The Kraft Heinz Company is best known for iconic global brands like Heinz Ketchup and Kraft Macaroni & Cheese. It’s also a leader in digital technology, having built an advanced cloud data platform that has allowed it to increase its pace of innovation, react faster to changes in the marketplace, and further its purpose to “make life delicious.”
In 2020, Kraft Heinz moved a largely on-premises data infrastructure based on Hadoop to the cloud as part of a wider digital transformation initiative to boost innovation and become even more responsive to customer needs. “The end goal is to create an agile, consumer-driven culture within the company,” says Mani Gopalakrishnan, vice president of digital transformation at Kraft Heinz.
The company evaluated new data platforms against three main criteria: scalability, agility/speed/performance, and being cloud agnostic. It also needed a platform it could migrate to quickly, without interrupting its ongoing development projects.
It selected Snowflake because the company’s Data Cloud met those criteria and for another important reason: The Snowflake Data Cloud makes it easy for users to access and consume third-party data sources, which Gopalakrishnan says has been highly beneficial for Kraft Heinz’s data science teams.
Despite the pandemic’s onset, Kraft Heinz decommissioned its on-premises data warehouses in just nine months, moving half a trillion records into the Snowflake platform running on Microsoft Azure. The company now has a single, unified data hub that drives its day-to-day operations globally and is also scalable and flexible enough to support new digital transformation projects.
A big benefit of moving to the Snowflake Data Cloud has been easy access to external data sets that are hosted on Snowflake’s platform. Kraft Heinz has been using Covid-19 data from Johns Hopkins University, allowing it to see instantly which areas of its business are most impacted. It uses this data to build predictive models that allow it to get products into the supply chain exactly where (and when) they’re needed, ensuring its partners and customers can keep supermarket shelves stocked.
New data sets must typically go through extensive preparation and testing before they are ready to use, which impedes agility and experimentation. With the Snowflake Data Marketplace, third-party data sets are already available on the platform and can be turned on in a few minutes. They are then available for use alongside Kraft Heinz’s own customer relationship management (CRM), inventory, and other data, making it easy for data teams to experiment and build new applications. That means the company can focus its resources on innovation instead of the mechanics of data plumbing.
The data teams at Kraft Heinz work alongside business function leads to ensure that they’re identifying the right areas for optimization and that new algorithms perform effectively before they’re rolled out. In 2021, the company’s data scientists experimented with several data models to calculate optimal “safety stocks”—the amount of inventory needed on hand to ensure orders can always be met. They tested those models on a particular stock-keeping unit (SKU) to demonstrate that they worked effectively.
“This example is near and dear to me, because we were able to achieve it in eight to 10 weeks,” Gopalakrishnan says.
This has been more than just a technology transformation. Kraft Heinz is building a more agile and data-driven culture across the company, and that means paying close attention to skills and culture as well. As part of this effort, Kraft Heinz is actively recruiting data scientists, data analysts, and engineers, as well as user experience (UX) designers and cloud-stack developers. It has put in place several enterprise-wide learning programs to foster a consumer-first mindset.
Looking ahead, the ability to share and exchange data among different organizations will become even more important, Gopalakrishnan says. He predicts a time when all businesses are linked in a data network where they can seamlessly share any data they need to, securely and with respect for privacy.
“Data is going from being something that’s guarded in a company to being a robust platform to being a service,” he says. “Having information is powerful, for sure, but sharing information can be even more powerful.”
To learn more about how Snowflake can help retail and consumer packaged goods (CPG) businesses, click here.