Payroll and HR solutions giant ADP generates vast amounts of data, which it anonymizes and delivers as data-driven insights to customers, governments, financial analysts, and university researchers. On a recent episode of our podcast, Rise of the Data Cloud, Jack Berkowitz, Senior Vice President of Product and Development for ADP DataCloud, shared how Snowflake is helping ADP build a new cloud-based data distribution business.

In business for more than 70 years, ADP processes payroll, paying about one in six people in the U.S. and issuing around 70 million W2 forms annually. The company provides its customers with HR systems and also handles insurance, retirement accounts, and money management. This work generates a tremendous amount of data, from which the ADP Research Institute can generate insights and economic indicators, including trends about hiring and wages.

“Broadly speaking, you can use ADP’s data to help understand the U.S. economy, which then amplifies out to the world economy,” Berkowitz said. ADP’s data is unique because it is derived directly from payroll and HR transactions. It combines this data with external information and uses neural networks to extrapolate from patterns it sees in its customer data to what’s happening in the larger economy.

ADP DataCloud provides people analytics, compensation, and HR benchmarking products to HR departments. Berkowitz’s team also builds and operationalizes machine learning applications across ADP’s HR, Payroll, and Time applications. Customers and partners can also purchase or license ADP’s data for their own decision-making, such as for demand planning.

Achieving Pace and Flexibility in the Cloud

Snowflake is playing a major role in helping ADP build a new kind of data distribution business in the cloud. According to Berkowitz, “Going to the cloud is about pace and flexibility.” ADP must get its information out as quickly as possible and then make it easy for customers to consume that data—from inside the specific data system a customer is using, for example.

ADP has been working with the Snowflake Data Cloud for the past year and is hearing from increasing numbers of customers that they want ADP to distribute its data through Snowflake. ADP uses the Snowflake Data Cloud to calculate, store, and distribute its anonymized compensation and benefits benchmark data, as well as its aggregated national data. It is also providing that information through existing SaaS applications, such as Real Income, using the Snowflake Data Cloud.

“So, Snowflake is essentially acting as a multi-tenant SaaS enabler for us for a visualization capability for our HR and our payroll customers,” Berkowitz said.

Powerful Distribution Capabilities and Richer Data Sets

Snowflake Data Marketplace is very attractive to ADP as a way to accelerate data sales and gain traction with more data scientists. In effect, Snowflake Data Marketplace streamlines and takes the friction out of the scientists’ ability to use ADP data in the course of their analysis.

According to Business Wire, ADP is also combining its information ahead of time with data from other partners, such as Intercontinental Exchange (ICE)1, to generate even richer data sets. ADP is working with ICE Data Services to build a value-add combination of their information for the U.S. municipal bond market and distribute it via Snowflake Data Marketplace.

“This is a use case which ADP would never have gotten into,” Berkowitz said. “But by using Snowflake, we’re able to combine ICE’s and our own data and come out with a new data product that no one’s ever even thought about before.” He’s excited by the future possibilities for more combinations of data for all sorts of different use cases.

Looking ahead, Berkowitz believes that everyone in the data world should be thinking about the ethics of when to apply and how to use data algorithms. “You’re going to see that trend increase as more and more people pick up machine learning operations and tools,” he said. “This all needs to be transparent, and you, as a technologist, will need to make a decision on how you want to behave and what you want to do.”

ADP has already published its AI data and ethics guidelines and holds regular meetings of an advisory board made up of ADP employees and outside experts who talk about the issues and review programs to ensure they align with ADP’s values.

Rise of the Data Cloud is a podcast hosted by award-winning author and journalist Steve Hamm. For each episode, he speaks with a data leader to learn how they leverage the cloud to manage, share, and analyze data to drive business growth, fuel innovation, and disrupt their industries. You can listen to more episodes here.