Anheuser-Busch InBev, the $60 billion brewer of Budweiser, Corona and more than 500 beer brands, had a data problem when Harinder Singh joined the company as global director of data strategy and solution architecture in 2017.
Thanks to acquisitions of more than a dozen brands over the past several years, AB InBev had accumulated a morass of data stored in on-premises and cloud systems in more than 100 countries. Singh’s objective? Aggregating and unifying the data and making it available to business users through a single lens.
“My peers here tell me that even as recently as three years ago, thinking of technology or data wasn’t on top of mind,” says Singh, who held a similar role at Walmart eCommerce before joining AB InBev. “Business transformation must be enabled by digital transformation and data is at the core.”
Data on tap
Treating business data as the new oil, enterprises are ponying up big bucks for software that can clean and organize data for cultivating business insights, with worldwide revenues for big data and business analytics software expected to reach $260 billion in 2022, according to market research from IDC.
But if data is the new oil, integrating it is the equivalent of extracting it from the ground, putting it into a digital barrel and making it ready for consumption. The problem? Data is growing increasingly fragmented across organizations, particularly as legacy applications are replaced with new, loosely-coupled applications.
AB InBev’s data was lodged in more than 100 source systems, including Salesforce.com, 15 SAP systems and 27 ERP systems, and the company relied on 23 separate ETL (extract, transform and load) tools to move data from one database to the next.
This “Band-Aid” approach made it difficult to obtain a single view of the data assets, Singh says. And with GDPR coming online, AB InBev needed to have global visibility into its data, some of which is about consumers and therefore subject to various privacy laws.
“We still need to standardize and integrate that data — another aspect of our data challenge,” Singh says.
Singh conducted a months-long proof-of-concept with Talend, a provider of data integration software in the cloud, before selecting the vendor. AB InBev uses Talend — essentially a cloud-based ETL tool for the modern era — to extract data from a range of sources, including cloud and on-premises systems and data from IoT devices, and store it in a Hortonworks Hadoop data lake hosted on Microsoft Azure.
That data is then processed and archived before Talend shuttles it to a “golden layer,” which data scientists, operations staff and business users can access via data visualization tools. AB InBev’s reusable data management architecture also includes open source tools such as Hive, Spark, HBase and Kafka, Singh says.
The analytics of beer
Whereas AB InBev staffers once spent 70 to 80 percent of their time locating relevant data across various systems, they now pull information for analyses from a single source. The work is ongoing, but Singh is certain the data platform is positioning AB InBev’s staff to glean more critical insights about sales, supply chain management, marketing, human resources and other business lines.
For example, AB InBev collects and aggregates consumer data from Nielsen and market surveys, and near-real-time data from social media, to analyze trends and deliver the right beers and more highly targeted marketing campaigns, including real-time coupons tailored to consumers at the point of purchase. AB InBev can also identify the best location in the store to sell beers, as well as how to create real-time events to drive more conversion.
The company also uses analytics to optimize its supply chain. IoT data from RFID devices help track “connected packages” to find the best routes for delivery drivers, and temperatures in millions of beer coolers around the world are monitored to ensure AB InBev’s products are being stored and served at the optimum temperature.
Singh allows that AB InBev is still working through technical debt that includes outdated processes and technologies, but he attributes AB InBev’s successful data transformation to three factors: a “cloud first” approach; using data as a business insight; and the development of reusable processes for rapidly extracting and providing access to data.
“At the end of the day, it’s about bringing the best beer and product to the consumers,” Singh says. “Using data has been critical part of our success in the journey.”