How to Develop Data Governance in Your Organization

Photo: Freeimage4life/Flickr

Whether your government shares data with the public or internally, it's essential to have a data governance process in place that addresses how data is maintained as well as when, how, where, and to whom data is released.

Good data governance prevents privacy and security issues and ensures only verifiable data from reliable sources is published — and only after it's vetted and approved.

At Socrata Connect, I had the opportunity to chat with my former colleague from Prince George's County, Ben Birge, CountyStat Manager, along with Mike Rowicki, Assistant to the Chief Strategy Officer in Fulton County, Georgia, on some of the important considerations for governments intent on sharing data.

Buy-in at All Levels is Key

Executive level buy-in for any data program is important to get it off the ground, Rowicki said. But so too is support from department-level partners, the custodians of data, and staffers who frequently engage with the data.

"A communication plan is key as to how you actually work with everyone," including directors, supervisors, and day-to-day staff, Rowicki said.

Staff training sessions can reveal how open data platforms make developing reports and basic work functions easier and speedier than waiting for reports from IT, Rowicki said.

In Prince George's County, data is tied to the budget. This gives agencies a significant incentive to share data. It's used to justify expenditures, headcount, technology, resources, and other agency spending, Birge said.

"In Prince George's County, a lot of what we're doing data-wise starts with agency data," Birge said.

As performance data became a key element to the budget process, the county saw a larger demand from agencies about how to use it.  

Varying Structures

Governments can choose how they want to operate with data.

There are three basic options: a centralized board; a decentralized, department-led structure; or a hybrid of the two.

Fulton County has a 20-person centralized committee, with members from legal, purchasing department, finance department, IT, executive management, and others. Together, this working group makes determinations about what data to publish and how to prioritize it. Department leaders make the final decision on whether to publish data.

This framework assuaged concerns from directors worried about losing data ownership, or having a lack of control over messaging, Rowicki said.

A Data Policy

The best practice is to have a plan ahead of any data management emergency — don't create a solution in search of a problem, Birge said.

This plan could include releasing data to executive leadership without going through a vetting process or putting up unreliable data. While the problems that arise with a haphazard release of data are clear, that doesn't mean that there's only one strategy for sharing data, Rowicki said.

"Everyone's jurisdiction is a little different in how they operate," Rowicki said. "You're going to have to define what works best for your jurisdiction. Don't be afraid to try anything."

Watch the Full Conversation

For more of the conversation on developing data governance, watch the full video: