You are not logged in. LOG IN NOW >

Ushahidi and the Long Tail of Mapping for Social Change

BY David Eaves | Monday, July 9 2012

What makes a mapping project successful? Image: Trafficking map Epawa SMS

A new website called DeadUshahidi launched recently with the express purpose of tracking Ushahidi mapping projects that experienced little use. While the Ushahidi team responded in good form, but it was hard not to see the website as a shot across its bow.

Ushahidi, which has become a darling of the social tech world, first became famous for mapping violence during the 2007 Kenyan election, after which a Kennedy School of Government review found the mapping platform had reported acts of violence more accurately than traditional media. It was used again to great effect after the earthquake in Haiti when over 15,000 reports were analyzed and 3,500 events were mapped. This helped some stakeholders decide where to deploy resources and target their efforts. A number of these high profile cases have been written about and analyzed by academics. Less well documented are the Ushahidi maps that don’t go on to become large projects. This is where DeadUshahidi comes in.

To qualify for DeadUshahidi, lucky sites have to meet one of the following criteria:

  • No one (let alone a crowd) has submitted a report to the map in the last 12 months.
  • For time-bound events, like elections and disasters, the number of reports is so infinitesimally small in relation to the affected community that the map never reached a point anywhere near relevance. (Our measure for elections is, for instance, # of submissions / # of registered voters > .0001).
  • The map was never actually started (there are no category descriptions or fewer than 10 reports). DOA, anyone?
  • The map is listed in the Ushahidi deployment list but the link to the map is dead. That definitely warrants a death certificate.

The thing about DeadUshahidi is that it makes a good point — one that we’ve made on techPresident before. If you want your project to be a success you’d better have a plan, a theory of change, or at least some basic hypothesis, that an Ushahidi map will help test. The downside is that DeadUshahidi is fairly snarky about how it goes about making this point, and frankly, makes some pretty poor assumptions about what makes for success or failure around a project. But perhaps most profoundly is that it may misunderstand the economics of the internet.

I’ve always suspected that, like pretty much everything else on the internet – website visitors, blog readers, number of start ups that succeed - the distribution of users on Ushahidi maps follows a power law distribution. This is to say that there is a long tail effect, with the most successful Ushahidi map of all getting probably twice as many users and the second most successful which had probably one and half more users than the third and so on.

What has the people at DeadUshahidi concerned is that long tail of “dead” projects. I mean, look at all those unsuccessful maps! What a waste! In part I agree. There are lots of people running around believing that rolling out a map will solve a problem when, without at least plan, it probably won’t.

There is of course, another way to look at that power law distribution. One could also say it is a sign of enormous success. Perhaps Ushahidi has made the cost of deploying a mapping platform so low that it is worth the risk of losing what is now a negligible investment of time and money to spin one up. I mean, isn’t this the effective application of Eric Reis’ "lean startup" Silicon Valley methodologies?

I suspect that the setup (not operating) cost of Ushahidi — while still not there — is slowly approaching the tumblr threshold. When I first set up a tumblr to experiment with, I forgot my password. It was literally easier for me to set up another tumblr than to futz around with my password. That’s an absurdly low barrier. Maybe Ushahidi is approaching the same level? If so, it means there are going to be a lot of failed experiments. It also means that may not matter. But more importantly, that is a remarkable accomplishment for Ushahidi.

I hopped on the phone with Patrick Meier, who recently left Ushahidi after spending three years with the organization. He agreed that the underlying idea of DeadUshahidi is good: projects are better off having a plan. But he was disappointed with DeadUshahidi’s narrow definition of success. As he subsequently pointed out in a blog post, many projects are bound by an event — an Ushahidi map with no new report in 12 months isn’t necessarily a failure because it could have been a wild success that was retired after the event it was launched to cover.

We also discussed how the mere threat of a crowd-sourced map might achieve some goals. He mentioned a map in India which tracked taxi cab bribes; soon after it launched and with few reports yet registered, local officials were shifting policies and cracking down on bribes. For DeadUshahidi this would qualify as a failure, yet it seems like a candidate for success.

Hopefully DeadUshahidi will cause people to engage in some more critical thinking before turning to mapping to solve their problem. On the flip side, DeadUshahidi’s mocking tone towards maps it deems ineffective, without any understanding of the goals of the groups behind them, is problematic. Moreover, mocking people or organizations whose maps don’t succeed hardly seems noble, and could even deter experimentation. In Silicon Valley, they say they celebrate failures and their lessons as key to a culture of innovation and success. While its intentions are ultimately good, it is hard to see how DeadUshahidi fosters such a culture in the world of crisis management.

Patrick and I did also talk about what makes for a successful Ushahidi map. Patrick said much of the advice on DeadUshahidi is good, but he added a few of his own:

1. Don’t think about the technology first
Many of the most successful Ushahidi maps are those deployed by established organizations with funding and a mission and goal. They end up using Ushahidi strategically to augment/magnify what they already do. The key lesson here is, it may be easy to set a map up, but running a project takes resources and requires discipline.

2. Get clear on your goal
IF you are starting from scratch on your own, have a strategy. What are your goals? What is your theory of change? Why is crowdsource mapping going to advance your goals? If you don’t have at least hypothesis to test it is going to be hard to know if you are successful, or why you are failing.

3. Ushahidi does not guarantee media attention
Often, when a map fails, organizers say it was because they didn’t get enough media attention. More explicitly they thought because Ushahidi has had media traction, if they use it, they’ll get some too. The fact is, your project has to be interesting and engaging on its own merits.

4. Be prepared to work
Many people found the work tedious and hard – truly successful projects often end up needing lots of volunteers, especially for live crisis mapping.

Personal Democracy Media is thankful to the Omidyar Network for its generous support of techPresident's WeGov section.