Yaw Anokwa is founder and CEO of ODK. ODK’s mission is to design, build, deploy and support offline data collection software for low-resource settings. ODK is one of the most popular data collection tools used in the aid sector, and forms the foundation for a rich global community of users active in every country on the planet. It’s being used for everything from COVID-19 tracking in Somalia, to monitoring rainforests in the Congo, to managing school attendance in Honduras.
In this episode, Yaw shares the story of ODK’s humble beginnings as a sabbatical / internship project. He recounts the serendipitous start of the effort, the first few users, and how the system got traction. In the second half of our conversation, he shares some of the secrets to success. He pays homage to Gaetano Borriello, his late mentor and advisor, who started him down the path that would set the course for the rest of his life. He talks about how to design tools for field workers and how to build the community needed to support open-source software. Lastly he shares the persistence he’s needed to navigate a complex sector and funding mechanisms, so that he could build a career creating software he can give away for free.
To find out more about ODK, you can visit getodk.org.
- ODK is a set of Open Source software for offline data collection targeted at low-resource and low-connectivity settings. Yaw is founder and CEO of ODK.
- Boldly going where no data tools have gone before: If you just can’t get enough of Yaw, you can here him speak on this episode of the Changelog podcast. Note that the Changelog podcast is targeted at software developers or those with a computer science background.
- Allen School recognizes Yaw Anokwa and Eileen Bjorkman with Alumni Impact Awards: You can read another account of Yaw’s life and work here, at the University of Washington page which honoured Yaw in 2018 with its Alumni Impact Award.
- SMS refers to text messages exchanged between phones
- Partners in Health is the organization which brought Yaw to work with medical record software in Rwanda and got him started in the space of digital health.
- Gaetano Borriello was Yaw’s mentor and PhD advisor at the University of Washington. Yaw credits Gaetano for introducing him to the idea of ODK, bringing him to Google, and getting him started on this journey.
- Neal Lesh gave a talk at the University of Washington which introduced Yaw to digital health. Neal also connected Yaw with Partners in Health and the opportunity to work in Rwanda.
- Google.Org is the charitable arm of Google, which gave Yaw and his team some initial support in the first year of ODK’s development.
- Vestergaard Frandsen is a social enterprise from Lausanne, Switzerland, that manufactures public health tools for people in developing countries. It is known for inventing the LifeStraw water filter and the PermaNet mosquito net. Vestergaard was one of the first organizations to equip thousands of data collectors with ODK, submitting over 40,000 forms a day. You can read more about this 2011 deployment here.
- MTN is a major telecommunications company active across many African, European, and Asian countries (similar to AT&T in the USA). Yaw references an early partnership with MTN Uganda as part of ODK’s initial deployment.
- The Grameen Foundation AppLab was a partner in ODK’s first deployment. In fact, as of the day this episode was published, the project page for ODK’s first deployment in 2008 is still available here.
- Java 2 Mobile Edition (J2ME) is the programming environment used to build applications on many feature phones (non-smartphones). It is distinct from the version of Java used today for Android application development.
- Kobo Toolbox is a set of open source data collection tools particularly popular in the humanitarian sector.
- Ona is a social enterprise based in Nairobi, Kenya, and Burlington, Vermont, USA. Ona provides software development and deployment services for the aid sector.
- Pursuing a Mission While Bootstrapping to Millions with John O’Nolan of Ghost is an IndieHackers Podcast Episode that Yaw cites, when asked about the moment he knew that ODK was a project worth pursuing. He shares a quote from John O’Nolan about traction, and how founders know when they have gained traction with their products.
- Luck Surface Area is an idea from Jason Roberts that Rowena cites, about increasing your chances of getting lucky, by giving luck more chance to get you.
- Mountains Beyond Mountains is a novel by Tracy Kidder about the life of Paul Farmer, the founder of Partners in Health and an inspiration to both Rowena and Yaw.
- Discourse is a tool for hosting online communities and dialogues. ODK users and developers have benefited from using Discourse to facilitate and organize the community.
- Evelyn Castle of eHealth Africa: Yaw gives a shoutout to Evelyn Castle, founder of eHealth Africa. He notes eHealth Africa an incredible through sometimes little-known organization doing digital health work in Nigeria and elsewhere.
- The Tao of Pooh is a book Yaw discovered in high school. He recommends it because of how it has helped him to think about and manage the many challenges he’s faced.
This transcript is an excerpt from the full interview.
The Costly Business of Free Software
Rowena: What were some of the low points, what were some of the risks, what were some of the points where you were like, “I just don’t know if it’s going to work out,” or “this is just too hard”. Because I’m sure you must have had some.
Yaw: My dear wife says: if Yaw loses a leg, what he’s going to say isn’t that he’s lost a leg, he’ll say “I’ve lost all this weight. I am now even more aerodynamic.”
Rowena: Hah! Oh my god, Yaw, do you still have both your legs?
Yaw: I have both my legs and both my hands right now. So for me, there hasn’t been an occasion where I thought, this is a disaster or this is not going well. We haven’t had huge experiences with data loss, for example, or where the software has failed. We’ve been very fortunate in many ways.
