A practical guide on building successful civic technology

“The first lesson is to build something that will help address the problems, be hesitant about what you build”

Interview with Luke Jordan, a practitioner at MIT GOV/LAB and founder of the civic technology organization, Grassroot

Jordan founded Grassroot, a civic tech organisation that enabled communities and individuals to participate in democratic processes by organising groups, using low-end phones and discussed issues affecting these communities. He is currently at MIT GOV/LAB and is exploring and researching ways artificial intelligence and machine learning can advance democracy in developing countries, according to the MIT GOV/LAB website.

How did you get into civic tech?

I studied maths and computer science as an undergrad. I had a technical background and ended up in International Policy and working at the World Bank in governance and economic development in Asia. Then I came home to South Africa and wanted to put what I had learned to good use and I spent a lot of time just trying to understand what was happening and understand where my skills and knowledge could be helpful. I ended up partnering with a couple of friends who were senior technology developers.

Through that partnership, we ended up creating Grassroot. A lot of research showed that the capacity for collective action and civic engagement is massively important in making formal democratic systems work. In South Africa, in townships and in rural areas, it was tough to drive collective action for ordinary people who had limited resources. But people had cell phones, but the phones had very different capabilities and so after that process of just talking to a lot of people, we thought maybe we could create a simple way for people to call a community meeting on their phones that was easier and efficient, but that works on any phone and can reach thousands of people at a time.

That led to integrating Grassroot, which was this blending of technical, governance and policy backgrounds. And then just talking to people and understanding the problems, and Grassroot evolved from there over many years. Then we started working with other large civil society organisations to think about how we’ve built a technology platform that could reach many people. One of the uses was with Amandla.mobi and other campaigning organisations to help them do campaigns across the digital divide.

What happened to Grassroot? Will the project come back at some point?

The core technology and platform of Grassroot was handed over to our partner, Amandla.mobi, for financial reasons, and they use the platform for many campaigns. The typical foundation funding model doesn’t work for a tech platform that is scaling to millions of users quickly and with very unpredictable, up and down sort of costs. We explored a whole range of ways to make the platform self-sustaining such as asking users for small payments and having other organisations use it for various things.

We did get up to 40 to 50% of our expenses covered by our revenue. So we were able to do quite a bit, but unfortunately, it just wasn’t enough. We had to choose between shutting it down entirely or reducing it to the scope of just the campaign, functions and then handing it out, and we took the second route. Would it come back, possibly I think not in the same form, so part of what I’m doing at MIT is exploring new technology, and if something emerges from that, and it makes sense to build into a new idea or a project Grassroot could continue.

What are some of the key lessons you learned from running Grassroot which you would like to share with the community?

The first lesson is to build something that will help address the problems, be hesitant about what you build. Civic techies need to research and understand a problem before building technology intended to address the issue.

And the second lesson, good processes and work can take time and often good processes can solve things people thought were impossible.

For example, when we did our project to teach over WhatsApp, how could we teach over WhatsApp? People will be too distracted. What we found was that when we worked to evolve the structure and change how we did things, we eventually found a fit where we could successfully teach over WhatsApp, despite people being distracted and having other lives. In 2019 Grassroot created a training course, Teaching on WhatsApp, that aimed to build community leadership skills by teaching community leaders on WhatsApp how to organise and mobilise community meetings.

The Guide: What inspired/made you realise the community needs this kind of guide?

The guide, “Don’t Build it. A Guide for Practitioners in Civic Tech is a guide that provides practical tips for people in civic technology. The guide is a distillation of what I’ve learned over the years in the last few years, I would regularly get people asking me for advice on things. I would periodically hear people telling me about projects that they’d heard other people doing that we all knew would not make sense I realised that there’d been a lot of research about why building new apps frequently, is a bad idea, but it hadn’t gotten through because it’s written in quite academic papers or in a ‘researchy’ way. It was a combination of having the time to do it and having had lots of people asking me for these practical tips.

In your practical guide on civic technology, “Don’t Build it”, you warn against people building their own technology. Why shouldn’t people build these technologies? Everything seems tech-based, so why shouldn’t someone build it, especially if it will help people/communities?

The latter part needs to be interrogated because we think technology will help a community or help people when it doesn’t work. It’s quite funny because somebody told me about how insurance companies are now saying that their technology for car monitoring will help fix potholes. And it’s like literally every three years; somebody comes up with a new pothole app, which they think will help people because they think reporting potholes to local governments will get the pothole fixed.

If you ask people what the problem is, they will tell you that the government knows about these potholes, but they don’t care enough to fix them. But you get these service delivery reporting apps just repeatedly, and similarly, get apps like talk to your local government. They are always based on the same ideas, they never interrogate the real problems, and they absorb a lot of time, money and absorb large amounts of community members’ time because of workshops. And they end up not helping anybody because they manage to arrange some press releases and a lot of media coverage. After all, the media loves to cover tech, and they get a government press release.

And so they get many people to download the app in the first week and say, look, we’ve got all these downloads, and then nobody ever uses the app again, and it dies. It happens repeatedly, and it absorbs a considerable amount of resources and time to work on more important things.

You warn people about,’ building technology for bad”? What is bad technology?

A lot of the time, funder and donor priorities can end up distorting technology. It’s not bad in the sense that it actively causes harm, that does happen, but often in the sense of you end up building and rebuilding things because you have a large staff budget that you need to justify to a donor. Or you have a donor that wants to see certain things or things that they can tweet about, even though it doesn’t serve a purpose.

So instead of building an iterating technology for the real problem and making it work for the people who are using it, you end up building it for donors. That leads to technology that nobody uses that everybody knows is a failure, but all the donors write in their reports was a success. I think to an extent;

Bad technology is the use of resources for things that don’t lead to much change in a period and in a time where the need for systemic change in the world is more than ever. So the more money that gets sunk into repeated apps that don’t serve a purpose.

Can you tell me about your research/project at MIT?

In the last few years, there’s been a lot of hype about AI [Artificial Intelligence] and machine learning. And a lot of it is excessive. Every start-up tries to say that it’s doing AI, and most of it’s just doing the linear regression model and nothing more. But beyond the hype, the capabilities that are being developed are interesting, useful, and have many possibilities.

It seems like such a big shift from USSD, SMS, WhatApp to AI. What is it that interested you about AI?

It’s still connected, for Grassroot; the idea was to take sophisticated technology and deliver it so that anybody could access it so that it could help them and make a change in the world around them. The fact that something is AI or machine learning doesn’t mean it will need a smartphone.

It might be an AI machine learning model sitting at the back, helping to process text and help people understand the consequences of laws that could be delivered via WhatsApp, phone calls or other methods. It’s closer than it appears and I felt like there was a space for understanding what could be done with this technology, as somebody who understood how it worked and could code it and understands the governance problems and understands the real conditions under which technology gets used.

Interestingly, you are working on a project in AI, but you warn against people building their technology?

Yeah, well, that’s why this process is about figuring out the problem, so I’m experimenting now; this is that pre-process, about making sure you understand what the problem is and a fit between the problem and the technology. So both the governance lab and I are very open about the possible endpoint of this research. This is an exploration of what problems might be and if the technology fits. And if at the end of that, we decide not to build it, then we’re not going to make it.