Code for America Summit 2024 : Day 2

Impressions from #CfASummit in Oakland - May 29, 2024

Four of us from the GX Foundry made it out to Code for America’s annual Summit this week, and we’re attending all 3 days. Below are some notes and thoughts from Day 2 (which is actually the first day of the full-blown conference).

Tech journalist Kara Swisher was on hand for the conference kickoff, accepting an award from Code for America. She also shared a few comments from her current bestseller Burn Book: A Tech Love Story.

Kristen

Day 2 was full of impressively good discussions! The sessions I attended were largely focused on how to best serve your users, whether they be the public or the employees you’re supporting. Here are two sessions of note…

I attended one session titled Making Lemonade: When Well-Intentioned Policy Is at Odds With User Needs. During this session, the speakers discussed legislation passed in Colorado that tasked the Colorado Behavioral Health Administration (BHA) with establishing a tool for providers to track and report their care capacity since it’s so difficult to find an available bed in a mental health facility when an individual needs support. The initial tool the BHA planned on building completely changed after talking to the providers themselves, who pointed out the new challenges the tool would create while only minimally improving the current placement struggle. User research is so important — without doing that research, you might risk implementing a tool that doesn’t support your customers, regardless of your intention.

We also discussed Product Management in a panel titled Unpacking Product Management in Government: What Makes It Unique and Where It’s Going Next. It was really amazing to hear how folks are doing product management across the government, the panel sharing the sentiment that “product management looks like whatever you need it to be.” The tasks involved in managing your product may change over time, or they may be different from one organization to the next, but the core of product management remains the same — you have to focus on delivering customer value through continuous improvement and innovation.

Sarah

I wanted to attend the research-focused sessions to help make my case for adding research staff to my team, honestly. I was able to hear a lot of examples of how research (design, user, community based participatory) was able to bring the problem and potential solutions into focus by centering the people most impacted. Yes, project timelines needed to shift, phased approaches and even changes to scope were all necessary. But in the end real people were able to get real value out of the work done instead of just measuring deliverables. To me, research (which is really just a formalized way of having conversations and listening to the people most impacted by the decisions you’re making) is a way to measure the distance between legislative, policy, or technical intent and real community impact.

The other big takeaway for me was that we’re on the right track. To be honest I’ve never built a website unless you count using GeoCities in 1998, and I’ve had plenty of second-guessing myself leading the website redesign project. But hearing the digital team at New York City talk about their similar website consolidation and improvement project and the questions people asked, I realized that while we may not have all of The Answers™️ we are absolutely asking the right questions, and we are centering people, trust, and relationships in the right way. It feels like a much needed reassurance to keep going.

Eric

In yesterday’s keynote, I really appreciated the focus on accessibility. 48% of federal websites have accessibility errors, and across public and private entities, the top 1 million sites have an average of 60 accessibility errors. It was proposed to go beyond compliance and tackle accessibility by design. Usability testing, questioning vendors’ accessibility documentation during procurement, and changing workplace culture are all part of solving the problem.

Yesterday I was impressed by the procurement session where I heard about how the process is being remedied in Portland, Los Angeles, and Boston. I was overwhelmed to hear that LA County has to coordinate procurement with the 88 cities located within the county. (Conversely, Franklin County contains 16 cities and 17 villages.) All 3 jurisdictions utilized gaining trust through equity and being nimble in dealing with small businesses.

I also attended a panel about designing local AI policy, interestingly without any AI experts. There was discussion on building an understanding of the opportunities and risks of AI, identifying opportunities to engage the public and across agencies, and using AI to drive towards equitable outcomes (think: AI has helped Flint, MI find lead pipes, which helps alleviate the water crisis.)

John

So about AI…

I’m done hearing platitudes and pronouncements about “AI” and how it’s going to transform everything. Thankfully Dr. Safiya Noble capped off the mainstage session with a reality-check around AI’s impact on the environment, it’s questionable value in practice, and the dangers of automatically trusting algorithms generally.

I would encourage everyone in government digital services to do a lot of soul-searching about what you’re doing with AI, what you’re intending to do, and what you think these tools are really going to achieve—and at what price. I suspect we will look back from 2034 and shake our heads wondering how the tech bros here in Silicon Valley duped us so completely into thinking this was another printing press moment.

