Technology for Social Impact with Jonathan Jackson, Co-Founder and CEO of Dimagi

(Joel Beasley at 00:00:03) Hello, my friends. Today we're talking to Jonathan, the co-founder and CEO of Dimagi. And we discuss how Dimagi built out tech for contact tracing and data processing during the pandemic, how Dimagi has augmented their engineering teams by working with Cactus Group over the years, and tips for approaching each new task with curiosity. All of this right here, right now on the Modern CTO Podcast.

(Jonathan at 00:00:34) Here we go.

(Joel Beasley at 00:00:35) This is the Modern CTO Podcast.

(Jonathan at 00:00:45) I got in basically since I was born. Both my parents are engineers. One worked for AT&T and one worked for IBM. And I was always just around computers, technology. My dad's basement is filled with ham radio equipment. So we have massive ham radio towers in our yard when I was growing up and endless rows of radios down in the basement. And that was something that I was just drawn to. And then when we got our first PC, I was hooked. I just was on it all day pre-internet. I was doing a lot of programming, learning BASIC and those types of things. Actually, two of my friends were also very into technology back when we were in grade school, and we'd pass notebooks back and forth, programming on paper, and then just read the code and figure out what the algorithm did and then write code back. And that was our version of passing notes. I was always kind of a huge nerd.

(Joel Beasley at 00:01:39) Yeah, that's awesome. So how did your career get started in tech? Were you always following that as a career path, or did you do anything else?

(Jonathan at 00:01:50) Yeah, I mean, I thought I might want to be pre-med, pre-law, or finance, but I always kept getting drawn right back into tech. So when I went to college, I did it here in the United States at MIT, and freshman year took my first college-level computer science course and was totally hooked. I was programming all night, doing extra effort on these problem sets, and really just loved the ability to code something and then hit compile and it's right there in front of you. And it's a working product that you can touch and feel, and that was something that I was always really drawn to. But I did take a slight detour in college over to the finance world. So I interned at Goldman Sachs on the mortgage desk and was really also interested in the world of finance, but my heart was always in technology and building stuff. And so I quickly realized that wasn't for me. And senior year, I tried to get a startup off the ground in social networking, and LinkedIn and Facebook obviously took care of that problem for us back in 2001. So then I landed on Dimagi, which is my current company that I've been running for the last twenty years. But that path through technology has always kind of been what was calling me and what I wanted to do.

(Joel Beasley at 00:02:55) That's awesome, man. So I know you founded a couple of companies and this might be a little bit of a tougher question, but early on, what was an assumption you made when starting one of your companies that you found out was just way wrong?

(Jonathan at 00:03:15) I'm wrong a lot. I'm trying to think of the most wrong assumption that I had. I think, you know, I'm twenty years into Dimagi and I still forget this lesson, which is, if people were trying to make the most impact per dollar they spent, the market would be functioning correctly. You wouldn't need a Dimagi who is so unique in the market, but they're not. And they never said they were in the first place. You know, the US government has political motives. The Gates Foundation has Bill and Melinda deciding what they want to do with their money. There's a lot of different people who are trying to create impact, but it's their own version of impact. It's not my version of impact. And then on top of that, you have the government, the host government that we're working with, and the community healthcare worker, and everybody's got these different motivations. And even twenty years later, I find myself just being like, man, if we could just do this X, Y, and Z thing, it'd be this huge impact. Why won't anybody fund this? And you're like, look, you know that's not how the market works, and you can't just wish that everybody agreed with you. And so one of the things that I've been wrong at in all my companies—I've learned a ton and been fortunate to sit on many great boards that's also taught me this—is when you look at the system dynamics that you're operating in, even if it looks like a totally normal functioning for-profit market, there's always weird stuff going on that is altering how the system functions. And as a starter of companies and somebody who participates in these markets and systems, you really got to understand the system you're operating in and what positive and negative effects it's going to have on your organization. And you can't wish that the system was going to be different. You got to play the game as it is today. And you can try to change it and advocate for a better system tomorrow, but if people aren't paying for health and education outcomes together, don't yell and kick and scream and hope that they do. Go figure out a different business model to sell the health or education, but if they don't want to pay for both, they're not going to. And that lesson of those companies early on was something that I still learn every day. But it's really kind of understanding if you're going into something and you're like, oh, this appears broken, I'm just going to convince people to fix it—that's typically not how that argument goes because somebody else probably tried to make that same argument before you. So what is the secret thing you know that nobody else does, or what is the novel innovation you're bringing to the table? And it shouldn't be you're just more convincing than the previous person because that's often not a good theory of change for a startup.

(Joel Beasley at 00:05:37) Yeah, that makes a lot of sense. So you're a co-founder at Dimagi, right?

(Jonathan at 00:05:41) Yeah, exactly. So my three other co-founders—Vikram Kumar, who is still with us and is our chief medical officer, and then two others who have since left, Tarjay and Vish. Vish is now the CTO of Mayo Clinic, so very successful. And Tarjay is an executive in biotech. So the other three co-founders are all immensely successful and really amazing technologists too. But yeah, Vikram and I, for quite some time, and I helped co-found it and get it off the ground.

(Joel Beasley at 00:06:09) So where did the name come from?

(Jonathan at 00:06:11) So Dimagi means brain in Hindi. Two of the four co-founders were Hindi, and Dimagi kind of means somebody who can be a smart fixer. And then I also learned it might also be a wise guy or smart aleck, which kind of fits with our Dimagi personality a little bit as well.

(Joel Beasley at 00:06:28) Nice. So what does the company actually do?

(Jonathan at 00:06:31) So we are a social enterprise that builds a platform that allows you—what now would be called a low-code application platform—so allows you to build your own applications and provide these to your users. Our most canonical use case is in the healthcare sector. So you'd be using our platform called CommCare, build a mobile application, give that to a frontline worker. She's in rural Africa, going household to household, registering pregnant women, helping them take better care of their child during the pregnancy and after birth. And our application is the system the whole time they're doing this. So it's kind of like if you were imagining your doctor's office—it's a lightweight, good version of that medical record system in a phone. And the key is that it works offline. So our technology and our company's ethos has always been to design for low-resource settings. We've always tried to build our software in a way that was able to be used by the most accessibility requirements and reach the most number of people. And that's really the main focus of our firm.

(Joel Beasley at 00:07:26) I know that identity in healthcare is a really tricky area. It's hard as a patient to even have access to my own records. What kind of problems have you guys run into in that space to be able to store records locally with a patient on them?

(Jonathan at 00:07:43) Yeah, that's an incredibly complicated challenge that has led to a lot of technology sophistication that we've got to build out around synchronizing records. So the identity in low-resource settings—often you may not have a national unique identifier, so a driver's license or Social Security number here in the US is very commonly used, and increasingly your mobile phone number. None of those may be available in a low-resource setting, and so you're getting things like a health ID. So it's kind of like a functional ID. It's specific to the use case. And then the software and the technology is able to recall that ID, but it's not able to identify you across different domains that may be more common here in the United States. But that portability of the record is something that we spent a lot of time engineering. So the ability for your device to synchronize with my device and move just the part of the record that we each need is something that allows us to provide very sophisticated care on kind of low-end technologies. So it's not like you can design for the latest iOS model or the latest smartphone, but typically two or three years on older technology. And so you need to make sure the power requirements you have, the battery requirements you have, all allow for supporting this workforce that may not have readily access to the internet, but also not ready access to charging or power stations.

(Joel Beasley at 00:08:59) I'm just having trouble, being surrounded by the internet at all times everywhere, kind of thinking about how you would make records accessible without the internet. How would that work? I'm trying to wrap my head around.

(Jonathan at 00:09:14) Yeah, I know. It's a good question. It's very difficult. So unfortunately, we've had to spend a lot of time on the engineering challenge of this. So basically, the record for you—let's say I'm your frontline worker. So I've registered you. I'm coming back to provide you support. My device knows that Adam's John's patient, and so I have Adam's record. But maybe I'm talking to you and all the other households in our village, so I have a hundred households in our village on my device that's available offline so I can deal with the processes I need to support you. But the other hundred thousand people or hundred million people in our country or community or district, they're not on my device. So we have to do a lot of work to understand for John to do his job, he needs Adam's record—what other records does he need—and put only those records on the phone. So you're both making sure you have strong data security but also can meet the performance requirements you need. So it's quite a complex problem in terms of understanding how a worker interacts with a given community and village. That has been a big technology investment for us.

(Joel Beasley at 00:10:14) Gotcha. Yeah, that makes sense. And that makes sense why it's so challenging too. I want to talk a little bit about Cactus Group because we were originally introduced—recently, not super recently, actually, time really flies—but we had Cactus Group on the podcast a little bit ago. And they were a really interesting company with how they—I really liked how they put social responsibility at the forefront of what they do with their custom software development. And I was curious, what's your relationship with them like?

(Jonathan at 00:10:46) Yeah, so Cactus Group is an amazing firm, based in North Carolina, and I've actually known the founder, Tobias, for over ten years. So we originally connected—we were both on a project together working with UNICEF. So we were building a messaging system in Zambia, in sub-Saharan Africa, that allowed people to text in and out of the system to get information or submit data. So you would just use an old Nokia phone and send SMS messages, and we can move that data through the system. And that's how we originally connected, and they were like Dimagi, a very socially driven firm, trying to build technology to make a difference. And we hit it off right away. And so over the years, we've worked with Cactus in a variety of different engagements where they provided their amazing engineering capacity to supplement ours. And, you know, the fact that they're so good at building custom solutions and client management was immensely helpful to us over the years, and particularly with COVID.

(Joel Beasley at 00:11:37) What kind of tools were you building in conjunction with Cactus Group to help out with COVID?

(Jonathan at 00:11:43) Yeah, so we—as I mentioned, our ethos has been in global health, really designed technology for low-resource settings. With COVID, we actually pivoted and moved into the US market as well. So the Center for Disease Control had called us up back in March 2020. This was when we had only had the outbreak here in the US in Boston, Santa Clara, and Seattle, but it was pretty clear this was going to be a worldwide problem. And we'd worked with them globally, but we hadn't worked together here in the US. And somebody who knew about us said, hey, do you think your software might be relevant here in the United States? And we're kind of like, don't we have systems for this here? And they're like, no, this is going to be very bad from an IT standpoint. Do you want to see if you might be a good solution for contact tracing? And we learned more about the use case, and we had done contact tracing with Ebola in West Africa, so we knew that the platform was a good fit, but we didn't quite know the health IT landscape here in the US quite as well. But we got that call Friday night. I talked to them Saturday. We had somebody on-site Sunday, and so we were working with a public health department in California and built the solution overnight, deployed it the next day, and were just iterating. We started out contact tracing on Monday. By Tuesday, they're like, we can't keep up. We're not doing contact tracing anymore. Can you help us with case investigation? So contact tracing is, I'm a close contact with you. I need to let you know you might have been exposed. Case investigation is you're positive. I need to talk to you and help you quarantine and see if you need any other services. Okay. We moved into case investigation on Tuesday. Case counts were skyrocketing. By Wednesday, they gave up on that, and we had to do nursing home support. So we kept rebuilding the technology and the application. You can imagine with that kind of flux in the application you're building, connecting to other data systems is a huge challenge. And so in the United States, you need to get positive lab results into our platform as fast as possible. And there's a lot of archaic legacy technology. There's a lot of challenging innovation problems. And so as we looked at, can we really make an impact here, we're like, if we don't get the data into our platform, CommCare, very quickly, it won't matter how good we are at case finding or case investigation because it'll be too late. You know, if you have a positive lab result but you don't talk to them for seventy-two hours, that's three more days they've been spreading COVID without knowing that. So we were like, well, we think the platform's a really good fit, but if we can't get data in and out fast, we're going to be stuck. And that's when I called Tobias and said, hey, you know, we're thinking about trying to really support here in the US, and we're going to have this huge integration challenge. Do you guys have capacity to kind of jump in and look at whether this is something you would do with us? And he was immediately like, absolutely. You know, let's see if we can do this together. And so our first project was with the city of Philadelphia. They had decided to go with CommCare, and then we needed to connect to their lab system so that when you got a positive lab result in Philadelphia, it would come into our system.

