Working in the Open: One Year Retrospective from USDR and CDS
By Janell Schafer, CDS and Jessica Cole, USDR
Over the past year, U.S. Digital Response (USDR) and Colorado Digital Service (CDS) tried something new. Less than a year into both forming, we decided to experiment with ways that a government digital services team and an organization offering pro bono technical expertise could partner to get things done in a moment of global crisis.
Colorado Digital Service launched in late 2019 within the Governor’s Office of Information Technology to develop user-centered solutions for Colorado’s most pressing technical challenges. Only months later, in March 2020, U.S. Digital Response was established to serve the urgent needs of communities across the country by pairing pro-bono, best-in-class technologists with state and local governments. Now, more than a year after our first work together, USDR and CDS have partnered on a series of projects, including supporting an IT help desk, coordinating exposure notifications, streamlining information about COVID-19 vaccines, and improving equity in the state’s public health communications.
Both USDR and CDS value working in the open so that others can learn from us and build upon our efforts. Together, we’ve reflected on our partnership and have jointly authored this piece to share candid stories and recommendations about how similar partnerships might work. Here, we’ll cover what went well, where we hit bumps, and our ideas for the future.
What Went Well
We aligned on a mission and end goal
“USDR and CDS are culturally aligned with this notion of setting up a team of people who are government workers and volunteers working together to accomplish something impactful.”
— Steve Leibman, USDR Volunteer
The Colorado and USDR folks came to the table with different cultural norms. By identifying our shared goals, like getting a specific service into the hands of Coloradans as soon as possible, we were able to rapidly build team trust in order to get things done. USDR volunteer Steve Leibman saw this firsthand. Steve has spent most of his career working as a software engineer with small companies and start-ups. After volunteering with USDR, Steve was placed on a project with CDS and immediately noted the unique dynamic of a team with both government and private-sector skill sets and backgrounds. According to Leibman, the diversity in experience was bridged by a shared desire to get things done and have immediate impact. The cross-functional work with USDR and CDS allowed the team to mature the product, automate key analytics, and share learnings with other states. The more we were able to align on our goals and tap into our diverse skill sets, the more we were able to be efficient and make an impact with our work.
We identified a discrete technical pain point to address first and then scaled from there
Exposure notifications is a technology that anonymously allows people who’ve tested positive for COVID-19 to privately alert others they have been in close proximity to possible exposure. The implementation of this technology in consumer phones was a collaborative effort among Apple, Google, higher education institutions, the CDC, and several states, with privacy as the shared common focus. This led to a challenge in gathering enough anonymized data across multiple platforms to inform and improve the service.
USDR worked with CDS to improve the analytics for the Colorado configuration of Exposure Notifications Express. The work focused on automatically aggregating metrics across multiple anonymized aggregate data sources to help the state understand the progress and effectiveness of the service. Prior to this collaboration, gathering metrics for the service was a manually intensive process.
As a way of starting by solving an initial need before scaling up, CDS worked with USDR engineers to automatically collect key metrics across different platforms into an aggregated database, updated daily for display on a spreadsheet. The team began with development in a private CDS repository, then gradually moved this into the state’s production environment.
We built more sustainably by involving additional state colleagues over time
USDR and CDS provide resources to projects on a time-constrained basis. USDR sources volunteers who have obligations after their volunteering concludes, and CDS as a team needs to stay nimble to address new high-priority efforts. Technologies like Exposure Notifications have longer time horizons, and for this specific technology, are joining a suite of public health technologies that will be around after COVID-19 subsides.
It was therefore important to transition product support to agency staff in the Department of Public Health and Environment (CDPHE). Additional IT staff joined the effort to further integrate the technology into the state’s existing tech stack. New CDPHE staff are now on board as the long-term owners of this product for the state.
We improved communication between CDPHE, CDS, and USDR teams throughout the project
“We kept a close loop with the volunteers throughout the project to understand their individual time commitments and blockers. It was also important to stay flexible with the amount of structure we created for the team. For example, formal sprint structures and daily standups could’ve been more burdensome for volunteers who didn’t have as much time allocated to us.”
