As the data services lead inside the Maryland Information Technology Department, Julia Fischer heads up a team of eight people.
The statewide office’s role is to direct strategic enterprise data efforts. Its customers are other state departments and employees, a force of greater than 15,000 people.
Some basic arithmetic finds there are almost 2,000 state employees for every member of Fischer’s statewide data services team – an unflattering ratio representative of the uphill challenges that confront government data and analytics teams.
Central data offices across government are outnumbered in staff, time and budget.
The aspiration for these offices is to coordinate incisive, cross-cutting analytical and data-based projects, improve data maturity and expand data services. The reality is more foundational: problem-solving and trouble-shooting backlogs of requests, while building basic data literacy and standardization.
“I believe that those fundamental conversations are even more important than those conversations on how to analyze data,” said Fischer, Director of Data Services, on GovLoop’s recent panel about data skills in government.
Stefanie Costa Leabo, Boston’s Chief Data Officer, counts herself as fortunate. She has a staff of 20, one of the more impressive tallies for local government data departments. But she too has found that it’s helpful to take a step back before launching into aspirational analytic projects.
Her goal is to foster data literacy across departments and, hopefully, stoke internal investment so that one 20-person department isn’t responsible for a whole city’s worth of analytics. Those efforts, which include one-on-one trainings and templates for how to design surveys, have had success.
The Fire Department has taken her up on her ask. After collaborating closely on several projects, the Boston Fire Department hired its own data point person, and it’s now a self-sufficient, standalone analytics operation.
Elsewhere, central data departments have crafted campaigns to improve data literacy and recruit data practitioners. Outsourcing analytics projects is often accomplished through reskilling and upskilling, whereby people without data science degrees are trained up and become functional data owners in their enterprise.
Such training is important, because it educates people on cybersecurity, data fidelity and analytics. These elements must kick in before moving to data visualization and analysis technology, experts agreed.
The U.S. Agency for International Development (USAID) is reskilling and upskilling its employees on data, including by setting up an informal learning group for Tableau, the popular data visualization tool. The agency is also part of a cross-agency pilot to hire data scientists with a standardized set of certifications.
“There’s just so much that goes into being able to use the analytical and visualization software that goes beyond just nuts and bolts,” Julie Warner, Data Scientist at USAID, said.
Expanding data’s role in an organization requires proving its value too, said Andy MacIsaac, Solutions Marketing Director for Public Sector at Alteryx, a self-service analytics platform. Impactful, small pilots with traceable results are a good place to start.
“Any analytics work you do, you really need to have transparency built into that,” MacIsaac said.
As data departments try to expand their influence and prove their work, time is of the essence.
Many government data scientists, familiar with existing systems and structures, are retiring or will be soon, leaving knowledge gaps and vital positions to fill. Though it can be tough to fit it into an already packed nine-to-five, training needs to happen now, Fischer said.
“You just have to force [the issue] at some point,” Fischer said.
This online training was brought to you by: