Creating secure and frictionless customer journeys

Overcoming data compliance and reliability headaches

As a Customer Data Platform (CDP), Relay42 gives digital experts complete control over customer data and enables them to create real-time personalized customer journeys on all channels and touch points. But customer data is tricky, especially with more compliance regulations and consumer wariness around data privacy. Relay42 stands out in that it’s compliant by design — meaning it doesn’t store sensitive customer data but instead ensures the data is clean and consistent, and it can help companies adhere to GDPR and other compliance regulations.

That’s why some of the biggest global brands across industries — from FedEx to Air France — rely on Relay42 to help them activate sensitive customer data. But that’s also where the company encountered challenges: “Since the beginning, the biggest challenge for us has always been the volume of data,” said Nikola Yovchev, the Head of Engineering at Relay42. “Whatever our clients want to model — such as every single click from one of their customers on their website — our platform can extract the data for them to see and make decisions accordingly.”

However, as data volumes grew, their data infrastructure began to falter. Relay42 was using a combination of Apache Cassandra and Hadoop to take in all of their customers’ data, which was not only difficult (and costly) to scale but also unstable and slow. Consequently, their infrastructure team spent significant amounts of time simply maintaining infrastructure and data pipelines.

“Not only was it expensive to store terabytes of data, but Hadoop was having trouble ingesting it. It was slow and unstable. Oftentimes, clusters would crash or pipelines would run so slowly,” explained Yovchev.

Unifying data to power more seamless customer experiences

But things turned around once Relay42 switched to Databricks Data Intelligence Platform. With a common view into their data via the lakehouse and with Delta Lake offering data reliability and consistency at scale, Relay42 is now able to quickly manage huge volumes of customer data coming from its clients with confidence. This means data insights get into the hands of their customers faster, which allows their customers to reduce “waste” from their online marketing efforts. Let’s say someone searches for a product or service online. A typical company might spend time and money retargeting the potential customer on Facebook, Google or other sites with ads, even though the consumer may have already purchased directly from the company website. This results in a tremendous waste of budget for something that turned out to be completely unnecessary.

But with Relay42’s CDP — powered by Databricks Data Intelligence Platform — companies can now access all of their customer data in near real-time and use it in smarter ways, saving millions of dollars a year in online marketing spending. “Databricks Data Intelligence Platform provides us with a centralized data layer that gives us the confidence that the data we’re giving our various teams and our customers is both reliable and complete,” added Yovchev.

For example, one of Relay42’s customers is a major telecommunications company in Singapore. When a shopper searches for a phone plan or buys a smartphone, Relay42 can use customer data to quickly predict how likely that shopper will convert. That prediction is then fed into a decision engine based on anonymized personal identifiable information (PII) to create profiles that can help inform the company’s next steps. In this way, Relay42’s client has a better understanding of what messages to send the shopper (and when) to drive conversions, all without throwing money into an advertising black hole.

Lakehouse architecture that delivers value, cost-effectively

For Relay42, one of the biggest benefits of turning to Databricks is the significant increase in the speed of data access and insights delivery to their customers. What used to take 20–24 hours to make complete customer data available (with frequent crashes) now takes 1.5 hours to move data through a pipeline to the eyes of their customers. But most importantly, they are now able to democratize this data more reliably.

“Our ability to ensure the rapid and reliable delivery of data has been a huge boon for our customers,” expressed Yovchev. “Now what used to take upwards of 8 hours for real-time analytics can be done in almost 15 minutes. This drives customer satisfaction, which has a direct correlation on retention.”

The stability of the Databricks Data Intelligence Platform has also resulted in huge productivity gains and cost savings for Relay42. Not only is the company saving over €180K in data storage costs per year, but it’s also realizing over €150K in savings from simply freeing its infrastructure team from having to spend precious time fighting a buggy system. Looking ahead, Databricks has laid the foundation for Relay42 to realize the goal of providing a self-service data platform that gives internal teams the ability to explore data for AI and analytics purposes. These gains aren’t only beneficial for Relay42, though — its clients also reap the benefits of data democratization, which in turn opens up a world of possibilities for serving their customers with more meaningful and seamless experiences.

Learn more about how Relay42 is delivering value to its customers here.