Why Civic Tech Keeps Failing, and What We Learned

Before we built much of anything, we spent time studying the civic-tech projects that came before us and did not last. We keep an internal catalogue of them: the platforms that set out to reconnect citizens and government, and the specific way each one came apart. It is a sobering read, and it is the most useful thing we have done. Two failures in particular sit next to each other in the file, because they fail in opposite directions, and between them they explain a lot about why this field is littered with good intentions.

The short version: one died rich, and one survived by becoming smaller than its mission. Both outcomes are traps, and avoiding them shaped how we are building Agora.

The one that died rich

Brigade launched in 2015, financed largely by Sean Parker, the founding president of Facebook, with an estimated forty to fifty million dollars behind it. The bet was that American democracy was missing its own social network, a place to take positions on issues, see where your friends and candidates stood, and organize around it. The team was strong, the mission was genuinely non-partisan, and the 2016 ballot guide reached roughly 250,000 people and generated about five million get-out-the-vote messages.

It wound down in 2019. The engineering team went to Pinterest; the technology and data went to a company called Countable. The reason is the one that haunts this whole field: there was no business model underneath the mission. You cannot charge citizens for participating without betraying the point, and no other paying customer was ever found, so the company ran on its founder's capital instead of revenue until both the money and the patience ran out. A product problem made it worse. When Brigade opened up its debates, its own CEO later said the change "empowered the loudest and most aggressive voices," which drove substance down and drove participants away.

The one that survived by shrinking

SeeClickFix is the opposite story. It started in New Haven in 2008, after its co-founder couldn't get the city to remove graffiti from his block. The idea was small and concrete: let residents report potholes, broken streetlights, and graffiti, route each report to the right department, and tell the resident when it gets fixed. Citizens used it for free; cities paid for the dashboard that managed the queue. On only about $3.4 million raised, a fraction of what Brigade burned, it reached more than 200 governments and was acquired by a larger public-sector software company, CivicPlus, in 2019, with the product still running today.

By startup standards, that is a win, and it deserves to be called one. But notice what it cost. The product that pays for itself is transactional 311 reporting, which is a far narrower thing than the civic reconnection the field set out to build. The paying customer is the government, which steadily pulls the tool toward operational service-management, and the citizen's "participation" shrinks to filing a request and waiting for a status update. The broad ambition narrowed to fit what a city would buy.

The two traps, named

Put side by side, these are the field's recurring ways to fail. The first is the business-model void: a mission with no one to pay for it, kept alive by venture capital until the capital tires. The second is the narrowing survivor: a project that endures precisely by giving up the hard, broad ambition and settling into a fundable niche. One runs out of money. The other keeps the money and loses the plot. Most civic-tech stories we catalogued are a version of one or the other.

What this changed for us

Seeing both failures clearly is the reason Agora is built the way it is. We decided who pays before scaling, rather than after: participation stays free, and the product we sell is the analysis built on top of it, structured inside a public benefit corporation so the mission is harder to trade away later. We refuse engagement-maximization and open citizen broadcast, because Brigade's debate pivot showed where that leads. And we are honest that SeeClickFix is the more seductive danger for us, not the more obvious one. The risk is not that we flame out like the well-funded consumer app. It is that we succeed, quietly narrow, and wake up one day as a service-request tool that forgot it was supposed to be about participation between elections.

We will add more cases to the catalogue as we go, and some of them will probably make us rethink things we feel sure about now. That is the point of keeping it. So the question we keep asking ourselves is the one the whole field seems to flunk: how do you stay funded without becoming small, and stay ambitious without running out of road?