What if we built a better Code for America Summit?

I’ve now been to the last two Code for America Summits (2024 and 2025).

But I don’t know whether I will return. Maybe? Not sure.

Why? Because the signal:noise ratio in the conference feels too low.

I’m a stickler for conference value. My standards are unusually high for a couple reasons. First, I used to work in public media (PBS, NPR, APM, PRX, etc.) and those industry conferences were superb. (It turns out if you know how to put on a show, you know how to put on a show.) Second, with the explosion of webinars, streaming, and virtual events in the last 10 years, and especially since the pandemic, I hold in-person events to an even higher standard because they are “more expensive” to attend.

Looking for ROI

In short, if I’m going to burn T time and D dollars to attend a conference in person, the return on investment (ROI) must be… ROI ≥ (T+D). And I don’t mean “return” in a financial sense. I just need to come away feeling like I got more value out than I put in—that being there made a difference.

So as I review the schedule, content, and costs for the Code for America Summit in 2025, I’m concluding the ROI isn’t making the cut, at least for me.

Part of my concern is the realization there’s much more value at the conference than I’m getting. There are tons of people doing excellent and fascinating work in the civic tech sector all across the country, we gather a few hundred of them in one space, I walk amongst them, and then… we hear from just a handful. Part of this is normal and unavoidable. But CfASummit’s structure and style unnecessarily amplifies the separation of talented attendees from one another.

So… what do I do about it? Well, I could just stick to virtual events and save my time and money. I could keep attending and just accept it. Or I could look for other conferences—like the NAGW in Denver this fall—and go to them.

Or I can use my “fixer” mode—skip the complaining and offer some ideas. And that’s what this post is about. I want to know…

• What if we built a better Code for America conference?
• What would that look like?
• How would we get there and who would help?

First, a few disclaimers

• I do not work for or represent any of the organizations I mention below and I never have. At best, I’m a “customer” or “supporter” of these organizations, but I have no role in their strategies or priorities. My commentary is from the outside looking in, and that means I may get stuff wildly wrong. I’d research my commentary further and build more specific recommendations, but alas… I’m not a journalist and don’t have time to do the extra work on my own. But in the spirit of blogging in the open, I’m hoping folks will correct me in the comments when I get stuff wrong.
• I did not get any input from any of the folks working in these civic tech organizations. Indeed, leaders in some of these orgs might even see this post and say, “Oh, bless his heart…” or “You sweet, summer child…” and I would deserve that! Indeed, I might deserve worse. 🙂
• While my attendance at the CfA conference was funded by my employer, my comments are entirely my own, and this post was developed on my own time. So this is just me.

Some #CfASummit wins and fails

I’ll get to the “big ideas” in a moment, but to start I want to at least offer some examples of what’s not working (fails) and what is working (wins) in the current CfA conference model.

I also want to call out the unsung heroes of this and other conferences—the organizers. Whatever criticisms I may have, I want the organizers to know I see their work and appreciate it. It takes a lot of work to execute any conference, especially a national one with hundreds of attendees. I deeply respect that work. Event planning is hard, and it’s almost entirely thankless. If you haven’t worked on creating or hosting an event of any size… you have no idea. As I love to say, everybody wants to go to the party, but nobody wants to throw the party.

If anything, I’ve spent the time to build this assessment and these recommendations because I can sense what the CfA Summit could be with some changes. My intent is to call out the latent potential value that some changes could unlock.

