Lessons from Civic Tech History
What happened to Brigade, EveryBlock, Sunlight Foundation and 37 other civic platforms. Structured post-mortems, full sources, and the recurring patterns behind why civic tech keeps failing.
Common questions
- Why do civic tech projects fail?
- Across the cases catalogued here, failure is rarely about the technology. Eight patterns recur: no durable payer; a cold start on the side of government that never arrives; engagement that displaces influence; transparency mistaken for participation; capture by one political side; survival bought by narrowing the original ambition; real participation that no government is obliged to act on; and absorption into a govtech vendor once the buyer is a procurement office rather than the public.
- What is the civic tech graveyard?
- A term for the many civic platforms that raised money, gained users and press, and no longer exist. The Civic Tech Field Guide maintains the best-known catalogue of them. This surface is a companion to it: rather than a directory, it is a set of structured teardowns of individual platforms, each with a dated fact panel, a named failure pattern, and its sources listed.
- Did any civic tech platforms survive?
- Several, and how they survived is the more useful question. The recurring route is narrowing: a project endures by trading the broad democratic ambition for a specific, fundable service a government will buy. SeeClickFix, Granicus and CivicPlus are catalogued here as examples of that trade.
The front door
51 cases, eight recurring shapes. Start with a pattern to see the evidence.
Every case we've studied
51 platforms. Filter by pattern, outcome, or era.
Americans Elect
A $35 million bet that a hidden centrist majority would flock online to nominate a third presidential ticket, undone when neither the voters nor the candidates it imagined ever showed up.
◆ The supply-side cold start2010–2012Shut downHigh confidenceAmericaSpeaks
The gold standard of large-scale in-person deliberation, which convened 180,000 citizens over two decades and closed the moment the grants that paid for it dried up.
◆ The business-model void1995–2014Shut downHigh confidenceBang the Table
The Australian pioneer of online community engagement, whose EngagementHQ reached 17 million people — and which ended, fittingly for this era, absorbed into the govtech giant Granicus.
◆ The vendor endgame2007–2021Acquired / absorbedHigh confidenceBetter Reykjavík
Iceland's post-crash participation platform that actually wired citizen ideas into monthly council decisions, and showed how much that wiring depends on a government that keeps choosing to listen.
◆ The advisory ceiling2010–presentSurvived (narrowed)High confidenceBlockboard
A neighborhood app for reporting potholes and watching out for your block that got bought by Klout, a company built to score social-media influence, and functionally ended as a civic tool the day the acquisition closed.
◆ The business-model void2011–2012Acquired / absorbedMedium confidenceBrigade
A lavishly funded "civic network" that tried to make taking political positions social, and found neither the scale nor the revenue to last.
◆ The business-model void2014–2019Acquired / absorbedHigh confidenceCauses
The first Facebook advocacy app, which signed up more than 180 million people to "causes" and then discovered that a click is not a constituency.
◆ The engagement trap2007–2014Acquired / absorbedHigh confidenceChange.org
The world's largest petition platform, where hundreds of millions of signatures show reach without reliably showing power, and where a for-profit built on a progressive base struggled either to be neutral or to be paid.
◆ The engagement trap2007–2021Still operatingHigh confidenceCitizinvestor
A "Kickstarter for civics" that let residents crowdfund public projects with municipal buy-in, and shut down in 2018 because it never worked out who its real paying customer was.
◆ The business-model void2012–2018Shut downHigh confidenceCivicPlus
The company behind thousands of city-hall websites, which bought up the citizen-facing tools around it — including SeeClickFix — turning the 'digital front door to government' into a vendor product line.
◆ The vendor endgame1998–presentSurvived / thrivingHigh confidenceCountable
The bill-tracking civic app that never found anyone willing to pay for civic engagement, so it took its engagement engine to Starbucks and Uber and left politics behind.
◆ The narrowing survivor2013–2020Pivoted awayHigh confidenceDecide Madrid
The most celebrated citizen-proposal platform of its generation, where hundreds of thousands proposed and almost nothing passed, and where a change of government showed how little of it was really theirs.
◆ The advisory ceiling2015–2019Survived (narrowed)High confidenceDecidim
The open-source participation platform that outgrew the city that built it, governed as a democratic commons — and still runs into the wall every participation tool meets: it can gather a proposal but cannot make a government act on it.
◆ The advisory ceiling2016–presentSurvived / thrivingHigh confidenceDemocracyOS
Argentina's open-source direct-democracy platform that tried to make legislators vote the way citizens clicked, and when it could not get the legislature to listen, ran for the seats itself and lost.
