Why We Invested: Siyavula

October 9, 2017

“In many countries and communities, learning isn’t happening,” World Bank President Jim Yong Kim notes in the introduction to the new World Development Report (WDR). UNESCO’s Institute for Statistics estimates that 617 million children and adolescents are not reaching minimum proficiency levels. This equates to 58 percent of the world’s school-age children on average, with that ratio rising as horrifically high as 88 percent across sub-Saharan Africa.

The WDR also reaffirms a clear tenet in Omidyar Network’s education strategy: the untapped potential of technology to help improve learning outcomes, as long as it is efficacious, leveraged in results-driven, cost-effective and gap-narrowing ways, and complemented by the human support and teaching that every child is entitled to.

On the same day that the WDR came out in Washington, D.C., one of our ed-tech investees, Siyavula Educationunveiled its exciting #1MillionMaths campaign in Johannesburg in partnership with Google.org. The partnership will massively expand the reach of Siyavula’s high-quality adaptive math platform, reaching hundreds of thousands more low-income students from South Africa to Nigeria.

The truly democratized, equitable offering of quality ed-tech to all students has been a core mission of Siyavula since it was founded, and they have taken many deliberate design decisions to achieve this. We first invested in Siyavula in 2014, and this partnership is but the most recent demonstration of the exciting potential we saw in their approach and their team years ago.

When we were first introduced to Siyavula in April 2013, they had a strong curriculum and showed early promise — but we did not invest initially because our belief was that good-quality content, while necessary, would not be sufficient to drive learning gains. One year later, Siyavula’s co-founders shared their learnings with us. They had noticed that while the content was indeed being used, in order to really drive learning gains they needed a world-class, adaptive, machine-learning based system that sat on top of this content — and that was what they had started to build.

The Siyavula Practice technology they built uses powerful and sophisticated machine learning to meet all students where they are, and then helps take them on a progressive learning journey. It uses a mastery-through-practice approach which maintains a healthy balance between consolidating current skills and being stretched to the next steps. In this way, students are able to learn and progress at their own pace, often without explicitly realising their own progression until later (when they notice their exam scores and improved self-confidence!).

Early on, the company was using this Siyavula Practice in 50 schools with roughly 2,500 learners. Many of these were more advantaged schools and learners, but Siyavula was also beginning to expand to lower-income students — and finding that the technology was helping students across the ability curve and across very different settings. This was the signal we needed, and in 2014 we invested.

Siyavula builds its technology to support teachers, not to substitute them. They provide dashboard analytics that allow teachers to see where each of their students is stuck — so they can provide laser-focused assistance. Teachers are innovating with this technology in a host of different ways — whether using it for small-group work in classrooms or for homework assignments.

Additionally, other entrepreneurs have used the technology to develop entirely new business models, such as efficacious lower-cost after-school support to serve middle- and lower-income families.

Finally, we invested because the Siyavula team built their technology to ensure that it would not be gap-widening. This included making one vital decision right from the start. They built their technology so that Siyavula Practice could be used even on an old-fashioned feature-phone, without heavy data and video requirements. This may sound quaint, but it is vital for reaching far-flung or poorer students.

A recent independent survey done by Acumen’s Lean Data team found that the income distribution profile of Siyavula Practice learners is almost identical to the South African national curve: It is a truly democratized high-quality offering, with 30 percent of its customers living below internationally recognised poverty lines.

Siyavula is improving learning outcomes at scale; from 2,500 learners three years ago they are projected to reach more than 300,000 learners next year with their adaptive technology alone. The Lean Data survey found that 97 percent of learners reported that it had helped their grades improve, and 74 percent said it had improved their quality of life.

Siyavula is home-grown proof of world-class adaptive education technology born from the Global South — for the more complex secondary grades and across the socioeconomic spectrum.

It has been thrilling to see Siyavula develop, and the impact and feedback from the learners is deeply energizing and humbling. But they haven’t, and indeed couldn’t have, done this alone. Watching the way so many other organizations like Vodacom, MTN, Sasol Inzalo, Zenex, MSDF, ITSI, Google, various central and provincial governments, and others have also engaged has been wonderful.

It underscores the point that while the learning problem facing us all is a massive one, it is indeed possible for collective action to emerge to help solve this problem. We are grateful to be part of this, and optimistic that together we can all help improve this awful situation.

Kevin Kent Mr at Forum Forge

I like reading articles like this. Well done to all. There is still a very long road ahead as South Africa has over 12.8 million learners. The government and Department of education also arnt helping as they keep dropping the pass mark of specific subjects in order to pass students and prevent a back log. I think math is sitting at 30% pass mark. Unacceptable. but they dont want to keep students back. the pass rate makes the department look good but isnt doing any of the students a favour as they are setup to fail with their tertiary education. We are busy developing a smart school app that will digitise all teachers work within the schools in SA (to start). this will also involve parents more in the learning process of their children. We have been working on this for over a year and hope to launch soon. This app is a disruptive app and will be able, over time to determine each individuals strengths and weekness allowing teachers and parents a pro active approach to their childs education. We also have plans down the road for a homeschool top box set that runs through satelite (data very expensive and not available everywhere). We currently have UNISA (University of SA) doing a paper on our app and technology in the classroom. We have partnered with Pearson education on the app to make teacher learning skills more widely available to teachers currently and learners down the road. Well done guys, keep up the good work and I am pretty sure that our paths will cross down the road somewhere. Together we can change the way our children learn and give all an equal opportunity.

Be the first to comment
It looks like there's some information missing
By clicking, you agree to the Terms and Conditions


New Competition Wants to Bring Ethics to Undergraduate Computer Science Classrooms

Tina Nazerian at EdSurge covers the launch of the Responsible Computer Science Challenge, the latest exciting initiative incubated at the Tech and Society Solutions Lab.



Silicon Valley Funders Issue Ambitious Challenge to Embed Ethical Thinking into Computer Science Curriculum at Colleges and Universities Nationwide

Final projects will be judged by an external review committee of academics, tech industry leaders, and others.