligence) initiative was formed in 2017, and is called "Deep Brew". According to Johnson, the company's vision has allowed Starbucks to recruit some of the best talent, attracting candidates who, in the recent past, preferred to join the big tech companies.
The main objectives are to improve the consumer experience through personalization, in addition to “freeing up” the workforce from administrative and backend tasks, so that they can focus on interacting with customers. According to Johnson, “It's not about robots replacing baristas. It's a technology that frees up baristas' time so they can better connect with the consumer.”
The technology is currently used in the US for workforce sizing (predicting how many people are needed per day at each branch), and also for automating store replenishment orders so that the right items are available at the right locations. right branches.
With regard to the customer experience, technology is used to personalize offers and recommendations in the company's app, expanding to digital menus in restaurants and drive-thrus. The system was built interconnected with store inventory data, in order to minimize consumer frustration with out-of-stock items.
The recommendation engine makes use of “regional” data (popular items in the location, weather, time of day, day of the week, if it's a holiday or not, for example), and is improved as the application is used. Through the analysis of purchased items and the user's reaction to previous offers, the algorithm can “understand” that a customer is vegetarian, or that he has a preference for lactose-free products, for example. In addition to considering the ingredients and product categories purchased, the consumer's sensitivity to higher or lower price items is also evaluated.
Through reinforcement learning techniques, the more transactions the user performs in the application, the closer the recommendations are to their preferences, both with regard to ingredients and product categories as well as price. For more details, check out the presentation made by the company at the Microsoft Build event in 2019.
In drive-thrus and restaurants, the technology does not rely on the individual history of users as in the application. However, according to Microsoft, 400 criteria will be used at each branch level to generate recommendations.
The company has also invested in IoT (internet of things) devices in its branches. Each Starbucks store has at least a dozen pieces of equipment (including coffee machines, grinders and blenders), which must be in operational condition for about 16 hours a day. Failures in any of these devices imply not only increased costs, but can also harm the consumer experience.
In partnership with Microsoft, the company turned its equipment into IoT devices, which collect data from the type of coffee beans used, to the temperature of the beverage, water quality and equipment pressure. The company has been using the data to ensure product standardization across thousands of stores, as well as using it to create predictive equipment maintenance plans.
Another use found by the company was the distribution of coffee recipes directly to the machines, which was previously done manually with USB sticks for each branch (for security purposes). “Think of the complexity – we need to reach 30.000 stores in nearly 80 markets to update those revenues,” says Jeff Wile, senior VP of retail and core technology services.
Starbucks is a company that does not shy away when it comes to technology, and the fact that they have managed to recruit professionals and develop skills internally is a great differentiator, especially when the workforce of artificial intelligence professionals is scarce and competition is fierce.
References:
Mobile Marketer – How Starbucks uses AI to counter mobile's isolating effect. https://www.mobilemarketer.com/news/how-starbucks-uses-ai-to-counter-mobiles-isolating-effect/570384/
QSR Magazine – Why AI is a 'Differentiator for the Future' at Starbucks. https://www.qsrmagazine.com/fast-food/why-ai-differentiator-future-starbucks
Big Data Beard – I'll take my Starbucks with a shot of AI. https://bigdatabeard.com/ill-take-my-starbucks-with-a-shot-of-ai/
Microsoft – Starbucks turns to technology to brew up a more personal connection with its customers. https://news.microsoft.com/transform/starbucks-turns-to-technology-to-brew-up-a-more-personal-connection-with-its-customers/
CIO Dive – Starbucks' AI project is turning past investments into data insights. https://www.ciodive.com/news/starbucks-ai-project-is-turning-past-investments-into-data-insights/566362/
Thatiana Nomi
5 years of experience in consulting projects in Logistics and Supply Chain. Worked on logistics network optimization projects, service cost, logistics master plan, inventory policy definition, sales forecast, S&OP (Sales and Operations Planning) and transportation planning and management.