(Jonathan at 00:14:24) And Cactus worked with them to connect their lab system to CommCare and did an amazing job, and it's been fully operational as soon as they built it and it's still running today.

(Joel Beasley at 00:14:35) Wow. So how widely did that end up rolling out?

(Jonathan at 00:14:38) So we worked together in Philadelphia. We worked together in Colorado as well. And now have a couple of other projects that we're looking at. But those two states in particular, we have operated in four different statewide projects and two other local geographies, and they've helped us in two of those six.

(Joel Beasley at 00:14:54) That's really cool. Yeah. You hear about all the different solutions that had to get rolled out to respond to the pandemic. And it's kind of hard to keep track of all the different people and companies that were pieces of the puzzle that helped us get through it. So it's really cool hearing your story and how you guys were able to help in those specific use cases.

(Joel Beasley at 00:15:16) That's awesome, man.

(Jonathan at 00:15:18) And it was crazy because even after we found that it was a good fit, we're like, come on, there's got to be other products here in the US, right? And so we're looking at Salesforce and Microsoft Dynamics and these other huge, huge behemoth players. And so when we talked to the public health departments, because we've done so much work for low-resource settings, making sure the user interface was really, really solid for low-literate users or non-tech-savvy users, all these things that we never intended to be features that would necessarily be highly differentiated here in the United States.

(Jonathan at 00:15:47) When you start to onboard a ton of contact tracers, like thousands of contact tracers who might be not the most tech-savvy users or might not be comfortable using certain types of sophisticated software. Actually, we beat out Salesforce and Microsoft head-to-head with certain customers, which we were really excited about. I'm sure there's use cases where they're way better than our platform, but to do that and to win on usability was a really exciting moment for us.

(Joel Beasley at 00:16:14) Yeah. That's incredible. So what kind of things do you have to do to make your software more usable for low literacy and low tech savviness?

(Jonathan at 00:16:24) Yeah. I mean, we recently got down this track because we had tried to deploy, I want to say this was back in 2011 in Afghanistan. So we went to Afghanistan, built an application on our CommCare platform that was for maternal and child health, and then we went to go train the users, and a lot of the users were not fully literate. And so, we're like, oh, what do we do? And then the client recommended, well, what if we do kind of audio and image-based user interfaces?

(Jonathan at 00:16:52) So rather than have it be all text, let's—can we do pictures? And can we use audio to kind of nudge you along the user interface? I'm like, oh, that's a great idea. And finally, this was a literacy approach that the user experience features were built for. We forgot to turn them off when we were demoing to somebody else in another setting.

(Jonathan at 00:17:09) And they were like, oh, I want that too, but not because my users are illiterate. I just think that that's going to be a better application. And what happened was using embedded video, embedded audio when you're interacting turned out to be a stronger way to do what in my field is called behavior change communication. So if your doctor is trying to convince you to exercise, like, hey, you should really exercise. You may or may not find that very compelling.

(Jonathan at 00:17:31) Where they have a really good three-minute video of Michael Jordan telling you to exercise, you might take it a little more seriously, right? So the idea that there's multiple ways to persuade people and the human contact might be one, but some other people may be more susceptible to a video or some other people may like to see an image or audio, was really an area that we've invested a lot in the platform. And so, we call that part of service delivery. You know, so it's in the interaction that I'm trying to support you as opposed to how a lot of people think about technology in the public health sector of just collecting data.

(Jonathan at 00:18:04) You know, so how do I get my data back so I can create an epidemiological model or surveillance system? And we also think it's critically important to empower the users. You know, how can we make that user's job better? And that led to a lot of the features we built that make it a very powerful technology for the settings that we're supporting.

(Joel Beasley at 00:18:22) Yeah. That's something that I've—that's kind of like a common line I've been hearing a lot recently about empowering the users. Like in cybersecurity space, you can build the most secure platform possible that hackers can't get in. But if you're not educating and empowering your users to practice proper cybersecurity hygiene, then it's still not going to be secure. That's something that we recently interviewed the CTO of Avast.

(Joel Beasley at 00:18:52) And he was really big on that and they're moving more into education for the users there. And that's really cool. And so I wanted to ask, because in your example, you said like if Michael Jordan told me to exercise, I'd be like, yeah, yeah, I want to exercise. Um, but so did you guys reach out to influential people in the areas that you were operating in to help convey these messages to the users?

(Jonathan at 00:19:25) Some of our clients have. So we typically would, you know, we're B2B, so we would sell to you and you would be trying to create that content. And so you would go reach out

(Joel Beasley at 00:19:33) to Jordan to

(Jonathan at 00:19:33) do it. But, yeah, some people have gone to local celebrities. And it doesn't necessarily need to be celebrities. You know, a huge part of these community-based services are built on local trust. So it might be a local religious leader or a local political leader, and it might not be somebody from the urban center who's the flashy athlete or the flashy musician who's the best carrier of that message.

(Jonathan at 00:19:56) In fact, the dubbing of audio for videos is actually really complicated because you have to practically think about, do you want that accent to sound like it's urban or rural? And depending on the culture you're in, a rural person might want it to sound like a sophisticated urban person or might want it to sound like a rural person. And there's all of these science of fascinating things just on how you dub videos to be most persuasive in different cultures and climates. And so, exactly to your point of thinking about the human factors, like, that is such a critical element of making technology successful in the markets we work in. But increasingly, we've realized it's a huge part of making technology successful everywhere.

(Jonathan at 00:20:35) You know, you can have the best security software as you just talked about, and if your users are still writing their password down on their computer, when they lose that computer, they're going to be stuck. And so, that combination of technology and human factors and making sure people actually want to use the technology you're building and you're solving a problem that both they feel and the organization they're part of is really something that we get excited about and is something we spend a lot of time thinking about.

(Joel Beasley at 00:21:00) So when you're going into new markets and areas of the world that you're not super familiar with, how do you interface with—how do you go about hiring locals there that can help guide you through the processes of just speaking with prospective clients in those areas and adhering to all the different cultural nuances that you might not be familiar with?

(Jonathan at 00:21:26) Yeah. It's a great question. And it's something that we spend a lot of time doing—local capacity building and local partnerships. So we are always going and working in a new market or new culture in partnership with somebody who's actually running the program that we're equipping.

(Jonathan at 00:21:40) And so we're very fortunate that kind of by design and by default, our approach is always to go in with a partner who knows a lot more about local context and local culture than we do. And we then are trying to be very humble and do what we call design under the mango tree, which is really listening to the users and trying to build an application that they really like. And that terminology came up from some of our early work in Tanzania where we were literally interviewing frontline workers, getting their feedback, and then going over to the tree, coding, redeploying it, and showing it to them. And it wasn't just great for us to really learn how to create a product, but it was super empowering for them. They said, hey, can we do this?

(Jonathan at 00:22:18) And then overnight, we'd go change it. And they're like, whoa, you actually listened to me and changed the software to do what I asked. And so, it was a win-win for both our users who were getting a better product, but also for empowering our users and making sure that they felt heard and that we were there to serve them and make sure they had a powerful application.

(Joel Beasley at 00:22:36) Yeah. That's really cool. Um, because, yeah, you definitely can't go into these new markets without local support. You hear about so many—I mean, I've been kind of disappointed recently hearing about like, there's been a lot of press around charities that aren't actually getting a high percentage of their donations to the end user. And a lot of that's because they don't have those local partnerships, boots on the ground that are helping them where it's actually needed.

(Joel Beasley at 00:23:06) They try to do it all themselves. So that, uh, it's really cool that there's so much effort put in on your side to do that.

(Jonathan at 00:23:13) Yeah. And I think that as I mentioned, there's both a goal, but it's also a necessity. Like, if we didn't do that, we wouldn't be building good technology. The users wouldn't like it. And to your point, you know, there's ways to do this that are crazy high overhead.

(Jonathan at 00:23:26) You fly people in, you do it all with high-cost salaries, or you can do a much smarter thing, which is use local talent, which is probably better than yourself anyways, knows more what they're doing. You're building local capacity, and it's even self-serving for us because we're building out our network. We're building out the ecosystem, people who know CommCare, you know, so it's win-win. And I think that's just a way better way to go about deploying technology for social impact.

(Joel Beasley at 00:23:50) Yeah, absolutely. So one thing I really like about Cactus Group is how they're super involved with the open source community. Are you guys also,

(Jonathan at 00:23:59) uh Yes. We are. Absolutely. Yeah. Our entire platform is open source.

(Jonathan at 00:24:03) And so this is something that we're really proud of. In our field, it's called Global Goods, so Digital Public Goods. Um, but we're one of the biggest. An independent kind of code analysis tool found CommCare to be one of the 53 best repositories on GitHub on an analysis they did back in 2019, which was pretty cool. I don't know how accurate the code analysis was, but it's something we're really proud of.

(Jonathan at 00:24:24) We have been open source our entire history at Dimagi and we've worked with Cactus really closely on a platform, that messaging platform I mentioned. We were working together with UNICEF. That's still in use today, so it's been 10 years. And that open source code base is still there. But, yeah, we're huge believers in open source.

(Jonathan at 00:24:41) And a great example is, you know, while Cactus was doing those integration projects with us, they said, hey, it'd be a little bit easier if you could do x, y, and z with the data model. And we were like, oh, could you build that? And they're like, sure. So we contracted with them to build core components of CommCare that they just pushed into the code repository. So it not only enables you to build great software, but allows you to build partnerships in ways that don't work very well with proprietary software.

(Jonathan at 00:25:09) So, there's a huge business benefit to the open source side as well.

(Joel Beasley at 00:25:12) That's awesome. Have you found any recruits through the open source community, like an engineer that was just working on it on their own? You're like, hey, you're doing really good stuff. You want to work here?

(Jonathan at 00:25:23) Yeah. Not directly because they're working on CommCare, but certainly a huge number of the recruits we have have open source code in their GitHub profile. Like, a ton of engineers these days are submitting their resumes with GitHub links to all the work they've done. It's a great way to see how big of an open source community member they are. The fact that we're open source, I know, draws a lot of engineers in.

(Jonathan at 00:25:45) We're hiring—if somebody who's listening to your podcast is looking for an open source job, feel free to come apply to Dimagi. But we do a lot of brand building around our open source efforts, and it's certainly a great recruiting tactic for engineers. But our software is really a sophisticated enterprise-class B2B platform, so it's not the natural type of open source project that's going to draw a lot of random software engineers into the project itself. But a lot of the libraries we've spun out have been pretty popular and have been ways for people to come in.

(Joel Beasley at 00:26:18) Yeah. So you mentioned it's not super common in an advanced B2B enterprise platform. Because when I think open source, I always think about a community of users working on it and improving it. But with being B2B, how do you guys balance the open source with also keeping it able to monetize? Like where are your core competencies that you guys are absolutely needed?

(Joel Beasley at 00:26:46) Someone can't just take your repository and run it.

(Jonathan at 00:26:49) Yeah. That's a great question. So people, some people do just take that repository and run it often. The most common reason for that is due to data sovereignty. So it'll be a national government in a market that we're serving.

(Jonathan at 00:27:01) Like, we love your product. We love working with you. We don't want to pay for your SaaS offering. We'd rather just host it on our own, and then we have the ability to support them to do it. One of the things that we're increasingly trying to make a more compelling argument, we're working with our hosting provider AWS on this as well, is to really do a better job articulating, you can host it on your own.

(Jonathan at 00:27:22) But to kind of replicate what Dimagi brings to the table, you have to do all these things. It's not just about turning a server on and running the code, but how are you going to secure it? How are you going to scale it? How are you going to keep it performant? And these are things that are kind of like—the same argument Amazon, Microsoft, and Google were trying to make to governments 10 years ago.

(Jonathan at 00:27:39) We're now trying to make to our customers today. And, um, it's a—in terms of what we think we bring that's hard to replace, it's really, um, we have a team of engineers on DevOps and code development production that are hosting multiple environments and multiple geos. We're SOC 2 compliant. We have a lot of our security infrastructure. All of that is really expensive to replicate.

(Jonathan at 00:28:00) It's not that you can't run CommCare on your own from a code base standpoint, but to run it on your own at that level of security, sophistication, and performance, that's the part that's really hard to replicate. And that's—usually we have a full SaaS offering as well, which the majority of our customers leverage and use in our environments.

(Joel Beasley at 00:28:18) That makes sense. So it's like the complexity of managing it all, because it's such a large platform is really difficult. And instead of hiring someone internally or multiple people internally to do that, they can work with you and that oftentimes is easier.

(Jonathan at 00:28:36) Yeah. And the analogy I give, you know, Pfizer has the new COVID drug coming out that's really effective and they're making the IP available. Well, like—

(Joel Beasley at 00:28:47) No way. Yeah. I didn't hear about that.

(Jonathan at 00:28:48) Yeah, so that IP is available in lower resource markets, but how are you going to manufacture it? So you have the blueprint for the drug, but the hard part is coming up with a clean manufacturing environment, you know, in that case. And similar to software, so much of software is becoming open source. And the movement for open source is exploding, and basically all of crypto code is open source in this next wave of Web3 technologies.

(Jonathan at 00:29:11) So there's increasingly tons of code, but you've got to manifest that code in a high scalability service. And that's becoming the more difficult skill set in some ways for some of these projects—is not the code, but it's the running of the code. You know, the manifestation of that code as a product that is usable to the end user.

(Joel Beasley at 00:29:29) Yeah, that's one thing that's been really interesting about running this podcast and learning from all these different tech leaders is there's so many tools out there to help just manage and implement the complexity of these various tools that are out there for running a business. Like AWS and cloud is so complex that we've had—I can think of like three different companies off the top of my head—that just help you manage your cloud strategy. Because that's not your company's core business, doing that. You want to focus on providing your CommCare platform.

(Joel Beasley at 00:30:14) You don't want to focus on how do I manage the cloud. But yeah, sorry, that's just something that's been on my mind.

(Jonathan at 00:30:24) Well, and I mean, it's exactly right. And I think pre-AWS, so we had a partnership with a database vendor that was based on CouchDB. So we were using that, and there was this great company in Boston called Cloudant that got acquired by IBM. And I remember they were one of our first big tech investments, you know. So we were running Couch on our own, and we're like, we can't keep doing this. We're having all these scalability and performance issues.

(Jonathan at 00:30:45) And there was this other really interesting hidden benefit that I hadn't anticipated. So we were paying them a lot of money. It was our biggest IT bill at the time. But because we were doing that, we were like, okay, we're not CouchDB experts. They are.

(Jonathan at 00:30:58) We're never going to think about this again. So also the problem is we have all these different technologies and cloud and these things. You're constantly like, oh, should that be the thing I do today? Let's go figure out how to optimize that problem. And you see this huge number of companies coming up that are super specialized, but it's great because you can pay that company to be what they're great at.

(Jonathan at 00:31:17) And not only are they better at it than you, but you also just remove it from your list of things to prioritize. Right? You're like, okay, that company handles this. I'm just going to wait for their product roadmap to do whatever it does, and I'm going to use it, and then I'm not going to think about this anymore.

(Jonathan at 00:31:30) And that's a huge benefit now in today's environment. This is something we spend a lot of time on also internally, is just getting that stuff off your plate. You know, I'm not going to think about this. I'm not going to think about that. And that alone just gives you this focus and ability to really let your engineering team be great at the thing you want to be good at and not worry about the stuff that you're not trying to be differentiated in.

(Joel Beasley at 00:31:49) Yeah. Well, speaking of focusing your efforts, I'm curious, how do you guys choose which projects to get involved with?

(Jonathan at 00:31:59) There's a complex answer I could give that we spend a lot of time thinking about. We have three priorities of the company: impact, team, profit, in that order. And that's roughly how we think about projects. So what is the current impact that we expect today and what's the potential leveraged impact tomorrow?

(Jonathan at 00:32:19) How good of a fit is it for a team? How much are we going to enjoy working with this client or government or use case? And then, obviously, can we afford to do the project based on the amount of resources available? And there's no clear criteria of, like, okay, it's got 100 units of impact and 50 units of team and 10 units of profit. But it's something that we look at a lot.

(Jonathan at 00:32:40) We are pretty risk-seeking in terms of being humble about what's going to end up making the most impact. So we still have a lot to learn. We've been at this for 20 years, and we keep running new stuff, new ways of doing our job, but more importantly, new ways of how community-based services could create outcomes. One of the interesting things in the U.S. market: healthcare has a huge impact on educational outcomes because for a lot of children, they're out of school because of health issues. And it may not be their own health issues, but a sibling's issues or a parent's issues.

(Jonathan at 00:33:11) So health and education intersect in this way you wouldn't necessarily expect if you're an education or health person, underneath the hood. And you can imagine, oh wait, if that's true, and engaging with a family as a whole family for healthcare is going to also improve education, what does that mean about how we could fund these projects? And what does that mean about the ROI of providing more community-based services for certain families? So these economic models, these financial models are something that I'm fascinated by and that we also love to take on projects that are pushing the envelope on what these economic drivers could look like for creating social impact. Because I think one of the embarrassing failures of technology for social impact is these are such high ROI returns for society, but they're not necessarily a specific ROI for one department in the government or one private sector player.

(Jonathan at 00:33:57) And technology could be such an amazing way to bridge that gap because you can carve out the impact that's created for the educational system, the impact that's created for the health system, get each of them to pay 50% of the project and do it. But it's pretty complicated. And you've got to have a friend like Dimagi who's really driven by the social outcome to be able to put some of those deals together. So that's really exciting for us. The other—you mentioned Cactus earlier—is looking at it.

(Jonathan at 00:34:24) We're often not the whole piece of the puzzle. Some projects, we are really a big part or the main part of the project, but almost always it's going to take a consortium of technologists, consortium of people who know what to do with that technology. And so we also look at the quality and skill of the other players that will be around those projects and try to select for the ones that really look like they have a successful team. And that's why when we got to work with Cactus on COVID and when we work with them elsewhere, having amazing partners like that is really something that also brings us to the table because life's too short, you know, to work with people that you don't want to. I heard a phrase that was like, if you don't work with somebody for 10 years, don't work with them for a day.

(Jonathan at 00:35:02) You know, just really pick long-term partners, and it's something that I think is really good advice. You know, whatever your cup of tea is, people have all sorts of preferences who and how they want to work, but finding those people is great because that's your career, that's your job. You've got to like the people you're working with.

(Joel Beasley at 00:35:20) So you started that by saying you have like a really complex answer to that. Does that mean you guys have teams internally that are working on these crazy complex calculations to prove the ROI of these community-based tech investments?

(Jonathan at 00:35:36) We do. We do. We have our internal research and data team that looks a lot at our impact, tries to do different analysis of, like, you know, per employee, is our impact going up? Per dollar coming into Dimagi, is our impact going up? What's our growth rate?

(Jonathan at 00:35:48) Like, do we think we'll sustain this? So there's a lot of internal research we do with our own data, but we partner a lot with academia on looking at the overall value of technology investments. A big group that we work with is the Community Health Impact Coalition. This is a consortium of many different technology organizations and nonprofits who focus on community healthcare workers in low-resource settings, and they do a lot of economic analysis. So one of the studies that WHO published back in 2015 showed that we have the ability to have a 10 to 1 return for community healthcare workers for investments.

(Jonathan at 00:36:21) Now, obviously that benefit's accrued to economic outcomes, health outcomes, a lot of different factors. But there's a huge amount of research that we partner on, and we're really excited to continue to push that part of the field forward as well.

(Joel Beasley at 00:36:34) Have you ever heard of a company called mPulse Mobile, like the letter M, Pulse Mobile?

(Jonathan at 00:36:39) I've heard of them, but I'm not too familiar with the work. But yeah, I've come across their name before.

(Joel Beasley at 00:36:42) Yeah, we interviewed them on the podcast like over a year ago now, I think. But they do like natural language understanding and like a conversational AI specifically for healthcare to help reach out via text message, like as you have experience with as well, to people specifically in lower-income communities to help get them to go to the doctor and realize their options to help improve the disparity in the health impacts of social situations.

(Jonathan at 00:37:15) Yeah, that's great. And, you know, you look at things like that that are, you know, they don't sound that difficult to do. And then it turns out they're very complex social engineering problems of, like, how you reach the target audience you're trying to reach. But the analogy I've been giving more recently—and I don't know how much you access the healthcare system here in the U.S.—but we spent $20 billion plus dollars, right, doing all this digital work over the last 10 years in health IT. I still can't pull up my phone, book an appointment with my doctor on any reasonably pleasant user experience.

(Jonathan at 00:37:44) Right? And I'm like, there's so many things you could do to make healthcare more accessible to people who have busy jobs or might be working two jobs. Or, you know, a lot of this is like, it's not necessarily health-seeking behaviors. Like, the system is incredibly annoying to interact with. The user experience of our healthcare system is terrible.

(Jonathan at 00:38:01) And so there's all these nudges you can provide, but there's also just amazing ways for technology to potentially create a better user experience. Not the classic user experience definition of the software user interface, but the user experience of that human interacting with the system. And technology is just one component of that, but it's such a great opportunity, in my opinion, for disruptive innovation and something that the United States Digital Service does a lot of and looking at these different areas. But absolutely, I think those types of technologies, like you mentioned with mPulse, could go so far because the user experience in some of these areas is so bad if you interface between a citizenry and the public services the government's trying to provide.

(Joel Beasley at 00:38:41) Yeah. And you're circling back to how so much impact can be found in areas where you don't necessarily—like when you're talking about health and education impacting each other in the way they do, thinking about how just the UI of an app for a healthcare provider directly impacts the health of the community that they serve because it makes people more or less likely to seek healthcare. That's crazy.

(Jonathan at 00:39:11) Yeah.

(Joel Beasley at 00:39:12) But, okay, so before we get into wrapping up, I want to talk to you a little bit about leadership because that's something that we like talking about on the podcast. Is that cool?

(Jonathan at 00:39:21) Absolutely.

(Joel Beasley at 00:39:22) Cool. So, actually recently we got to interview Zscaler's CTO. And that was, I thought, a super cool episode because going in, I knew that they were a huge security company and big on zero trust and stuff, but I had no idea they actually have like 150 data centers strategically placed globally that help to process data in a way that decreases the to-and-from time. Instead of like sending your data across the country, it just goes like a couple cities over to their center.

