One of the best things about a room full of friends is the potential for ruthless honesty. And honesty is exactly what happened last week at the Code for All summit in New York City, at Personal Democracy’s recently opened Civic Hall. The intimate setting provided one of the most upfront conversations in the overcrowded history of civic tech panels: a discussion about how, when and why civic tech doesn’t work.

In a small room full of friends, talking about failure should be easy. But for some reason — maybe because of the relative novelty of using more expensive technologies for social innovation — people working around civic technology are not used to admitting when their projects don’t work. And while this sort of dishonesty might help with short-term opportunities (especially when it comes to funding), it has a serious effect on long-term sustainability: We cannot learn from our mistakes.

Today, we’re asking you to be honest with us, and to submit your failure stories as well. Tell us about civic tech projects or ideas that might not have worked out through this short survey. If you wish to remain anonymous, that’s all right too — just let us know, and we will only use your failure story for more general conclusions.

Luckily, though, an increasing number of organizations working around civic technologies dare to admit publicly if some of their ideas have proven unsuccessful (or maybe were just poorly timed). The ever-inspiring Tom Steinberg, mySociety’s outgoing director, has recently announced that the Popit component of Poplus is not succeeding in its mission of making structured data on politicians more widely available. Therefore, the organization will put development resources into other projects that are about producing data on politicians, specifically YourNextRepresentative and EveryPolitician.

Detective.io, the collaborative network visualization and analysis tool from Journalism++, which was supposed to reform data-driven investigations, is also retiring soon; the platform does not have enough paying customers to become profitable, and most of its targeted users are not sufficiently enthusiastic about open source software.

At Sunlight, we also have quite a lot of retired projects, some of which are tools that did not work out the way they were intended to do. We’ve written about the process and reasons behind sunsetting projects, too.

Larger software organizations and nonprofits are, however, in a relatively good position when it comes to admitting failures, or shifting the focus of resources. A national-level advocacy organization with more limited funding opportunities and tech capacities might not have the luxury to evaluate failure in such an honest manner, which is why learning from each other`s mistakes gets even more crucial.

We have long been thinking about what works and what doesn’t work in civic tech, but our emphasis so far has mostly been on the success stories. That’s why we’re trying to get the other side of the story with our short survey — to paint a more complete picture of the true civic tech landscape.

Too much hype around failures would be another mistake, though. Some projects do not work in a specific environment and in a specific time, but the very same idea might flourish in a different context or under new circumstances — maybe even a few miles or years away.

And remember: