Our four day discovery sprint with San Jose

A big group photo of everyone involved in our San Jose discovery sprint

In mid-September, a small team from Ad Hoc conducted a four day discovery sprint with the innovation team in the city of San Jose. Ad Hoc is experimenting with different consulting models aimed at helping governments adopt new ways of working, starting with understanding the problem they’re trying to solve, and experimenting with both citizens and public servants to rapidly identify opportunities and shape solutions. We think these engagements can not only help our clients solve real problems, but also demonstrate that more experiment-driven ways of working are possible and worth investing in.

The focus of our sprint was MySanJose, a mobile app that residents use to make service requests of five services, such as reporting abandoned vehicles, or graffiti. The app had a successful launch in 2017–in fact, too successful. The app was generating requests much faster than some service lines could respond to them. The team was on the verge of a new procurement, and not sure exactly what direction they should take with the product.

Our goals

Our goals with this engagement were to:

  • Break down silos between departments and promote shared awareness of the key challenges and opportunities facing the San Jose team
  • Make assumptions explicit, and test them with rapid prototyping and usability testing
  • Identify areas for exploration and learning opportunities
  • Build capacity in our government partners by helping them learn by doing

The San Jose team lacked a concise, consistent vision statement and set of priorities for what to do next. There were a lot of possibilities, and the product meant different things to different stakeholders. At the very least, we wanted to bring some alignment across the teams and stakeholders, and help the new product owner, Michelle Thong, establish a shared vision and set direction.

There were also a lot of assumptions as to why users behave in a certain way and how the app could be changed to influence their behavior. By stating these assumptions explicitly and crafting hypotheses we could test to better understand user needs and behavior, our partners not only gained exposure to lean product methods, but also became key participants in the design process.

We also wanted to give our partners a framework that they could weave into their existing processes that helped them explore and validate new ideas. One benefit of discovery is that you can identify areas where the biggest unknowns are. These unknowns include both inherent risk if not explored and enormous possibility. Over the course of four days, we showed the San Jose team how to explore a wide range of possibilities, and then narrow in on places where we could learn the most and have the biggest impact.

Image showing a chart that diverges over time, starting with a big difference, that later decreases.

The sprint

The basic sprint format went like this (if you’re not interested in details, skip to the bottom for the result):

Day 1:

AM: We held a kickoff with executive leadership from the city manager’s office and from all affected service lines, which included a high level service blueprint exercise

PM: We further fleshed out the service blueprint with “on the ground” folks from each service. We then identified a vertical slice of the blueprint for one service (illegal dumping) to focus on for the rest of the sprint.

Day 2:

AM: We went on ride alongs and conducted user interviews with staff fulfilling service requests. We also held deeper conversations with city staff on their triaging and fulfillment processes.

PM: We conducted research interviews with users who have used illegal dumping in the past, synthesized the results, and started crafting some initial hypotheses with our partners. We then prioritized those hypotheses with our product owner. Our offsite designer then started working on a few concepts for usability testing.

Day 3:

AM: We conducted remote usability testing with clickable prototypes in InVision. Remote testing using Zoom allowed a lot of our stakeholders to observe at once, without overwhelming the user. After our initial tests, we crafted a new hypothesis and conducted a design studio to come up with more options. Then our designer incorporated the ideas into a new prototype.

PM: We conducted more usability testing with more iterations of the prototype! Overall we did 3 iterations in one day.

Day 4:

AM: We synthesized our research, crafted a vision statement, and held a product roadmapping workshop with the core product team.

PM: At last, we held an executive read-out with all of the stakeholders with a presentation of the new vision statement, and roadmap.

The result

In the end, our product owner Michelle came up with a vision statement for the new direction of MySanJose:

“MySanJose makes it easier for residents and government to work together to keep San Jose safe and clean.”

This statement captures what MySanJose will and won’t be for the foreseeable future. It’s an articulation of value that our product owner can use when evaluating requests, prioritizing features, and coordinating stakeholders. It also captured the ethos of the feedback we got from many of the civil servants in San Jose. They were overwhelmed with the volume of requests from the app and weren’t set up for success, but they also weren’t going to be able to pick up every piece of litter someone sees on the ground. The emphasis on residents and civil servants working together resonated with them.

This creates lots of opportunities for the MySanJose team to explore not just a better user experience with the app, but also how citizens can help, either by providing more details about a request or learning how to recognize situations that require special attention and routing (i.e. a homeless encampment will not be handled by the illegal dumping crew). This creates space for a much broader impact than just focusing on one piece of the user experience.

The players

Special thanks to all the folks from the Ad Hoc team who participated, including our product director Larry Bafundo, and our visual designer and prototyper Savannah Million. And of course, we couldn’t have done it without the amazing product team in San Jose, including Michelle Thong, the MySanJose Product Owner; Charles Amith, the MySanJose Product Manager; and Nira Datta and Julie Kim, the Code for America fellows that were a huge help with user research and recruitment.

What’s next?

Interested in working with Ad Hoc on something like this? We love thinking and talking about hard problems and how to improve government digital services. Drop us a line at innovation@adhoc.team.