I think the things that are failures in a sense are… you know, this notion of a community oriented global health project, the sustainability and how that actually gets paid for isn’t so clear. So I think people have this notion of open source, where people will show up and they’ll build software and it’ll be a big Kumbaya community project. And for some projects it’s like that.
“You realize, realistically speaking, for the last 5, 10 years, the project is always a few months away from bankruptcy.”
But the reality for ODK is there’s a lot of code there. It’s very mission-critical code that people are using for Ebola vaccine tracking and monitoring elections.
And it’s a 10 year old code base. So it’s not the kind of thing that somebody would just show up as a volunteer and just jump into. It’s very complicated. The reality is, even though the software is free, it costs real money to build.
Yaw, gathering feedback from a Ugandan farmer during ODK’s first ever deployment. 11 years later, the Uganda government used ODK in an effort that vaccinated 19.5 million children.
ODK started in 2008. So for the last 10, 12 years, it has always been hand-to-mouth figuring out how do we actually fund those kinds of things. So that’s the kind of failure and challenge at this stage in the project. That’s what keeps me up at night, is how we solve that problem, how we figure that out. It’s when you realize, realistically speaking, for the last 5, 10 years, the project is always a few months away from bankruptcy.
And this is the case for most open source projects.
Rowena: That does sound stressful.
Yaw: It is very stressful, it’s just like, well, if another contract doesn’t come through or if we can’t find a bunch of contributors, there’s nobody who can take on the responsibility of this enormous code base for free. It just doesn’t work.
Rowena: It’s really great to hear you talk about that openly. When I hear ODK, I think of it as such a success story. And hearing that even with such a success story, sustainability challenges are still present, and are present for years. It makes it clear that this is not an easy space to work in.
If there are other people out there that are trying to build something like this, and they’re like, “Aaaagh, why can’t I pay for this all?!” Maybe it’s just a hard sector to pay for it all.
Yaw: It is, in many ways… And maybe this is a little inside baseball for folks, but there is a lot of money, like actual money, in global development.
If you think about how many Land Cruisers there are driving over Africa, what the diesel costs are, the cost in the pre-pandemic days of flying people around the world to have conferences where they discuss sustainability. Whereas if they just took the money to discuss sustainability and gave it to a project, that project would be sustainable!
Yaw: So there’s plenty of money in global development, billions of dollars, but not all of it filters down to the projects. And so there are structural problems that make this a very difficult problem to solve.
“If you think about the cost of flying people around the world to have conferences where they discuss sustainability… if they just took the money to discuss sustainability, and gave it to a project, that project would be sustainable.”
Rowena: How has that affected you, in your journey, as you’re struggling through this?
Yaw: You know, I don’t get a lot of sleep sometimes. But that’s just the nature of the beast.
There’ve been two approaches that I’ve tried to take over the last couple of years.
[One approach is this] people at Google get paid one hundred thousand, two hundred thousand dollars a year to work on software that is less complex. And so we’re making software as volunteers for free, giving it out. And obviously that doesn’t work because there are not a lot of volunteers to do that.
Rowena: Yeah, and it’s tricky because that’s also a lot of the appeal of ODK. You download it, use it, you don’t have to talk to anyone. But then how does it not stagnate?
Yaw: I’ve been talking to funders for the last, I don’t know, forever, saying that if funders and countries like this model, where it’s free to use and open source in this way, then there needs to be funding for it somehow. So that could come from the countries, that could come from the funders. And to me, it’s not unreasonable because…
For example, malaria bed nets are not sustainable, right? Like giving away malaria bed nets. Nobody says, “Well, this isn’t financially sustainable.” People do say, “Here’s a good thing that we want to do. These are public goods that we want to give to the world. And so we should find some way to, you know, pay for it.”
Yaw: And so in the same way, if you want software that is a public good, there has to be some way to pay for it. And it can’t just be Kumbaya magic. So that’s what I’ve been telling funders. And I think structurally it’s going to be very difficult for funders to make that change. But that’s the change that needs to happen if they want more things like ODK to exist in the future.
[The second approach is this] But as of today, what we’ve done on ODK is take more ownership of the problem. And so we are now providing folks who want to pay things of value that they can pay for. So if you don’t want to host ODK yourself, you can pay us and we will put it on the cloud [ODK Cloud] and provide technical support for it. And you pay for that service.
Rowena: That makes sense.
Yaw: Or if you want a feature, then you can hire the developer team and we work to put it in the software. So just going away from the model where it’s just like, oh, people, when they want to contribute, can contribute to it. These are very clear ways that you can pay us money and that money will go to making the software.
So that’s where we’re doing. But for the broader ecosystem, if funders and countries want more things like ODK, like the next kind of ODK to exist, there has to be a way for there to be consistent, reasonable, small but meaningful amounts of money to make it happen.
In the last 12 months, ODK has been used in every country in the world except Greenland and North Korea. If you know about a project in either country, let us know! – https://twitter.com/getodk/status/1376563820106252299