About 6 weeks ago we arrived back from Monrovia, having just undertaken our first design exercise out there. Paul wrote about our experience in this blog as a broad overview. After further long distance design calls we wanted to delve a little more deeply into the process we’re following and what we’re learning about it.
To begin with, I should mention is that this is the first project where mySociety International will be leading on the implementation of a project using Design Thinking (the South African was a trial using a cut-down version of the approach and furthermore the implementation is being carried out solely by a local team).
Another important point for us is that the Design Thinking approach encompasses far more than just thinking about software development. The aim of the process is to develop an understanding what is required to ensure that the users’ need is addressed. Some of the solution might be technical, but much is likely to be about the processes and people that are required to ensure that needs are met.
For example, in the case of the Liberian FOI project where the internet penetration is low and the day to day obstacles people need to overcome are significantly more difficult than in the global north, a large proportion of the project time and resources will be dedicated to delivering offline services.
These provision of these services will tend to take a shape that fits into citizens’ current experiences. An example might be setting up an SMS short-code that allows people to contact a support team to call them back, in order to draft an FOI request on their behalf. They will then physically deliver hard copies of those FOI requests to the relevant ministries in Monrovia. This type of solution could be particularly beneficial for people who live outside the capital and do not have the time or resources to travel there to submit requests directly themselves.
There are two critical differences between the Design Thinking approach and other projects we have run with groups in the past. The first is that, with non-design centred partnerships, most groups start the process with a firm sense of the “type” of thing that they want from the outset – for example an instance of our Pombola platform that is used to power Mzalendo.com.
This is totally understandable, and in many cases what the funders of these projects are looking for, but it doesn’t necessarily mean that the users will get the most impactful solution.
Where we have sufficient funding to undertake this process available we intend for all of our new international projects to be starting with no expectations about the ultimate product – the outcome might even be that we don’t end up producing any software at all, because the best solution might consist of a desk, a phone and some radio advertising.
The second difference is that we have usually relied upon the local implementing partner to provide the insight and define the specification.
For our Design Thinking based projects we’ll have a closer relationship with the local partner and together we’ll identify potential end user groups we can talk to about their needs.
After doing a first round of in-depth interviews, the team then synthesise the information – essentially sharing what we’ve learnt with the rest of the group to pick out the most important points. The next stage is empathy mapping, where we figure out what people have said, thought, felt and done. This is a key stage in helping to identify the needs of the users.
It might seem simpler to ask them what they need, and often we did say something like “What would make this process easier for you?”. Yet actually analysing what they’re saying about the process and at what points they seemed frustrated or blocked – that tells us a lot more about points where we could change and hopefully improve the process than a straight up “What do you need?” question.
This is the stage we’re in at the moment with the Liberian project, though we have done some brief forays into Ideation – coming up with ideas for how to address the needs, and we are now starting to thinking about prototyping these ideas.
Of course, this method doesn’t mean we’ll completely stop using software solutions, or looking at A/B testing and Analytics as measures of the success of website. However we will also be looking at other measures of success or failure based on the product we’re building and the change we’re trying to achieve.
In the case of the Liberia FOI project, many of the users are likely to have no direct contact with the software themselves so we’ll need to design a monitoring system that measures the effect the changes have on their experience of making FOI requests.
One thing we’ve learnt is that a Design Thinking approach doesn’t only affect the first iteration of a solution. This may seem obvious, but from our brief work with this process we’ve seen that uses/users can be hard to predict at the outset – though in the case of the latter group we worked hard to spread the net widely in order to find potential users in Liberia.
So we’re interested to see, when we get to that point, what the prototype testing brings back and what new changes, improvements and tweaks need to be made.
More about our experiences with this process will be shared the lifetime of the project, as we learn, change and iterate ourselves.
Last year Dave Whiteland wrote about our first experience of using design thinking in creating the specification for a Freedom of Information Project in South Africa.
For those of you unfamiliar with the design thinking approach, it is worth me providing a little context (borrowing heavily from Dave’s text).
The design thinking way
Traditionally mySociety has built international digital projects by working closely with our local partner to define a specification. We then build to this specification and seek to continually improve it . We perform usability tests, we apply A/B testing, and we think hard about what our analytics tell us. The problem is that much of this is reactive, iterative design: it’s being applied after the core product has already been built.
