In 2015, OneWorld was asked to join an Oxfam-led project in Mali to promote accountability in local government. Our role would be to design an innovative service using new and/or traditional media -- like mobile phones, social media, and radio -- to engage local people with local government activities.

The project has been running for over a year now, and while we're not as far along as we had hoped when we set out -- more on the challenges of working in rural Mali further down -- it's been an extremely interesting and rewarding project that has forced us to be creative in both traditional and non-traditional senses of the word.

I was responsible for the overall process to ensure good design and effective implementation of our intervention. My colleagues Aissa and Aminata, based in Mali, would be overseeing all local activities, but they don't have design backgrounds.

I had been working (remotely, from London) on a couple other projects being implemented in Mali since 2012, so I already had a good understanding of the demographic and technology landscape -- the distribution of wealth, access to technology, how different types of people use media in their daily lives... that sort of thing. But we needed a deeper understanding of the hopes, expectations, and life situations for the key target groups for this project: youth, government officals, and project focal points. What were they hoping to get out of this project? What were their goals in life? How do they already interact with local government services -- if at all? What technologies do they use -- which ones do they love?

So I designed an interview questionnaire and sent it to Aissa and Aminata to use to begin collecting and sharing background information.

Once I arrived in Mali, we started by conducting a series of interviews with Oxfam's project partners to better understand their activities, and consider how we might design something to fit within them.

After meeting a number of partners working on local governance, a key insight emerged that seemed to explain the central problem quite clearly: many local people don't pay their taxes because they don't trust local government officials to do anything useful with the money. Yet government officials complain that local people don't know about the good work that's being done with tax funds. And they say they could be doing even more for the community if everyone would pay their share!

So we had our focus.

In fact, in many places the local council meets quarterly to discuss the budget -- and regular council meetings happen much more often than that -- but public town meetings take place only once a year to announce the results of the previous year's spending and discuss what was decided for the next year. And even when those meetings do come around, many local citizens don't participate. And of those who do, many lose faith again in the long interval before the next year's meeting.

These long periods without information seemed to be feeding a viscious trust spiral. We saw an opportunity: mobile phones and other media offer a great channel to share information in small bites throughout the year, keeping people informed and potentially building trust in the system -- bit by bit, message by message.

Now we just had to figure out how to do it...

Our ideation process was helped along by the fact that we all had significant experience in the country, so we came in with a pretty solid understanding of what tends to work best to communicate complicated issues to various target groups.

We mapped out the key actors, their goals, and their activities (within the larger project we knew we had to fit our service into).

We worked up our thoughts until we converged on the concepts that seemed most likely to make best use of the opportunity we had unearthed: use new (and old) media to share information about local government activities in small bites throughout the year, keeping people informed and building trust in the system bit by bit.

As this work was done during a single short trip to Mali, we had to steal ideation time when we could. We brainstormed in my notebook in the taxi on the way back from meetings, with post-its on a wall back in the office, and even some more on the whiteboards with colleagues back in London.

From these initial discussions, and building on our previous experience in the country, my colleagues and I developed some preliminary ideas:

1. Q&A: an SMS question-and-answer system staffed by local experts to answer people's questions and receive feedback about local government issues
2. Bulk messaging: citizens subscribe to get regular updates on community issues based on their interests/demographics
3. Citizen Observatory: a citizen representative attends local council meetings and shares summaries with the local community via SMS, Facebook, radio, and printed documents posted in key public spaces.
4. Participatory Governance: youth or other civic groups discuss their priorities and fill out surveys, which are submitted by focal points via coded SMS or smartphone application. The results are presented to government officials. Focal points analyse the results of the budget process and feed back to the youth groups, who then fill out another round of surveys, to be presented to the government officials, forming a continuing cycle of advice, action, and reaction.
5. Smartphone application explaining key aspects of local government functioning, using stories, simple games/quizzes, and infographics.

We came to these 5 ideas through a combination of practicality and imagination -- thinking both about what we had done previously, in Mali and elsewhere, on governance and otherwise; what other companies and organisations had done in this and similar spaces; and what human and other resources would be at our disposal through the network of partners working with Oxfam on this.

The Q&A idea, for example, was a natural for us -- it would be a simple adaptation of a quite straightforward platform that we had already implemented in several countries for health information projects -- including in Mali. We had even piloted a short-term Q&A platform in Mali for a month ahead of the 2013 elections, so we had a great deal of experience and even logistical contacts to facilitate set-up for that one.

