Starbucks is not simply a business that sells huge amounts of beverages worldwide. The brand also collects vast amounts of data from more than 100 million transactions per week. How did they use this data? And how do AI and IoT participate in this strategy?
The way Starbucks uses data and modern technology for its competitive advantage is an example for all businesses, regardless of size. For example, the brand is a pioneer in combining loyalty systems, payment cards and mobile applications. But that is only superficial.
This article highlights the 5 most interesting examples of how Starbucks uses data, AI and IoT for its competitive advantage. There may be a compelling argument that Starbucks is no longer a coffee business but is now a data technology company in the food and beverage industry.
Starbucks has no shortage of data. They have more than 30,000 stores worldwide and complete nearly 100 million transactions per week. This gives the brand a holistic view of what customers consume and enjoy. But perhaps surprisingly, they’ve only really focused on the value of this data for more than a decade.
This proves that it did not use the data in the first place. But, like so many big changes in a company, a crisis has led to change. It was the 2008 financial crisis that caused Starbucks to close some stores. The lesson of Howard Schultz, CEO at the time was: “Starbucks’ use of data requires more analysis, especially in deciding where to shop.”
Earlier, like many other companies, Starbucks decisions were made by people based on experience and judgment. Data is clearly important, but not systematically. There are very few people writing about it, but it seems that is the usual approach of using data to validate and inform human ideas and decisions.
What Starbucks does is test all sorts of new ideas using data and technology, then use more data to find ideas that will continue. As with real estate, of course, the way Starbucks uses data today extends to a range of marketing and product activities. This shows the intelligence of the brand’s supply chain management. A core part of this is the Starbucks Awards loyalty program, which also began in 2008.
What’s less common is the way Starbucks uses data to make the Internet of things, especially store operations, start with a coffee machine and is now expanding to in-door devices. Other goods like ovens. Five examples of how Starbucks uses data, AI. and IoT to create competitive advantage:
Among a lot of great things, I chose 5 highlights. I chose these because they show that good data usage improves Starbucks’ business, alongside technologies like AI, IoT and the cloud:
# 1: Personalized ads
The use of classic customer data is personalized preferences based on consumer preferences and Starbucks is no different. With over 16 million members in the US alone, the loyalty program accounts for nearly half of all store transactions in the United States.
Knowing customers’ order preferences and favorite products allows Starbucks to send personalized recommendations more precisely. Using AI to identify such campaigns is becoming a standard application of artificial intelligence, and Starbucks has been doing this since 2017 with its “Digital Flywheel” program.
An important focus of this is to propose new products that consumers may like, based on what they have already bought.
But this is not just an individual advertisement. A large portion is still offering regular mass campaigns, but directly affects each consumer in the target segment. These can include cold drinks on hot days, product launches, or seasonal menus.
# 2: Develop products that suit your tastes
Personalized promotions are certainly effective, but equally important for Starbucks is to use customer data in their product development.
Starbucks uses data arising from the buying habits of large numbers of consumers. Details from this data suggest variations and developments from existing products. For example, there was a cute idea more than 15 years ago that gave birth to a pumpkin-flavored drink during Halloween. This has started a series of pumpkin-inspired products around the globe.
A second type is the use of data on channels. The most typical example of this is probably the company promoting the introduction of coffee into the home space in 2016. This is the launch of the product in a supermarket, for customers who like to make coffee at home. Store data gives Starbucks a solid basis for deciding which products to target home drinkers. They can even test take home products like instant coffee in regular stores.
# 3: Analyze the store location thoroughly
Planning where to open a Starbucks store is a complex piece of data analysis. AI supports location economic factors for a store planning model. These include population, income levels, traffic, the presence of competitors, etc.It uses this data to forecast revenue, profits and other aspects of economic efficiency.
The system also looks at the location of existing Starbucks stores. It considers the impact of a new proposed store on existing sales in nearby areas. The main AI technology of this application is location-based analysis.
# 4: Change menu flexibility
Another point of the examples above is that Starbucks has the ability to continually adjust its products. The way Starbucks uses data makes it possible to modify based on customers, location and time. This affects products, promotions and prices.
However, if at the store you visit, there are still pre-printed menus behind the counter, the ability to adjust this is no longer available. This is one reason why imperfect solutions such as blackboards are still popular with retailers. But for Starbucks, the answer is the appearance of digital signage in stores, with menu screens set up by computers.
This completes a chain that allows changes that can occur anywhere in the customer experience to be reflected in the store.
There are obviously lots of questions, and there are lots of areas to complicate things. However, since mid-2018, Starbucks has been experimenting with this in a few stores. They have focused their efforts on pushing selected products based on local circumstances such as weather or time of day.
# 5: Optimization of machines
Our last example is the maintenance of coffee makers and general store machinery.
Trade at typical Starbucks stores with relatively low cost and short time. The larger the number of customer transactions shows the success of a store. Therefore, if a machine is broken, it can significantly disrupt business performance.
Starbucks did not arrange the engineers on duty for the incidents. Instead, they send them to the processing center to deal with the repairs and perform scheduled maintenance. So bringing engineers to broken machines quickly makes a difference.
There are common approaches to this problem. This means collecting data about errors, usage as well as repair requests, etc.Data analysis is often good at finding trends. AI can help forecast incidents and maintenance needs.
The case of Starbucks has taken a new step with the Clover X coffee machine. This machine is currently only used in certain stores. Along with the ability to make superior coffee, it also connects to the “cloud”. This not only allows for more comprehensive operational data collection. It also allows remote diagnosis of errors and also remote repair.
The same concepts apply to other machines. For example, stores now have a standard oven, also controlled by computers, to prepare products globally. However, current machines need to be updated with the USB drive. This happens whenever a change in machine configuration is required, such as new products. In the future, these machines will surely connect to the cloud directly, while creating more opportunities for AI to analyze data more effectively.