Design thinking challenges this approach by suggesting that the user on which initial designs are often based is purely imaginary. As a result, the site inevitably includes the assumptions and prejudices of its creators. This won’t necessarily lead to a bad design — especially if the creators are benign and experienced — but it must fail, by definition, to account for the unexpected things that may motivate or concern actual users. The design thinking process attempts to change this by approaching the initial problem in a prescribed way and following a process that isolates genuine, existing requirements. This includes, in design thinking terms, processing the initial interviews into empathy maps from which requirements emerge, and which themselves become features that are rapidly prototyped in isolation from other parts of the system.
Our commitment to the process
mySociety is dedicated to maximizing the impact of its projects by tailoring solutions to the local context. As part of this, we have committed to only carrying out new large-scale international digital projects where we can follow these design principles. One challenge with this approach is that, unlike traditional projects, we are unable to provide funders with a clear description of what the project will deliver. At the start of the process the tools that will be built and the processes and infrastructure that will surround them are unclear. Furthermore, all this prototyping and piloting adds to the time these projects take to complete.
In Making All Voices Count we were incredibly fortunate to find a funder who was not only undeterred by these concerns but actively appreciated the value of the approach. Earlier this year they awarded us a scaling grant to work for a two-year period on a Freedom of Information project in Liberia.
On the ground in Liberia
Last week the project team met up in Monrovia to start the design process. Our partners on the project are the iLab Liberia and Public Works at Stanford, an offshoot of the d.school which focuses on applying design thinking to governance in the developing world. The staff working on the project are Luther Jeke, Carter Draper and Teemu Ropponen (Ilab); Jenny Stefanotti (Stanford); and Paul Lenz, Dave Whiteland and Jen Bramley (mySociety).
So what does this actually entail? Briefly, it first involves meeting with a large range of different stakeholders, users and potential users and building up an understanding of their current behaviours, their needs, challenges and perspectives. We interviewed more than 20 people, including the Information Minister, the Independent Information Commissioner, investigative journalists, Public Information Officers, FOI NGOs and community groups.
Each day we would break down our notes from these interviews into things said, thought, done or felt, and group them by type of stakeholder. This is a very active and visual process, resulting in sheets of paper being covered in hundreds of Post-It notes.
In working through this process it became clear to us that there were a number of common issues. Firstly, while the FOI law in Liberia is legally very strong, in practice adherence is pretty poor – partially due to simple process failures, and perhaps sometimes due to willful avoidance. Secondly, despite significant resources being invested in “sensitising” (educating) citizens about the law, very few FOI requests have ever been made by individual “average” citizens; rather they have been submitted by NGOs, journalists and activist groups. These two finding might not be hugely surprising, but others were perhaps less obvious. For example, almost without exception, the Public Information Officers that we met were deeply proud of their work and wished the FOI law was used more; even though the law allows for electronic or phone call requests to be made in reality each request must be hand-delivered, hard copy, and a receipt obtained; finally, even skilled and experienced investigative journalists can end up spending years chasing requests through a Kafkaesque bureaucracy that takes advantage of the requester’s ignorance of the detail of the law.
The next stage, that we are really only just starting, is to identify some approaches that might address these issues, and then to find a way to attempt to prototype the solutions. To save time and cost and in order to enable maximum flexibility these prototypes often take a “Wizard of Oz” approach – human intervention in lieu of building a technical platform to trial. An example would be rather than building an SMS gateway that interfaces with a computer system the prototype relies upon a person simply receiving the SMS on their phone and typing it into a form.
The week was incredibly intense and rewarding driven by Jenny’s fantastic energy in overseeing the whole process, and the great commitment and engagement from the iLab team – particularly given the developing situation with Ebola while we were there.
Freedom of Information: really?
You might ask: given the infrastructure, income and health challenges facing Liberia , does FOI really matter? Is it perhaps a right to be addressed when more essential needs have been met? Are people even worried about being able to request information from their representatives, due to the other factors at play?
I can make no more compelling case in response than one that was made to me by one of our interviewees:
“Liberia went to war over the mismanagement of natural resources in our country. FOI can help hold people to account and stop it happening again”