So let’s get into it…

High-level concerns with #CfASummit

• There are too many “mainstage” sessions—4 of them over 2 days—where all attendees are pulled together to see a program with mixed applicability for any given attendee. Conference “keynote” sessions are where you set tone and theme for a conference, but really you only do them at the start and the end, not multiple times daily. That’s wasted time that could be spent on sessions or networking.
  • Additionally, the mainstage sessions were over-produced, including excessively scripted hosts, pre-recorded “morning zoo” announcements to move the program along, and an obvious talk show staging I suspect was designed to make people look famous and influential rather than substantive. (Charlie Rose was a substantive program, and all they had was a round wood table on an all-black set.)
• The schedule was uneven. Some sessions were too short. The breaks were sometimes there, sometimes not. And compared to the best conferences I’ve attended, the day started a bit late and ended a bit early. To be fair, there’s an interplay of venue, schedule, content, and attendee population that’s complex and hard to get right. It’s close to right at CfA, but not entirely. The schedule is where value really gets unlocked, and it needs a refit to give more space to presenters in content tracks.
• Vendors are strangely included and excluded at the same time. I love how CfA does not have a vendor space with booths and so forth, and it’s not awash in vendor branding or pitches. Yet vendors still ended up in some sessions, and there were some vendor-sponsored side events. I also had one vendor’s sales rep pester me ahead of the conference. So… which is it? Vendors in, or vendors out? Let’s pick one.
  • And as I will note below, Code for America (the organization) is itself a vendor, so… that’s awkward.
• The awards handed out at the conference felt like afterthoughts. Rather than being a rallying cry moment, celebrating civic tech work, they appear to be denouements to some kind of hidden award process that didn’t involve attendees. What could be a celebration is just confusing and immediately forgotten. Just minutes after the award presentation, I couldn’t tell you who won or why. That’s not good.
• There’s no clear feedback loop at the end of the conference. The best conferences I’ve attended in my career included live feedback sessions to capture ideas for future improvements.
• I have some other concerns around basics like name badges, wayfinding, swag, AV systems, and the ratio of attendees to space, but those are comparatively minor and are concerns any conference must address.
• The conference is not clearly demographically layered. The needs and interests of (a) different government strata—federal, state, local—are similar, but not identical, as are (b) the interests of different professions in attendance, and (c) the needs of different experience levels in attendance. The structure and content of the conference didn’t promote clear “tracks” for these layers and verticals. This is likely a sizing problem for the conference—it either needs to be a lot bigger, allowing for specialization, or a lot smaller, focusing on a narrower theme or audience.
• The inclusion of headline-grabbing celebrities at the conference—Kara Swisher in 2024, Chef José Andrés in 2025—was unnecessary at the very least. Those folks are not practitioners in our field and while they may be entertaining or inspirational, they are not part of our world. Hearing about the work of Terrance Smith in the opening session this year was more inspirational and relevant than Swisher and Andrés combined.
• Finally, Code for America (the organization) isn’t a pure-play nonprofit advocacy or capacity-building organization any longer. They are now a government contractor, too. While CfA’s intentions are clearly altruistic, this structural fact is… problematic. I believe it’s time for CfA to take a “spiritual leadership” role with the Summit, but step back from actually owning the whole thing.

What #CfASummit gets right

• There are too few national conferences around civic tech, and this is one of them. That sounds basic, but it’s the most important thing CfA gets right—it’s available every year, despite the strategic changes inside Code for America (the organization). Honestly, given those strategic changes, I’m not sure I would have kept the conference if I were leading CfA, so I appreciate that they have so far. Having the conference at all leads to a unifying feeling, emotional attachment, national mindshare… it’s powerful and valuable, and CfA deserves thanks for keeping it going.
• The rock-solid consistent visual design of the conference is high-quality and ubiquitous. That’s hard to pull off, but they do it well, right down to the standardized slide decks, the venue decorations, wayfinding, name badges, website… the works. There’s no doubt which conference you’re attending.
• They’ve assembled a solid annual process for content submissions and reviews. This is also hard to pull off because you need a lot of volunteer labor and folks have to learn / re-learn how to do this work each year.
• Using different locations in alternating years is a great idea, and using the two “seats of power” for government (DC) and tech (Silicon Valley) is thematically smart. (Although next year it’s in Chicago for some reason.)
• The venues are good enough without feeling wasteful or extravagant. CfA has a “premium economy” sensibility that feels right for the audience and purpose.

The dream team for a future conference

Okay, so if the “value proposition” is a little off—in my estimation, of course—then what could we do to fix that? How would we get started?

First, I would assemble a dream team to develop and drive a new national civic tech conference. We’ll need to tap a few players:

• Branding: Code for America
• Design: Civilla
• Organization + Convening: The Beeck Center
• Vendor Management + Additional Support: U.S. Digital Response

I also want to call out Technologists for the Public Good here. They are doing great work, too, and if there's a role they would like to play in all this, I would be delighted to see them on board. They're smart, committed, and have built the only truly active online community to keep something like a "virtual summit" going all year long.

What do I mean by all this? A bit of explanation…

Branding: Code for America

The history of CfA and the Summit is nothing short of legendary. And their control over the graphic design and identity elements is excellent. Code for America—the organization—has changed strategies a lot over the years, and now focuses on a different set of priorities than in the past, but the meaning of the Summit branding is solid and should be retained if at all possible. So ideally we keep that. But the CfA organization shouldn’t be burdened with running the whole conference—it’s not their core mission. I only ask that they donate the branding, their blessing, and that they actively participate each year.

Design: Civilla

If you haven’t done the tour at Civilla’s Detroit office, you are missing out. It’s a literally visceral experience that takes the concepts of human-centered design and turns them into a palpable encounter that takes all the big words and ideas of civic tech and turns them into a confrontation with the realities of the people we ostensibly serve. I am not exaggerating. It is stunning.

Civilla was born as a design shop and lives it, in the biggest sense of that word, and they should be in charge of the design for the conference itself. That includes theme, scope, schedule, venue selection, space prep, content goals, the opening and closing sessions… the overall “thing” that is the conference.

