One of the strongest examples of an early pioneer in data mesh, a highly decentralized data architecture, is not a telecommunications firm, large-scale manufacturing business or even a financial industry giant. It’s HelloFresh SE, a German publicly traded meal kit company.

The data team at HelloFresh found itself in a whirlwind of growth as the company rapidly scaled its business over the past decade on its way to becoming number one in the world in its field with 239 million meals delivered across 14 countries in the previous quarter. That growth required the team to reimagine the company’s data architecture from a monolithic focus to decentralized models, which would position the expanding company for the digital future.

“In 2019, more people realized that this model doesn’t really scale, and the leadership of the company came together and identified data as a key strategic asset,” said Christoph Sawade (pictured, right), global senior director of data at HelloFresh. “If we could leverage data in a proper way, it gives us a unique competitive advantage which could help support and fully automate our decision-making process across the entire value chain. HelloFresh is now able to build data products that have a purpose.”

Sawade spoke with Dave Vellante, host of theCUBE, SiliconANGLE Media’s livestreaming studio, during the AWS Startup Showcase: New Breakthroughs in DevOps, Analytics, and Cloud Management Tools event. He was joined in the interview by Clemence Chee (pictured, left), global senior director of data at HelloFresh, and they discussed the business issues that led the company to reevaluate its data structure, a reorganization that changed governance and ownership, and the need to foster acceptance among data users. (* Disclosure below.)

Issues with technical debt

HelloFresh built its original system using a Hadoop-based, on-premises model. However, when HelloFresh was going into hypergrowth, its data team realized that it needed to reevaluate how it thought about data. There was already plenty of evidence to support why this was becoming a priority.

The company had a centralized data function that was a huge asset as HelloFresh was beginning to scale, but it had begun to drag the business down following its IPO in 2017.

“In the business domains like marketing, supply chain, finance and HR, analytics teams started to build their own data solutions,” Chee recalled. “The data pipelines didn’t reach the engineering standards, and there was an increased need for maintenance and support from the central teams. Most of those datasets turned into a huge debt, with decreasing data quality, trust and transparency. A majority of time was spent in meeting rooms to align on data quality.”

HelloFresh embarked on a phased reorganization to build data products with a purpose, creating trustworthy assets for data users. A key element of this involved governance coupled with a focus on standards.

“We realized that we needed to focus on helping our teams develop those governance capabilities and teach the standards for how work is being done to truly drive functional excellence in the different domains,” Chee said. “We’re teaching ownership and responsibility to our colleagues via individual learning paths and helping them upskill to use shared infrastructure and self-service data applications.”

A gamified experience

Teaching those learning paths also required buy-in by HelloFresh data stakeholders, so the data team incorporated forms of gamification to facilitate motivation and progress.

“They can earn badges along the way, which simplifies the process of learning and engagement of the users,” Chee noted. “What we see in surveys is that employees value this gamification approach a lot and are even competing to collect learning path badges to become number one on the leaderboard.”

What HelloFresh learned in the process was that the journey toward a data mesh required a new way of approaching data management and infrastructure. A data mesh decentralizes ownership and turns data assets into products that become distributed companywide.

“Data mesh is an architecture of scale,” Sawade said. “It’s necessary for huge companies who want to build data products at scale. Make sure you take people with you on your journey.”

Watch the complete video interview below, and be sure to check out more of SiliconANGLE’s and theCUBE’s coverage of the AWS Startup Showcase: New Breakthroughs in DevOps, Analytics, and Cloud Management Tools event. (* Disclosure: This event was sponsored by participating companies. Neither Amazon Web Services Inc. nor sponsors have editorial control over content on theCUBE or SiliconANGLE.)

Photo: SiliconANGLE