IE11 Not Supported

For optimal browsing, we recommend Chrome, Firefox or Safari browsers.

S.F. Data Academy Spreads to the State, Sets National Model

Joy Bonaguro launched the San Francisco Data Academy during her time with the city. Now as California’s chief data officer, she’s expanding and improving the model to support agency-level analytics.

When Joy Bonaguro was San Francisco’s chief data officer (CDO), she set a national example for improving public employee data literacy by pioneering the SF Data Academy, whose mission was to galvanize the use of data in decision-making and service delivery throughout the city. The initiative assisted departments with preparing data, dashboarding and digital storytelling, advancing statistical modeling, and using tools such as Power BI, ArcGIS and Tableau.

Now, Bonaguro is taking this approach to the next level in her new role as the state of California’s CDO, iterating and enhancing her original model.

According to Bonaguro, the DataSF academy “was a victim of its own success. We never could teach enough classes to meet demand. And sequencing of classes was especially difficult due to limited room capacity and availability.” At the state, she is taking the lessons learned in San Francisco and combining them with the state’s newfound familiarity with telework.

The 2.0 model continues the basic premise of free courses, practical hands-on training and peer teachers. What is different is how they plan to administer courses and their overall ambition. The state will explore combining live teaching with recorded content and participatory activities. As constraints of room availability and capacity diminish, the focus is on nurturing and maximizing the impact of its most valued asset: peer instructors. According to Bonaguro, “If a peer instructor has eight hours a month to give, what is the optimum use of that time? Is it eight hours live instructing; two hours live instructing, combined with recorded content, and then six office hours; something else? We aim to find out.”

In addition to experimenting with a remote-centric format, they also intend to use this newfound capacity to reimagine what’s possible with the learner’s journey. In SF it was a challenge to get a single person through a standard sequence of Excel courses in a timely fashion; sometimes there would be a six-month waitlist between intro and intermediate. According to Bonaguro, “As a result we had a collection of courses, and not the true academy we envisioned.”

An analytics accelerator will complement the 2.0 model and “aims to provide an omnibus training experience for entire data teams,” said Blake Valenta, the statewide data programs manager. The accelerator will give interested data teams access to a curated schedule of classes and access to one-on-one technical assistance and office hours. In exchange, it requires department stakeholders to identify a relevant pilot project and commit staff time for “out of classroom” homework. In this way, attendees immediately apply what they learn. This approach is more experiential. Contextual learning means that learners don’t need to complete a multi-month course before applying skills.

Valenta added that having self-service components and optional office hours “helps individuals complete a project that they’re working on so there’s a tangible deliverable at the end... . By tying training to a particular project and making sure the project has buy-in from the learner’s stakeholders, there’s a bit more guarantee that the learner will be given the time to apply the learnings and allow them to stick.”

Bonaguro recognizes that many agencies frequently endure “data fire drills,” where too much time is spent answering basic and ad hoc questions while scrambling to pull the relevant data. To address this, the analytics accelerator will initially be focused on skills and tooling to automate reporting. With that freed-up capacity, the team wants to provide toolkits, training and technical assistance to support departments moving up the data maturity levels from improving their metrics, building out modern data teams, leveraging agile and user-centered processes, and incorporating data science methods.

The data literacy initiative complements broader efforts to transform state operations. Data literacy will empower data analysts at the agency level, while the work of the CDO’s office supports recruiting, hiring and retaining employees with these skills. Broad reform powered by data requires training that extends from top to bottom in each agency. Executive-level officials will have dedicated trainings tailored to their needs, so they understand the best data and innovation approach for a given decision. Staff-level training will be available to support any given data framework.

“We are here to build capacity,” said Bonaguro, “not be capacity.”

This article first appeared in Government Technology, sister publication to Industry Insider — California.
Stephen Goldsmith is the Derek Bok Professor of the Practice of Urban Policy and the director of the Data Smart City Program at Harvard's Kennedy School of Government. He previously served as deputy mayor of New York and mayor of Indianapolis, where he earned a reputation as one of the country's leaders in public-private partnerships, competition and privatization.