◆ The supply-side cold start2012–2015DormantHigh confidenceElectNext
A candidate-matching "eHarmony for voters," built by two Penn students, that pivoted twice chasing revenue, first to selling political data to media outlets, then to sponsored comments, before being folded into Change.org.
◆ The business-model void2011–presentAcquired / absorbedMedium confidenceEveryBlock
A pioneering "news feed for your block" that told you everything happening around you and gave you nothing to do about it, twice bought and twice switched off by owners who could not make it pay.
◆ Transparency is not participation2007–2013Acquired, then shut downHigh confidenceFixMyStreet
Britain's durable street-fault reporting map, a nonprofit success that lasted by doing the small, fundable thing, reporting the pothole, rather than the large civic one.
◆ The narrowing survivor2007–presentSurvived (narrowed)High confidenceg0v
Taiwan's leaderless civic-hacker community that 'forked the government,' seeded vTaiwan and Cofacts, and survived for over a decade precisely because it never became a company or a product — with all the fragility that model carries.
◆ The narrowing survivor2012–presentSurvived / thrivingMedium confidenceGo Vocal
The venture-backed 'citizen' engagement startup that scaled to 500+ governments — and then dropped the word 'citizen' from its name, quietly telling you who the real customer had become.
◆ The vendor endgame2015–2024Survived / thrivingHigh confidenceGoGov
The 311-app maker that gives hundreds of small towns a branded 'report a pothole' button — the humble, sustainable end state of the citizen-engagement dream: a service line, not a voice.
◆ The vendor endgame2016–presentSurvived (narrowed)Low confidenceGoogle Moderator
Google's crowdsourced question-voting tool, which powered Obama's online town halls and showed how fast an organized minority can vote its pet issue to the top, before Google switched it off for the whole civic world that had come to depend on it.
◆ The engagement trap2008–2015Shut downHigh confidenceGovPilot
A digital-government platform that replaced city hall's paper forms with 125 tidy workflows — a genuine upgrade for administrators that has almost nothing to do with a citizen being heard.
◆ The vendor endgame2015–presentSurvived (narrowed)Medium confidenceGranicus
The company that started by streaming city-council meetings and grew, acquisition by acquisition, into the govtech giant that now owns much of the plumbing between governments and the public — mostly one-way.
◆ The vendor endgame1999–presentSurvived / thrivingHigh confidenceHotSoup
A bipartisan political debate site built by an actual roster of Bush and Clinton-Gore campaign veterans, edited by a former AP chief political writer, that folded within a year for the plainest possible reason: nobody advertised on it and nobody kept coming back.
◆ The business-model void2006–2007Shut downHigh confidenceInterTwinkles
An MIT dissertation project turned open-source toolkit for small-group consensus decisions, undone not by funding or users but by a single third-party login system it depended on being discontinued out from under it.
◆ The business-model void2011–2016Shut downMedium confidenceJumo
Chris Hughes's nonprofit "social network for social causes," which drew a million users and 15,000 nonprofits in weeks, watched traffic collapse by 93% within six months, and was folded into a media company a year after launch.
◆ The engagement trap2010–2011Acquired / absorbedHigh confidenceLiquidFeedback
The German Pirate Party's liquid-democracy platform, where the power everyone was handed flowed to a handful of "super-voters" and the participation drained away almost as fast as it arrived.
◆ The engagement trap2009–2013DormantHigh confidenceLoomio
A worker-owned deliberation tool born from Occupy that kept its values by staying small, and found that inclusive online decision-making lives inside organized groups, not open civic populations.
◆ The narrowing survivor2011–presentSurvived (narrowed)High confidenceLouder (Loudsauce)
A crowdfunding platform that let small donors pool money to buy political and social-cause advertising, backed by the Knight Foundation and real investors, rebranded once, and was ultimately acquired by Change.org.
◆ The business-model void2009–2015Acquired / absorbedHigh confidenceMapzen
An open-source mapping infrastructure company wholly backed by Samsung, whose tools are still used today, and which closed in 2018 with no explanation the moment its one corporate backer's interest moved on.
◆ The business-model void2013–presentShut downMedium confidenceMindMixer
A well-funded "virtual town hall" that raised more than $23M to gather residents ideas, then found the paying product was selling city hall its own data, and became a government analytics company.
◆ The narrowing survivor2010–2015Pivoted awayMedium confidenceMyGov
India's vast government "citizen engagement" portal, where tens of millions submit ideas into a channel that no ministry is required to answer, and the conversation stays one-way.
◆ The advisory ceiling2014–presentStill operatingHigh confidenceNationBuilder
Software built on the promise that any citizen could now run a campaign like a professional, which thrived by becoming the shared, sell-to-all-sides infrastructure of the professionals instead.
◆ The narrowing survivor2009–presentSurvived / thrivingMedium confidenceNeighborland
A beloved civic-input tool grown from a New Orleans art project, used by millions and hundreds of institutions, that was bought by a neighborhood social network and quietly switched off.
◆ The business-model void2010–2019Acquired, then shut downMedium confidenceOpenCongress
A nonpartisan hub for tracking bills and contacting Congress, built the exact thing a later critique said the field never needed, and lost not to failure but to free government tools finally catching up.
◆ The business-model void2007–2016Shut downHigh confidenceOpenGov
The Silicon Valley company that set out to make government finances transparent to citizens and grew into a $1.8-billion back-office ERP — the transparency dream matured into accounting software the public rarely sees.
◆ The vendor endgame2012–2024Survived / thrivingHigh confidenceOpenGov Foundation
A nonprofit that built a genuinely elegant tool for citizens to mark up legislation, and spent years discovering that legislators would praise it, pilot it, and then not use it.
◆ The supply-side cold start2012–2019Shut downMedium confidenceParticipateNYC
New York City's official participation platform, where residents get a real vote over a symbolic slice of the budget, and the harder question of power over the rest goes untouched.
◆ The advisory ceiling2022–presentStill operatingMedium confidencePledgeBank
A mySociety tool that let anyone say "I will do X if N other people will too," which quietly predated Kickstarter and Groupon by years and was retired in 2015 because the resources it needed had to go to more-used projects instead.
◆ The business-model void2005–2015Shut downHigh confidencePolis
The opinion-mapping tool that can find the hidden consensus in a crowd of thousands — and then hands that consensus to a government that is under no obligation to do anything with it.
◆ The advisory ceiling2012–presentSurvived (narrowed)High confidencePOPVOX
A scrupulously neutral marketplace for telling Congress what you think about a bill, which delivered millions of messages but never found a citizen who would pay, and became legislative software for institutions.
◆ The business-model void2010–2021Pivoted awayMedium confidenceRousseau
The Five Star Movement's "direct democracy" platform, where a shrinking minority cast Yes/No votes on pre-cooked questions while a single unelected owner held the data, the fees, and the real control, until the movement walked away.
◆ Partisan capture is terminal2016–2021DormantHigh confidenceRuck.us
A political social network built by two former state-party operatives' sons to organize engagement around issues instead of parties, which drew real press and funding, then quietly became a campaign fundraising toolkit once the network itself never caught on.
◆ The supply-side cold start2011–2014Pivoted awayMedium confidenceSeeClickFix
The 311-style issue-reporting platform that survived where flashier peers died, by selling to city halls and narrowing from "fix democracy" to "fix the pothole."
◆ The narrowing survivor2008–2019Acquired / absorbedHigh confidenceSunlight Foundation
The flagship of the "sunlight is the best disinfectant" movement, which spent fifteen years and brilliant engineering making government legible, and proved that legibility is not the same as a public able to act.
◆ Transparency is not participation2006–2020Shut downHigh confidenceTextizen
A clever Code for America SMS tool that let cities reach residents by text, which grew to forty governments and then dissolved into a big government-communications vendor after two years on its own.
◆ The business-model void2012–2015Acquired / absorbedMedium confidenceThe People's Money (NYC)
New York's first citywide participatory budget, where 140,000 residents voted to allocate a few million dollars out of a $100-billion budget — real participation over a rounding error.
◆ The advisory ceiling2022–presentStill operatingMedium confidenceVoter.com
A lavish dot-com political portal, named the best on the web and edited by a Watergate legend, whose traffic lived and died with the 2000 election and whose business died with the bubble.
◆ The business-model void1999–2001Shut downLow confidenceVotizen
A verified-voter social network that made sure every voice was a real constituent, and learned that the verification which made the signal trustworthy also kept the network too small to matter.
◆ The supply-side cold start2010–2013Acquired / absorbedHigh confidencevTaiwan
Taiwan's admired experiment in crowdsourcing law by consensus, which worked while it had a champion and the novelty held, then stalled once its recommendations proved to be, in a co-creator's words, a tiger without teeth.
◆ The advisory ceiling2014–2018Survived (narrowed)High confidenceWe the People
The US federal petition platform whose entire promise was a response, which raised the bar as the public showed up, gave boilerplate when it answered at all, and went silent when the administration changed.
◆ The advisory ceiling2011–2017DormantHigh confidence
Take the whole thing as data
Every case here is available as structured JSON and CSV, openly licensed, with the sources attached. No signup and no key.
The dataset →