By Gangandeep Nanda and Kabir Sethi

The impact of India’s edtech unicorns has turned out to be just as fictional as their namesake. Initial excitement around how remote education tools would promote greater access has died down. A Dalberg study on education during Covid-19 showed that only the top slice – largely, urban and private school students – benefited from the edtech boom of 2020-22. In India, where there are 87% low-income households, provisions for last-mile access are crucial to the success of innovative technologies.

Artificial intelligence (AI), which is finding practical application in the education sector, could face a similar trajectory. Alarm bells are being sounded over how tech will ‘outsource thinking’ and kill creativity. Yet AI’s integration in education is inevitable. If done right, it has the potential to undo the mistakes of the past and bring a second revolution to India’s edtech sector.

The first wave fell short

The first Indian edtech wave, which peaked during Covid-19, exposed how underprepared we were for a large-scale shift to remote learning. As schools were putting together makeshift remote learning environments for students, the market soared. The demand for edtech products expanded, decisions were driven by market expansion, and R&D took a backseat. Only a small group of impact funders and investors used learning outcomes as a metric for designing products. Instead, it was companies that were able to project rapid growth that raised most of the funding. Growth predictions turned out to be wrongly estimated – and smaller players got left behind. Outside of a few such as Educational Initiatives and ConveGenius, most impact-focused players continue to operate at a small scale, despite their demonstrated evidence of delivering high impact.

As a result, edtech solutions that emerged were not localised to low-income, low-infrastructure and rural settings. For instance, the average annual price of e-learning products dominating the market begins at Rs 20,000 for primary and secondary grades, going up to several lakhs for higher education. This creates a barrier for low-income households at the point of entry itself. Features such as relying on a ‘one device per child model’ or programmes requiring high-speed internet deepen the challenges. Meanwhile, the content itself is tailored for English-speaking, urban audiences and relies on examples and imagery, such as travelling in airplanes, that alienates low-income and rural students.

How to make second wave work?

If we want to prevent AI from falling into this trap, the social impact community needs to have a greater share of voice. This includes non-profits who play a key role in reaching underserved communities and work collaboratively with public-school systems, as well as philanthropic funders who are not only at the cutting edge of thought leadership, but also take an impact-first lens.

Ecosystem players can agree on a set of fundamental principles for responsible and ethical use of edtech. These would serve as guidelines on things to consider as an edtech product makes its way from lab to market. Edtech providers need to be cognisant of the datasets based on which they design and build their products. For instance, a product built on representative and India-specific data would have greater local resonance and uptake. Further, use of robust uptake and impact data can help build a culture of evidence, where tracking learning outcomes and using them as a key metric of decision-making becomes central to edtech design and delivery.

Affordability is another consideration. Social impact players can help subsidise operation costs and this can lower the price of products and services. We’ve seen that working with governments is the best way to scale and reduce cost of delivery. The social impact sector, with its demonstrated history of working efficiently with governments, can leverage existing relations and deepen the reach of quality education. The sector is best equipped to safeguard the interests of the most important stakeholders – students, parents and teachers. The first edtech wave rode on the hubris that digital classrooms would replace physical ones. This time around, players have a chance to work towards supplementing educators’ efforts rather than attempting to replace them.

Finally, any attempts at collaboration need to be attuned to maximise impact, with an eye on reducing the information gap, bringing down costs and thus market price, and increasing awareness on risks and safeguards. Concentrated, coordinated and time-bound efforts that bring big and small players on the same page can help get the basics in order. Developing these metrics and measures to hold edtech actors accountable is the only logical way forward.

Authors are associate partners at Dalberg Advisors, where they lead the education practice.