Data Product Teams : Best Practices for a Modern Data Team

Data teams are becoming software engineering teams.

On December 14th, we welcomed Philip Zelitchenko, VP of Data from ZoomInfo, to talk about how he has built this discipline within his team & it was fascinating. The video is here.

Like the Devops movement, the Dataops movement aims to scale the use of data within companies without increasing the headcount of the data team.

To do that, Philip defines data products using DPRDs, structures his data team with five key roles, & defines clear roles between the data team & others in the company.

DPRDs, or Data Product Requirements Documents, contain the key information about a data product: what it will provide, how it will produce value, how the data will be governed including data quality alerting.

Unlike code, data is stochastic or unpredictable. Data may change in size, shape, distribution, or format. This adds an additional dimension of complexity to the DPRDs.

In addition to the DPRD, the ZoomInfo data team employs TEP or technical execution plan that aligns the internal technical teams on architecture & governance.

The data team has five key roles:

1. Data PMs : quarterback the DPRDs. They gather feedback from users, define the value, solicit feedback from the rest of the team, then manage the execution of the plan.
2. Business logic : the data engineering team build the ETL pipelines while the data science team researches & implements machine learning algorithms for ML\DS driven data products.
3. Data analysts : embedded/seconded to the different operating teams, analysts analyze the data each team needs using the infrastructure provided by the data platform.
4. Data governance : ensures data quality/accuracy, defines the access control policies for security, sets the operating procedure for alerting & monitoring, and help define data contracts between producers, processors, and consumers.
5. Data platform : builds the universal data infrastructure for the company.

Last, the ZoomInfo team is building an internal product called Heartbeat that measures usage across the main data products, evaluate the priority, SOPs for impact on SLAs and communication with data practinioers across the org in an automated way.

For Philip, leading the data team is about focusing on the data products that drive meaningful value to the company. I learned a tremendous amount about the way modern data teams, who leverage software engineering disciplines, operate.

Thank you, Philip!