To be clear, I believe AI-ish features have some value, and there will be further refinements in our application of algorithmic auto-completion bots. But it’s not going to wipe out humanity nor solve our most intractable problems. When the dust settles, we will still have to solve the thorny problems of policy vs. delivery and development of trust with the public ourselves.

Other mainstage notes

• Katie Fiore from New Jersey had a killer section of the mainstage presentation that resonated a lot. She talked about communicating with constituents clearly and directly and with accessibility in mind. The only problem with the presentation was that it was too short; she deserved more time.

• Seeing Kara Swisher was entertaining, for sure. I listen to Pivot every week and enjoy it. Although I have a nagging suspicion the award granted to Swisher was mostly to generate buzz for the conference, since her work is pretty tangential to government digital services. I would have preferred to learn more about the other award winners that were on the stage too briefly. Let’s celebrate the practitioners.

Low-code tools

The low-code tools session was great. I didn’t pick up a lot of new info, mostly because we’ve already embraced this notion (in our case primarily with Quickbase) and we’ll continue to do more with low-code solutions in the future. But it was awesome to see the infectiously positive attitude Melissa House from USDR brought to the presentation. Not a lot of engineers can pull that off, so kudos to her.

Community-based research

The session on Code for America’s “opportunity youth” research project in Shelby County, Tennessee (Memphis area) was stunning. Not enough people attended that session. This project was a real achievement because they were able to connect with local youth and build a powerful report to highlight the needs of community youth that are struggling with a lack of education and employment opportunities.

There’s way too much for me to summarize, but it was wonderful to see a join effort between Code for America, Shelby County government, a local nonprofit, and young people right in the community that actually conducted the research and compiled the results from their peers themselves.

The only problem with this session was we wanted more—we wanted to understand some of the mechanics of how the research was developed and done, how the results were compiled into the impressive report, and so forth. I hope CfA shares more details in the future.

Product Management

Similar to Sarah’s comments above, this session was one where I felt like the direction we started to take nearly 2 years ago was the right one — pivoting toward product ownership (or management). We’re not there yet, but we’re headed that way and I’m happy to see we picked up the thread others laid down ahead of us.

One interesting takeaway, which played into a following session, was the idea of teaching “product thinking” across the board, to agencies and others that provide direct public service. It seems simple, but I think it’s lost on a lot of folks. We need to share these ideas broadly.

Training government workers: a breakthrough

We’ve needed to hire a “trainer” for years in our organization. We spend millions on a variety of software platforms every year, but we don’t support those technical investments in building skills in the employee base using those platforms (which is insane, but here we are). So I wanted to pick up more training delivery ideas from folks here at #CfASummit.

And the bi-coastal presenters — Caitlin Seifritz (Philadelphia) and Amy Martin (San Francisco) — delivered big-time. They had lots of specific recommendations, the best of which may have been this slide about which training approaches carry the most value:

This idea, plus the idea of creating a training coordinator role was a big insight. It raised a question for me… What if we pitched a training role that was more than just training development and delivery? What if all the training they developed was really just cloaked consulting, where the person goes into a specific agencies or team, finds a service delivery problem they are struggling with, and figures out how to apply a technology they already have in hand? That would make the training specific, custom, and focused on an active need the team has in the real world.

Further, what if this training coordinator drew on the Product Owners we have in place? For example, if you have an M365 product owner, and the target team would benefit from better collaboration using Teams, the training coordinator would make an introduction between the team and the M365 product owner, and the product owner is on the hook to ensure the users get the training they need, tailored to their situation.

Finally, as former librarians the presenters addressed a problem we specifically have: rejecting customer inquiries that are “malformed” according to our own definition. We allow ourselves to get salty about customer requests that come in the “wrong door” or are “incomplete” in vision or description. (This is common for tech teams.)

Instead, we need to focus on developing a Reference Desk model, similar to a library. It’s expected that users come to the Reference Desk with incomplete concepts or questions, then the Reference Desk team will collaborate with the user to clarify their ideas and find resources.

That’s what we need to do in all our work.

---

There’s one more day at the #CfASummit and we’ll share that tomorrow.

---

Check out our notes and comments from the other 2 days of the Summit:

• Code for America Summit - Day 1

• Code for America Summit - Day 3