Learning from failure: FAILfare

Katrin Verclas from Mobileactive.org writes on what I have for years strongly advocated in Sri Lanka – openly discussing failures so that we may learn from them. As the email promoting the event, called FAILfare, notes,

While we often focus on highlighting successes and gains in our industry, it’s no secret that many projects just don’t work – some aren’t scalable, some aren’t sustainable, some can’t get around bureaucratic hoops, and many fail due to completely unanticipated barriers. But a failure is no reason to be ashamed! Instead, FAILfaire looks to bring together a collection of failures so that we can learn from what hasn’t worked in the past in order to make our future projects stronger and better.

It’s not easy to talk about failure. Whenever I mention the groundbreaking technology adaptation behind the OneText initiative I helped design in Sri Lanka to support the peace process, I also noted that the peace process was a dismal failure and resulted in even more brutal war. Going into the reasons why ICTs didn’t change the dynamics of inter-party negotiations, and how partisan politics overwhelmed the best virtual interactions is a lesson in how ICTs can be leveraged for similar, fragile and vexed processes in the future.

Katrin’s FAILsafe initiative is anchored to the use of mobiles in development. But failure is an integral part of ultimate success in other ICT fields as well, including citizen journalism. I have already written about the failure of the first citizen journalism initiative I was involved in, Moju, leading to what is today the success of Groundviews. This is echoed by Sunlight Foundation’s Ellen Miller in an interview recently published on the Nieman Journalism Lab’s blog. Her submission is worth quoting at length,

I think the ability to fail is absolutely part of the culture in which we live. And so, someone will try lots of things, which you know sometimes just don’t work — but because we don’t know how people want to engage with, you know, either fairly wonky information about legislation or critical information, if we don’t build it, we never give people the opportunity to test it. And some of the things have worked far beyond — much better than — what we expected. And some of the websites just weren’t popular, and we couldn’t quite figure out why, and we said, “Oh, they weren’t popular, let’s just take it down.” So there’s not much cost to the experimentation. But partly I think it’s because you have sort of a new and largely successful of the project because Sunlight, you know, is an institution without any legacies. It’s just — it’s really built into the DNA. But it’s something major other institutions, you know, have to work on. Now you can’t really build it into the DNA of reporting a story: Failing, getting the facts wrong, telling the story that’s wrong. But there are certainly elements in terms of engaging citizens, in getting them to tell the stories that work. So, I mean, if it doesn’t work this time, you know, you might try it again or — or not. But we’re beginning to learn.

So one of our examples was — it was successful, but it was a failure in the end. We did a series of distributed research projects in the early days. We do one investigating members of Congress’ spouses, and whether they were employed by their campaigns. And then we did another one on getting people to contribute to a database on earmark requests when they started posting them. And then we realized that if people who worked on Project A, we had no idea if they’ve been secretly working on project B, or who worked on Project C. We said, “Wow, let’s stop that.” We created one platform, Transparency Corps, so that anybody who worked on A or B or C had the opportunity to see what was D, E, and F coming down the road, to begin to build more of a community. Because if you’re interested in these kinds of distributive projects, you’ll be interested in, you know, any number of them, and you get deeper engagement in them. So it worked in the individual pieces, but we knew we were losing these people because we didn’t know quite how to reach out to them again. So I think it’s that experimentation or constantly, constantly iterating on something that worked or that didn’t work until you find things that work.

It’s not easy to talk about failure because doing so risks future support for initiatives outside a model strictly entrepreneurial or philanthropical. Donors and funders, with their log frames and results based management frameworks, rarely demonstrate patience with projects that fail, blacklisting those individuals and organisations that do against future funding. This creates a vicious cycle, where openness about failures is subsumed by manufactured success, leading to an essential dishonesty that is bad for everyone. There could of course be managerial, organisational, financial or other valid reason for donors to chastise project failure. There could also be contextual, processual challenges that only came about after the project and the introduction of ICTs into the mix. Sometimes innovation takes years to be recognised as bringing in value to a process, set of actors or particular context. On other occasions, unintended consequences of a project can outweigh the benefits initially planned for and expected. In sum, thinking robustly about the reasons for failure, and openly sharing lessons learnt can be a valuable exercise. Framed in a supportive, non-judgemental environment and facilitated well, FAILsafe’s findings and case studies can buck the trend of publishing only that which is successful.

I have always found failure instructive. It is good to see others recognise this. For as Oscar Wilde noted, experience is simply the name we give our mistakes.

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s