— Michelle Park, CDS Product Manager
At the beginning of our projects, CDS prioritized getting to know the volunteers to help the team dive deeper into the work together. After the initial onboarding process, the team focused on creating smooth communication channels via chat for frequent communication and designated a clear contact person for volunteers to ping if things came up and help prioritize requests for their time. CDS also maintained close feedback loops with the volunteers to understand when they wanted more or less structure for their work throughout the course of the project. Perhaps most importantly, we had conversations along the way about where the work could live long-term and how a clean hand-off and transfer of knowledge could be structured when volunteers needed to roll off. Having those important discussions throughout the course of the engagement made it so the team could seamlessly move from one phase of the project to the next without confusion or communication gaps along the way.
Where We Hit Bumps
We needed to iterate to determine the right opportunities for volunteer support
Before collaborating on exposure notifications, Colorado first sought USDR’s support with a help desk. The ask was to help rapidly source volunteers who could support the state’s workforce as it transitioned to telework. Many state employees were using some remote collaboration tools for the first time in the early days of the pandemic. Volunteers were seen as a way to quickly divert help desk tickets that didn’t require intimate knowledge with Colorado’s systems, but were more generic in nature (video conferencing tools, chat tools, etc).
CDS reached out to USDR for volunteers and started onboarding them with the help desk. The team immediately hit setbacks as volunteers were required to go through expedited background checks, even for this lightweight engagement. After a volunteer cleared background checks, they were then required to take a several-hour training course on the ticketing system so that they could be assigned help desk tickets. Getting access to that training and system took additional time. By the time the first volunteer cleared the process, the surge of help desk tickets was abating and the help desk had found contractor resources to better handle demand.
“This was the wrong project to use volunteers on. The type of help we needed required background checks and system access, which is a high barrier to entry for volunteers and slows down our ability to respond in a rapid fashion.”
— Matthew McAllister, CDS
We learned to clarify ambiguous problem areas and broad scopes into discrete needs
CDS and USDR learned that an important variable to success is defining an appropriate scope to match the State of Colorado’s needs and timelines with volunteers’ availability. Over summer 2020, CDS began supporting a project called ‘COVIDTech,’ a new organization structured around the technology products supporting the COVID response. The CDS team recommended that each focus area under COVIDTech be structured as a cross-functional technology team with roles such as public health experts, product managers, user experience designers and researchers, and engineers. Roles such as ‘product manager’ and ‘user experience designer / researcher’ are often newer concepts at state agencies, so the CDS team knew they would have to prove their value before CDPHE would be interested in hiring them.
USDR offered volunteer product managers and user experience designers/researchers to embed within each COVIDTech content area. The volunteers were tasked with helping onboard their teams to a new way of working that was centered around modern software development practices like agile and human-centered design. The volunteers were committed to 10 hours a week, but each had full-time jobs and were often only able to engage after hours. This provided unique challenges as the culture of COVIDTech at the time was extremely meeting heavy, making it difficult for volunteers to fully embed and contribute to the team. After this experience, we designed volunteer opportunities differently so that they did not rely heavily on organizational knowledge in order to be successful; USDR and the State of Colorado worked together to identify not just the total amount of time that volunteers should have available per week, but any specifics around daytime availability for shared meetings.
With a year of collaboration under our belts, we plan to continue working together with these lessons top of mind and a clear sense of what makes engagements with pro bono support successful. These lessons include:
- Tightly scope projects that can be completed primarily outside of normal working hours
- Communicate the state’s development environment and process to volunteers early to increase alignment and set expectations for access delays
- Identify a technical lead from the state to pair with volunteers
- Iterate on how to deploy code inside state infrastructure early on (if the project involves development)
We hope that by working in the open, we can encourage other government digital teams and nonprofit organizations to partner successfully so that they can more quickly meet the needs of their communities. We also believe in making “only new mistakes,” and we hope our experience encourages ongoing reflection about the best ways we can support each other in this work.
Colorado Digital Service is helping teams throughout the State of Colorado recruit and hire various roles in the digital services space. Learn more about CDS and opportunities with the state here. If you are a state or government interested in partnering with U.S. Digital Response, fill out our brief intake form and we’ll get back to you shortly.