We (still) need to do better at failing in civic tech
How and why do civic tech projects die?
Over the past three years of studying the refugee tech space, one of the dominant themes has been a sense of decline. Dozens of projects have expired — some with a decisive cut, many with a slow fade into oblivion.
One challenge with this is methodological. We want to track the fortunes of more than 100 such projects we know about. A website might still be online, but not especially active. Maybe it was given up on some time ago, or maybe it’s still drawing laboured breath. It can be hard to know.
But the more significant challenge is how to interpret or react to this mass demise. We’re clearly looking at the inevitable and natural return to a more stable equilibrium following the extraordinary explosion in activity in 2015–16. Some projects served a specific and transient need — like Volunteer Planner, which coordinated the thousands of people wanting to volunteer at refugee shelters — and once the moment has passed, it makes sense to become dormant.
But at the same time, not all the factors are so systemic — some of the deceased were just not good projects, whether based on misguided ideas, or poorly executed, or both.
In short, a lot of our work over the past three years has been that of a forensic pathologist: is this project alive or dead — and if it’s dead, how and why?
This is valuable information. The more we collectively understand which approaches succeed and why, the more intelligent and impactful the system as a whole will be. And details about unsuccessful attempts are at least as valuable as the (often rose-tinted) success stories which active projects like to tell about themselves.
When I joined betterplace lab back in 2013, my first weeks were spent working on the Trendreport, in which we described new ideas percolating in the civic tech space, and civil society more broadly. One of the earliest trends was entitled “Produktiv scheitern”, which might be translated as “fail forward”. The idea was that there was a growing acceptance of the value and importance of being vocal about things that don’t work well.
If only. Six years on, I can see precious little appetite for embracing (some) failure as a necessary part of experimenting and learning. The initiatives which that trend highlighted such as Admitting Failure seem to have become largely dormant. And in the refugee tech space, almost all the opportunity for discontinued projects to publicly share what they learned has been squandered. The stand-out exception to this is clarat, who deserve serious praise for this blog-post frankly reflecting on their journey.
A part of the reason is psychological. Failure is for most people closely entwined with feelings of sadness and shame, which we often expend a lot of energy trying to push away or hide. On a human level, it’s little wonder people shy away from speaking up about experiences that might feel painful or embarrassing.
But there are other reasons that are more cultural and structural. The world of commercial tech start-ups has been very successful in promoting innovation by defusing the taboo around failure. It’s commonly understood among start-ups and investors alike that nobody knows what will ultimately succeed, and initial funding is a licence to give something a try, not an implied demand for guaranteed success.
This is the mindset that’s still lacking in the social sector (the only example that comes to mind is the Prototype Fund). Recent years has seen the steady advance of impact measurement and trying to objectively demonstrate and quantify how much a programme achieved. On its own terms, there’s much to be said for this. But my sense is that so far, it’s led towards a greater conservatism, in the form of demanding proven impact before funding is awarded. This might be good for optimising the paradigms we’re currently in, but to get us to brand new paradigms and more transformational change, we will need something different.
And a part of that will be learning to fail better.