(Joel Beasley at 00:39:53) And in doing that, they've really smartly thought about security from the ground up. But anyway, their CTO, Amit, was also just an excellent leader. And he was talking about how he's a big believer in the power of marginal improvements. Like if you improve at something by 1% a day, that becomes an absolutely huge amount of progress over the course of the year. So that's been on my mind and I'm curious, how do you carve out time for self-improvement in your schedule and what do you do for that?

(Jonathan at 00:40:26) So I'm very fortunate that I have an amazing chief of staff, Jillian. And she and I talk a lot about how to make sure the schedule I have is balanced in terms of internal one-on-one meetings, external meetings, working with my direct teams and indirect teams. And so I try to really make sure I have a healthy diet of a balance of meetings. Most of my calendar is pre-programmed every week. But I do a lot of reading, and that's where I find I have the most amount of time to self-reflect at night when I'm reading a book. Right now I'm reading John Doerr's "Speed and Scale" about climate change and some of the work he's been doing. And when you're reading about these different areas that you may not be totally immersed in every day, it often brings you back to your own company or your own industry. So you're reading a book or a science fiction book or something else that has nothing to do with your sector, but it starts connecting the dots in your head. And then you can take that and say, okay, what did I take from that, and how do I want to improve how the company's operating or my own skills? For me personally, you know, I've been at the helm of Dimagi for 20 years.

(Jonathan at 00:41:25) I've seen it go through multiple iterations, but I've been basically the leader of a tech company my entire career. And so with that comes a lot of blind spots on what it's like to join an organization, you know, what it was like to join us when we were 10 people or what it's like now when we're 250 or, you know, how we interact. And so I also try to do as good of a job as I can being humble and curious, and I don't know that I'd be accused of being too successful at that on some days. But curiosity, I think, is one of the best ways to do self-improvement. You know, everybody has meetings they don't want to be in or conversations they don't want to have, but coming in with that sense of curiosity to say, you know, maybe I'm annoyed I have this meeting, or maybe I don't like this topic, or maybe I'm kind of frustrated that my team member didn't do X, Y, and Z.

(Jonathan at 00:42:08) But you can always be curious about why are we in this situation? Why am I frustrated with you? Why did you not do the thing I thought was obvious that you should have done? And I don't always successfully bring that curious mindset into meetings, but when I do, you know, I had a leadership coach a couple of years back and she was like, you know, before every meeting, spend 60 seconds, think about what you want this meeting to do, what do you want to get out of it, how you want the meeting to play out. Like, run through the conversation you're about to have in your head. And it was such an eye-opening way to do self-improvement of kind of going into those meetings, being much more self-aware of, like, how do I expect this meeting to go?

(Jonathan at 00:42:41) How did it go? And so I think that's something that doesn't require a huge amount of time. I mean, a lot of leaders are quite busy, but for self-improvement for me, it's really trying to figure out how you want to show up and what you can do to do that. And when you're constrained for time, a lot of that can be thoughtfulness in how you're going into meetings. As I said, a lot of my time spent in meetings and so that's the unit of work that I think of when I think about how to do self-improvement.

(Jonathan at 00:43:05) But I'd say curiosity is such a big factor of being able to improve yourself.

(Joel Beasley at 00:43:10) Yeah. Being intentional about just setting goals for meetings, like you're saying, is actually something that I've been thinking about recently too, because I found myself kind of falling into a dip where I'd be showing up to a meeting and figuring out on the fly why we're even here, who am I even talking to? And it was just like a loss of productivity that I wasn't okay with. So like you said, just taking that time before each meeting, I started setting little like a reminder five minutes before each meeting that just automatically goes off on my calendar to like stop what—that's when you have to stop what you're doing. You're not allowed to just work until the start of the meeting anymore.

(Joel Beasley at 00:43:55) And take that, for me it's five minutes, to set up in my brain my mental model of how it's going to go, who I'm talking to, so I can come in more educated on the situation and get more out of it.

(Jonathan at 00:44:11) Yeah. And I come from an engineering background. A lot of the people that we recruited early on at our company were coming from engineering backgrounds. So it's very easy to kind of be like just schedule and routine, show up at the meeting, what are we doing, interact on it. And I think, you know, given how much time is spent both in meetings, but also just how much of work is now team-based, you know, collective progress, getting good and better at that is something that I think is well worth your time or anybody's time to constantly be trying to improve. Like, how can I run a better meeting? How can I make sure everybody's getting the most out of this meeting? How can I make sure different personality types can be successful in meetings with me? You know, I'm a really big extrovert, so I don't mind talking. Well, that can be really counterproductive if I want to hear somebody's voice who's very introverted, may not just volunteer that.

(Jonathan at 00:44:56) And so also how you manage multiple personality types to be contributing during meetings or getting successful closure on topics, that's a skill that I'm constantly trying to get better at. And I definitely would not say I excel at it, but it's something that I want to get better at.

(Joel Beasley at 00:45:12) Yeah. Something I've heard that was like so simple, but kind of blew my mind was just kind of getting rid of the sense of urgency to come to a conclusion by the end of the meeting. And that it's okay to just ask the question in the meeting and give everyone time to start thinking about it and come to the answer in a few hours or tomorrow. And that's especially good for situations where you have a mix of introverts and extroverts, because yeah, like everyone's gonna have good ideas, but some people are just gonna speak first or louder or both.

(Jonathan at 00:45:53) Yeah. Yeah. Yeah. Well, as you said, I think that's a really good point and great advice for your listeners because I think if the meeting's easy to come to a conclusion, then like you probably didn't need to have the meeting in the first place. Write down the answer, send it around, get feedback when you're done. Most of the meetings we do are really what in the leadership curriculum was called adaptive problems. So you're even struggling to define what the problem itself is. Once you define the problem, the solution often becomes easier and might be a better written document so people can review it on their own time and provide input. So a lot of the meetings that I have are more, exactly to your point, the point is not to conclude the meeting with a solution, it's really to better define the problem. You know, what is going on here? Why do these two teams not feel like they're clicking? Or why is your go-to-market not working? But if it's a very clear known problem, often we don't need to have a meeting about the solution. Somebody writes up the solution and we're like, is this it? Yes or no? And critique it on paper.

