by Mala Kumar, GitHub Open Source for Good program
My name is Mala, and I lead a program at GitHub called Open Source for Good under our Social Impact team. Before joining GitHub, I spent the better part of a decade wandering around the world designing, managing, implementing and deploying tech for international development (ICT4D) software products. Throughout my career, I was told repeatedly that open source (OS) would revolutionize the ICT4D industry. While I have indeed worked on a few interesting OS products, I began suspecting that statement was more complicated than had been presented.
Indeed, after joining GitHub this past April, I confirmed my suspicion. Overall, the adoption of OS in the social sector – defined as the collection of entities that positively advance or promote human rights – lags far behind the commercial, private sector. Why? You may ask.
Here’s one hypothesis we have at GitHub:
After our team’s many years of experience working in the social sector and through the hundreds of conversations we’ve had with fellow social sector actors, we’ve come to believe that IT teams in the social sector have significantly less decision making power and autonomy than commercial, private sector IT teams. This is irrespective of the size, the geographic location, or even the core mission of the organization or company.
In other words, decision-making power in the social sector does not lie with the techies who typically have the best understanding of the technology landscape. Rather, it’s non-techies who tend to make an organization’s IT budgetary decisions. Consequently, when budgetary decision-makers come to GitHub to assess OS tools and they see something like the below, a GitHub repo, they have no idea what they’re seeing. And this is a problem for the sector at large.
We want to help bridge that gap between private sector and social sector tech development. The social sector is quite large, however, so we’ve had to narrow our focus. We’ve decided to target the social sector’s M&E vertical. This is for several reasons:
- M&E as a discipline is growing in the social sector
- Increasingly more M&E data is being collected digitally
- It’s cross-sectoral
- It’s easy to identify a target audience
- Linda is great. ☺
How We Hope to Help
Our basic idea is to build a middle “layer” between a GitHub repo and a decision maker’s final budget. I’m calling that a MERL GitHub “Center” until I can come up with a better name.
As a sponsor of MERL Tech DC 2019, we set up our booth smack dab in front of the food and near the coffee, and we took advantage of this prime real estate to learn more about what our potential users would find valuable.
We spent two days talking with as many MERL conference attendees as we could and asked them to complete some exercises. One such exercise was to prioritize the possible features a MERL GitHub Center might have. We’ve summarized the results in the chart below. The top right has two types of features: 1) features most commonly sorted as helpful in using open source and 2) features potential Center users would actually use. From this exercise, we’ve learned that our minimum viable product (MVP) should include all or some of the following:
- Use case studies of open source tools
- Description of listed tools
- A way to search in the Center
- Security assessments of the tools
- Beginner’s Guide to Open Source for the Social Sector
- Installation guides for listed tools
Aggregation of prioritization exercise from ~10 participants
We also spoke to an additional 30+ attendees about the OS tools they currently use. Anecdotally, mobile data collection, GIS, and data visualization were the most common use cases. A few tools are built on or with DHIS2. Many attendees we spoke with are data scientists using R and Python notebooks. DFID and GIZ were mentioned as two large donor organizations that are thinking about OS for MERL funding.
In the coming weeks, we’re going to reach out to many of the attendees we spoke to at MERL Tech to conduct user testing for our forthcoming Center prototype. In the spirit of open source and not duplicating work, we are also speaking with a few potential partners working on different angles to our problem to align our efforts. It’s our hope to build out new tools and product features that will help the MERL community better use and develop OS tools.
How can you get Involved?
Email firstname.lastname@example.org with a brief intro to you and your work in OS for social good.