Agriculture has long been viewed as a fertile ground for technological innovation. Over the years, governments, donors, development agencies, and private companies have invested heavily in ICT-enabled solutions to address persistent challenges such as low productivity, limited market access, climate risks, financial exclusion, and weak extension services. 

Drawing on my work in ICT4D, D4D, D4Ag, National Agricultural Data Infrastructure (NAgDI), and National Data Exchange Infrastructure (NDEI), I argue that digital technologies may emerge, peak, and fade, but data remains the enduring strategic asset in agricultural transformation. To benefit from the current advances in Artificial Intelligence (AI) and Digital Public Infrastructure (DPI), countries should prioritise investment in trusted and reusable data rather than follow technologies blindly.

Over the last decades, agricultural innovation has progressed from radio, telecentres, and offline information tools to mobile services, digital platforms, remote sensing, AI, and Digital Public Infrastructure. Each wave has promised transformation, yet many solutions have been replaced, failed to scale, or disappeared when funding ended. The history of ICT4Ag and D4Ag reveals a recurring pattern. New technologies emerge with considerable enthusiasm, attract substantial investment, and are often promoted as transformative solutions. Over time, many become obsolete, are replaced, or fail to achieve widespread adoption. The agricultural sector has therefore experienced multiple cycles of technological enthusiasm and decline. 

Each technological wave has promised to revolutionise agriculture (smallholder agriculture). Some of these technologies have delivered value during their time; many have been overtaken by newer innovations or failed to scale sustainably.

Yet throughout these cycles of technological change, one element has remained constant: the need for trusted, accessible, interoperable, and reusable agricultural data.

This raises a fundamental question: 

If technologies continually change, what remains constant in supporting agricultural transformation?

The earliest ICT interventions in agriculture focused primarily on information dissemination – radio programmes, community telecentres, among others. These technologies extended the reach of extension services but were largely one-directional and limited in their ability to collect feedback from farmers. Although many of these tools have either disappeared or become marginal, the information they carried remained valuable. 

The value was not the technology itself; it was the agricultural knowledge and information being communicated.

The rapid expansion of mobile networks transformed ICT4Ag into SMS systems and interactive voice response (IVR) platforms that supported mobile-based extension services, farmer management systems, e-voucher programmes, digital financial services, e-commerce platforms among others. Hundreds of donor-funded projects emerged across Africa, Asia, and Latin America. However, these platforms created new silos of duplicate farmer registries, fragmented value-chain databases, proprietary information systems, and lack of interoperability. The challenge was no longer a lack of technology but the inability to exchange and reuse data across systems. 

As new applications emerged, older platforms often lost relevance. Many become unsustainable once donor funding ended. While specific applications disappeared, the farmer profiles and registries, market data, weather records, and advisory content remained critical assets. 

The current generation of digital agriculture technologies includes Internet of Things (IoT); Artificial Intelligence (AI); Big Data Analytics; Satellite Earth Observation; Digital Twins; Blockchain; Robotics; Precision Agriculture; and Digital Public Infrastructure (DPI). These technologies promise unprecedented capabilities for data-driven decision-making and automation. Yet history suggests that many of today’s technologies will eventually be replaced by new innovations. 

My experience over the past four decades of agricultural information and knowledge management reveals an important insight – Technologies are temporary. Data is enduring. Every technological generation has depended on User data (farmer profiles/registries, data on research, extension, input networks, polygons of fields, etc.); and Content data (land data; weather data; soil data; market data; trade data; production data; financial data, etc.). 

When technologies become obsolete, these datasets remain valuable. In fact, many modern AI systems derive their power not from novel algorithms but from access to large volumes of quality data. Without trusted and accessible data: AI cannot learn; digital services cannot function; precision agriculture cannot operate; supply chains cannot optimise; policymakers cannot make evidence-based decisions. 

The next phase of agricultural transformation should focus on building National Agricultural Data Infrastructures (NAgDI) (not just technical data platforms). As argued in my Commonwealth digital agriculture work, digitalisation is not merely about technologies but also about the data infrastructure that enables those technologies to function effectively.  

The long-term value of digitalisation lies not in the technologies themselves but in the data they generate, exchange, and utilise.

A simple advice for national governments is not to chase each new technological trend, but rather to build robust national infrastructure for their agricultural data. Neither DPI nor AI is the solution to food systems transformation in countries. Private sector companies will build their businesses on them. But for national governments, your asset is the trusted, accessible, interoperable, and reusable data can support the private sector services – AI, digital services, precision agriculture, resilient supply chains, and evidence-based policymaking across successive generations of technology. And to take sovereignty over this asset, countries need to build an infrastructure for it.

Digital technologies will come and go; data is the enduring asset on which resilient, productive, and inclusive agricultural systems can be built.

Lear more about how to invest in National Agricultural Data Infrastructure: https://thecommonwealth.org/national-agricultural-data-infrastructure