(Joel Beasley at 00:46:44) Yeah. Yeah. That makes a lot of sense. So I want to ask you a little bit about—because you said you've been with the company for twenty years. You've grown it to two hundred and fifty people now, which like, wow, congrats. That's crazy. And to my brain, our company is like six people right now. Anyway, what were some of the biggest challenges you ran up against in scaling and what did you learn from those?

(Jonathan at 00:47:09) Oh, man. That's a great question. I think I was constantly surprised at when we outgrew processes and it snuck up on us, you know, as we grew or when we put something in place that worked really well in a surprising way. So when we were at about twenty-five, I formalized the management team, and it was a leader of five or six people, and we would meet weekly, and that worked amazingly well. And so from twenty-five to fifty, we were just humming along. And then we get to fifty, and we're like, everything just like doesn't quite feel right anymore. What's going on? It's like, you know, our structure just kind of was outgrown. We had started to go into new geographies. We needed a new management team structure. And then we did some iteration and kind of figured that out and went from fifty to a hundred. And then we got to a hundred, and we're like, stuff doesn't feel right again. And so we kept, you know, it'd feel great, and then we'd keep growing, and then it'd feel like it didn't work anymore. And we went through a really big reorg going into 2020, moving from a mostly functional approach to a much more business unit oriented approach. And we were incredibly fortunate to have done that transition because if COVID had hit, we would not have been able to kind of take forty people out of the existing divisions and put them on COVID without kind of collapsing under our own complexity. And so we had gotten really fortunate that we had done that reorg just in time to be able to be prepared to move a bunch of people into what was effectively a new division overnight and be able to support such a massive public health response both here in the US and globally across so many different use cases. And so that—it's the org structure. And they always talk about how leaders aren't necessarily the person who takes a company from zero to ten or ten to fifty or fifty to a hundred isn't the person to keep carrying you forward. But that's also very true of the org structure. I didn't appreciate how immensely important the org structure was going to be towards how you could scale, where you could grow, and then when you would outgrow that org structure. And that's been one of the really surprising things as we've kept growing.

(Jonathan at 00:49:08) But one of the things that I'm super proud of has been our culture has been pretty solid and improving this whole time, in my opinion. Obviously, we had a different culture when we were ten people, but the core elements of it as we've grown and being open and candid and socially driven and really wanting to work with amazing people, that's been such a big draw. And we would not be anywhere, you know, at Dimagi without having built such an amazing team that really is all pulling together for each other and for our users. And so that was probably the most critical success factor of success as we scaled was we knew who we are, and we knew what we wanted to do together. And having that purpose together made it easy to overcome some of the challenges as you like outgrow your organizational structure and realize you screwed something up or need to change something. Having that energy to do it is really dependent on having an aligned team that's all trying to get to the same place.

(Joel Beasley at 00:49:58) Yeah. And I mean, it sounds like your culture is simultaneously like a competitive advantage and also a self-perpetuating cycle. Because I mean, something I've been hearing a lot about is how with like so much of the engineering jobs being remote today, wages have somewhat flattened across like dev jobs and whatnot. And the differentiating factor in terms of hiring has become like your mission and the culture at the company. And you guys being impact focused, I mean, it sounds like you're gonna win a lot of the time on that front and by that token be attracting people that are likely to help keep your culture strong. So, I mean, just looking at it from how I see your company from how you described it today, it makes sense that your culture has grown so well with the scaling.

(Jonathan at 00:51:02) Yeah. It's exactly as you said, you know, it's got a nice self-perpetuating culture. I think some cultures are really hard to keep intact, and ours is actually really easy to keep intact because it attracts a certain type of person who's mission driven, who's really talented, wants a team-based approach. And so you don't get people who are super egocentric or not into the impact, but they just wouldn't take the job. And it'd be really obvious during the interview process to both parties that it wasn't a fit. And so we're really fortunate for that, which isn't to say you don't have to work at it and improve it. And we're doing a lot of work to do that right now, but it's something that has been a huge differentiator. And as you said, it's also a great recruiting tool because it is so competitive now to hire engineers in all markets. You know, we have offices in India and Cape Town, and those markets are going crazy. The US market's crazy. So it's really, really hard to hire, and our West Africa team as well. And having this impact approach is a key differentiator for us and allows us—it's not for everybody. You know, it's not whatever engineer wants to do, but for those who do want to do it, we're usually a great mutual fit.

(Joel Beasley at 00:52:04) Before we wrap up, is there anything that we want to make sure we get out to the world that we didn't get to touch on today or any extra shout out you want to make?

(Jonathan at 00:52:12) Yeah. I'd love to, you know, for the listeners, I know you have a lot of technologists and engineers. The shout out I'll give is saying, working in fields that have public sector engagement or public services or interacting with these like really big bureaucratic problems, it can be super frustrating, but it is so rewarding because if you crack it, the impact you're gonna have is huge. So I just really encourage everyone to think about the contribution they can make in those different sectors because technology is changing at such a crazy rate. The gap between what people who understand technology can do and what people don't is unfortunately growing. And I think technologists themselves can play a role trying to fix that and contribute more to closing that gap. And so I just encourage everybody who's listening to really think about that and how they might want to spend, not all of their career per se, but a portion of their career or side time, thinking about how to use technology to really improve public good and social impact. And it's something that I've found huge purpose and love in my life, and it's something that is definitely worth trying out even if it ends up not being fully for you. And yeah, just leave that with your listeners.

(Joel Beasley at 00:53:20) Thank you so much for listening. And if you found this episode useful, please share it with a friend or colleague who you think would get value from it. And if you have topics that you'd like to hear discussed on the podcast, either add me on LinkedIn or send me an email [email protected]. Every time I get an email or LinkedIn message, it absolutely makes my day and inspires me to keep going.