Civilla, however, is a small shop, so they will need help in execution, which is where the next two players come in.

Organization + Convening: The Beeck Center

The Beeck Center has been a tireless booster of civic tech for years and their events and online resources are second to none. They innovate, they lead, and they are committed to the mission in a pure, nonprofit, think-tank kind of way. They should run the new conference, using the design developed with Civilla. Beeck has a deep reservoir of Georgetown student talent (super-smart, super-committed kids), they have DC connections, and they have all the right (non-conflicted) intentions.

Indeed, the best conference I’ve attended in years—rivaling my experiences in public broadcasting—was hosted by Beeck in early 2024. And their work on FormFest has been a raging success, in collaboration with Code for America.

Vendor Management + Additional Support: U.S. Digital Response

If there’s one thing USDR does well, it’s pulling together disparate groups and getting them to work together for a greater good. They do that all day, every day. And if we’re going to have vendors at the new conference—and I think we should—I would trust USDR to find the right collaborative balance between vendors and attendees. Additionally, USDR has a wellspring of volunteer talent they can tap into and use for any other kinds of support the conference may need. USDR and Beeck already work well together, so let’s amp up that collaboration and focus it to build the back-end of the new conference.

If we put together USDR and Beeck and Civilla (and maybe TPG), they can share the load, share the ideas, and build something that honors the Code for America legacy but builds something new, powerful, and exciting for our national civic tech community.

Decisions needed for a new CfA Summit

To build a new Code for America Summit, there are a few key problems to tackle and decisions to make. This is why careful design up front would be so important. Blending goals, scope, themes, venue, and audience feels nearly impossible! A few considerations…

What is the underlying theme or goal?

Why are we coming together? Civic tech is a big space with a lot of players and intentions and methods. People need to know why they would attend and what they will get out of it. Is it capacity-building, like skills training (UX research, coding, project management) or is it mission reinforcement around thematic goals like human-centered design, trust in government, digital service? Is the conference a series of victory laps, showing off successes in the field via presentations and awards—so folks know what to aim for in their own operations? Is it connecting people in similar roles or at similar layers of government? Maybe it’s a bit of all of these, but ultimately there must be a unifying strategy at the heart of this event so organizers know what to say “yes” to and what to say “no” to and attendees can figure out who needs to be there.

How do we address the complex demographics?

This is probably the hardest nut to crack. Government has State, federal, and local layers. There are also multiple professional tracks in digital services. Plus, attendees have career arcs spanning from novice to expert, not to mention variable experience levels in government itself. To satisfy all those needs fully we would need a huge conference, which probably isn’t realistic. So conference organizers would need to make some choices. What’s happening at the CfA Summit now is a well-meaning melange of all of the above, which makes it less valuable to each audience.

Knowing who should be there and what they need will help set the scope and scale of the conference and allow for targeting the right people to attend.

Are we including vendors or not?

If vendors are out, that makes for a “pure” practitioner conference, and that may be attractive for a lot of reasons. But that changes the financial picture for the conference and it also ignores the reality that software and services vendors play a role in our ecosystem (and should be encouraged to act more like partners). But if we’re including vendors, what’s the right level of participation? I want to be supported by vendors, not preyed upon. One way or another, we need a clear plan here because the current approach is… confused, at best.

Conference mechanics

Once the themes and boundaries are set, there are “mechanical” decisions to be made, like venue, target price for attendees, food, AV systems, site wayfinding, swag, and so forth. The conference also needs a clear process for drawing content from volunteers. And it needs a clear schedule—how many days, how long is each day, how long is each session and how does the length of session interplay with the content of the session? You get the idea.

So what’s next?

Honestly? Probably nothing. Maybe some folks out there will feel I’m calling the #CfASummit baby ugly and will not react well to it. I might become “that jerk that hates CfA.” But I’m not calling the baby ugly. I want the baby to undergo a growth spurt.

In any case, let’s be clear: I’m not in charge of anything. I’ve only been to two CfA Summits. I’ve only been in government for 6 years. I don’t work for these organizations. So what do I know? It would be easy to ignore my comments (and perhaps you should!).

Maybe nothing happens next, and we get the same CfA Summit again next year. That would not be awful. It would be fine.

But if there’s any desire out there to build a better, more impactful national civic tech conference, y’all are gonna have to speak up. Drop a note to the folks you know at CfA or Beeck or USDR and suggest a re-think. Add your voice to the conversation on LinkedIn or Substack, WordPress, Bluesky, etc. Shoot… there may even be folks at Code for America (the organization) that would be delighted to let this conference responsibility fall to a new team, so they can focus their energies on more strategic work. Maybe they can start a conversation privately amongst their nonprofit peers to see what’s possible.

So I don’t know what happens next. But I’ve shared my thoughts.

What are yours?