The situation with the smartphone app idea was similar -- we had successfully created one in Myanmar on democracy and citizenship, and so we had a lot of experience to build on. The context, however, would be quite different in rural Mali, where far fewer people used smartphones, and it was unlikely we would find the resources within our current budget for an expensive content development process. But we wanted to float the idea anyway to see how it would be received -- you never know, we might find a way to develop a pared-down version if the people demanded it! Or maybe we could use their feedback to raise complementary funding.

The bulk messaging idea was one that would be fairly easy to add on to a basic Q&A service, and I've been intrigued by the idea since we first raised it during a scoping study in Myanmar a few years back. To me, it's analagous to the marketing SMS messages that have become extremely common across West Africa in recent years. They're frustrating users, but they must be bringing value to companies, since they keep sending them out. Could there be a way for us to seize on the positive components of this practice, without angering our users -- a sort of marketing-for-good approach?

The citizen observatory emerged as we linked what partners were already planning with the intent and opportunity we were focusing on. We realized that, to provide useful information to citizens, someone had to generate the content that would be shared. To be locally relevant, content would need to be generated locally. Thankfully, the other project partners were setting up networks of youth leaders and "focal points" -- could we use them as "citizen observers"?

The participatory governance idea also came about as we were considering the resources and activities our partners were putting in place. Youth leaders would be responsible for communicating regularly with youth groups in their villages. Could we use these networks to highlight youths' perspectives, and then channel those opinions into recommendations for elected officials? A survey tool seemed like a natural way to do this, as it could be done in organised groups (to ensure regular and continuous participation), and we already had experience collecting, transmitting, verifying, and visualising data via SMS and apps, through our election monitoring projects. If youth had a way to express their preferences, leaders couldn't ignore them for long without consequences.

We knew the ideas would take shape and transform once we started implementing, but we also knew we didn't have the resources to build out all 5 ideas, so we summarized them in a simple powerpoint presentation and met with groups of young people, a women's group, elected officials, and a citizen's watchdog group, to get their feedback.

Unfortunately, the trip was too short -- and we had to take advantage of small windows of opportunity to discuss with the various user groups -- so we were unable to develop prototypes in advance to show to the groups we met.

We did the best we could with schematic drawings, but it became clear that while there was great enthusiasm for all these approaches, the ideas were too far removed from anything that already existed for anyone to be able to provide specific feedback at this point. The only way to really understand which approaches would add the most value would be to develop working prototypes and begin testing.

We decided to start with the Q&A service and the Citizen Observatory for a few reasons:

- The Q&A was a low-hanging fruit -- we had a similar platform already running in Mali for a different project and could replicate that quickly to get something up and running. This could eventually be adapted fairly easily to provide bulk messaging as well, once subscriber lists had been developed.

- The Citizen Observatory matched well with other activities being conducted by project partners -- the focal points already in place could be used to test the concept. Plus, it could be tested with technology that was already widely in use even in these rural areas. In a way, this made it a bit of a low-hanging fruit as well, but also an approach that, if effective, could really pave the way toward more participatory governance at a local level across Africa. After all, if something like this could work in the deserts of Mali, it could be done pretty much anywhere on the continent.

So upon returning to London, I asked our developers to create a new instance of the Q&A platform. Aissa and Aminata set about finding local experts to staff the hotline and training them on how the system worked.

I then communicated the basic concept of the Citizen Observatory to our software developers, to make sure we were all on the same page...

To get things running as quickly as possible, we broke the process down into just a few simple user stories -- uploading a report of a meeting as text and image -- and they built a very basic version of a platform, just expanding on the infrastructure that already existed for the Q&A platform. Aissa and Aminata organised a training of project focal points to show them how it worked and get initial feedback.

As you can see, these first reports could only be printed or closed -- and I'm not sure "print" even worked. But it was extremely important to get a prototype version of the platform working as quickly as possible as this allowed my local colleagues to finally get past the puzzled expressions they had been receiving for the previous months when they tried to explain what OneWorld's platform would be doing. The concept was finally real -- even if it was nowhere close to fully operational. Now we could start to get some real feedback, and begin to understand if our preconceived ideas about how citizen observers would use the platform where anywhere close to reality.

In the meantime, we had got the Q&A platform up and running as well and recruited the first two hotline operators. They were trained on how it works, and about 300 questions had been answered between August and the beginning of December.

That was when a colleague of mine was able to travel to Mali, so she took the opportunity to get initial feedback on the platform.

We developed an interview questionnaire that focused on usability and perspectives for the next phase of development, which she was able to use with the two hotline operators and two citizen observers.

Their main feedback focused on:
- basic usability improvements (workflow, notifications, text size, accented characters, colors, etc.).
- new features (allow voice recordings, provide vocal instructions in addition to text, enable video sharing)

Some of the feedback were things we could act on immediately, while others (like video) would have to wait for future development cycles.

In the weeks following the training of the first citizen observers, the content that was uploaded to the platform was quite sparse. It was useful to see the types of content they wanted to share, including a citizens watchdog group meeting to analyse their local budget, an experience-sharing meeting between councillors from neighbouring regions, and voters checking their registration on Election Day.

But we knew we had to build out the functionality to enable the observers to begin sharing their observations with the community -- only then would they see the fruits of their efforts and be motivated to share more. This mean enabling editing of content and publishing to multiple channels -- we decided to start with Facebook and PDF, as these, again, were low-hanging fruit, as audiences already existed for each of them. SMS will come soon, but that requires generating subscriber lists, which will take some time.

So I developed a mock-up of how the citizen reporters' interface could work to enable editing and publishing to Facebook and PDF.

The developers had lots of questions, as it seems developers always do, and so I worked with a colleague to develop some functional mockups in Invision. 

We're planning to head back to Mali in early February to do more user testing with the hotline operators and citizen observers. That should provide enough time for us to deploy this latest round of improvements and for the citizen observers to try them out.

As the platform is finally reaching a stage where there is meaningful activity happening on it on a regular basis, we can begin to analyse more deeply how the platform is truly being used. Before we go, we'll do some quantitative and qualitative analysis of the way the Q&A and Citizen Observatory platforms are being used -- how many messages are passing through, when, from where, and what are they about?

Once we've drawn initial insights from that, we'll build out a user testing plan to guide our time in Mali. We've been thinking about how to create a lean version of David Travis' very comprehensive 27-page User Testing Toolkit, to make it more useful in the less ideal circumstances we often find ourselves working.

We'll interview the hotline observers and focal points, again focusing on usability and their suggestions for where the platform needs to go next.

We'll aim to observe the hotline operators and citizen observers in their work as well, to identify any unexpressed usability needs and opportunities for enhancements.

We'd like to interview some "women on the street" this time. Now that we're mostly through the very difficult opening stages of just getting the prototype to work in the field, we'll need to start focusing more on how we can expand the reach of the service beyond the usual suspects.

And we'll interview government officials and citizen watchdog groups again this time. Now that the service really exists -- and they can hopefully get their head around it a little more clearly -- it will be great to hear how they think it can support their efforts to share information in small bites throughout the year, keeping people informed and building trust in the system bit by bit.

And this time we'll show everyone some real prototypes -- examples of information they could have received by SMS, Facebook, or heard on the radio. We'll find out what's interesting to them and look for surprising new ways we can get them engaged with local government activities.

The project has definitely been a challenge to get off the ground, taking much longer than hoped. This is largely due to the difficulties of working in rural Mali -- communication is slow, intermittent, and often not stable enough to permit voice, let alone video. This makes it very hard to explain and delegate design tasks like conducting user interviews, and impossible to ideate and synthesise insights during the long periods I am not in Mali. But we're working on this.

We're getting better at finding workarounds. We are now comfortable switching quickly between WhatsApp and WebEx, Skype and SkypeOut... or just paying 28p per minute to talk things through when it's needed.

I've learned to provide very detailed interview guides by email, with direct links between questions and answer boxes, to ensure the person conducting the interview extracts as many of the insights we're looking for as possible.

When we are in Mali we generally don't have a lot of time with users, as the citizen observers and hotline operators are based in Mopti, which is off-limits to internationals right now due to insecurity. So we have to bring them down to Bamako or meet somewhere in the middle, which is expensive, time consuming, and exhausting for them, us, or both. This means we're not at our best when the work gets done, and we need to plan well to make sure we maximise the time we do have together. 

The language is also a barrier -- while I speak French, many of the user interviews have to be done in local languages that I don't speak. My colleagues are good at getting the insights I've asked for, but this limits our ability to identify surprising new opportunities, as they are not experienced in the abductive reasoning or design thinking that's often needed to take conversations down interesting unforeseen paths.

We're also slowly finding the right balance between advance planning and allowing for serendipity. We know that our next trip needs to be very well planned if we're going to get full value out of the short time we'll have with users -- and do all the activities I outlined above -- but we also know that some of the best insights come when you're sitting around a lunch table for three hours because it's too hot to drive back to town in the mid-afternoon.

No one said it was going to be easy. But of course, if it were easy, someone else would have already done it!