Farming Data


A large proportion of the population in developing countries are smallholder farmers. Selling excess produce accounts for at least 50% of their family income. They are unfairly disadvantaged and face exploitation as they sell comparatively small volumes and are geographically isolated. The vast majority of smallholders, globally, have access to 2G ‘brick’ phones and local buyers have access to smartphones. We have developed a digital farmer-to-buyer transaction platform as a scalable solution to the market access, data generation, and data delivery problem in smallholder agriculture. The platform is user-centric, built for the technology available to smallholders and local buyers. We will be undertaking our proof-of-concept pilot in Colombia for 12-18 months from 2019-20. We have secured collaboration agreements with three major agricultural organisations and a telecoms company in Colombia to pilot with ~1000 farmers. In addition, our team has extensive experience in sub-Saharan African agriculture, and we are actively exploring pilot projects in East Africa. Our six team members met as PhD students at the University of Cambridge. We cover a broad range of skills, from predictive modelling and data analytics within agricultural systems, project implementation in developing countries, machine learning, and agricultural policy.

Currently, the majority of smallholder agricultural activity is informal, which means there is no centralised, transparent record. Digital platforms bring informal activity ‘online’. Providing a farmer-to-buyer digital transaction tool simultaneously generates large amounts of agro-economic data, and has the highest probability of adoption, as it is connected to the most direct incentive in the system: increasing household income. Moreover, whilst farmers are not paid a fair price for growing additional produce, there is no incentive to increase farm productivity. Therefore, we view market access as the first order problem that must be solved as productivity boosting tools, such as fertilizer or improved cultivars, depend on it making financial sense for farmers to grow more.

We believe this problem is unsolved because you cannot efficiently communicate adequate spatial information with ‘brick’ phones alone. However, uniquely, we hypothesise that this problem can be solved without both sides needing smartphones. We have designed an end-to-end tool that manages the entire transaction from the point of connecting farmers and buyers to facilitating payments, where the buyer is a smartphone user and the farmer can use a ‘brick’ or smartphone. We take advantage of phone calling where possible to minimise barriers to entry due to illiteracy. Importantly, by facilitating payment, this allows a small transaction fee to be applied, which provides the mechanism for the platform to be sustainable and drives expansion to new users.

Aggregating transaction data on the platform generates highly spatially and temporally resolved dataset on the crops farmers are growing, location-specific market prices, and market connectivity.

The team

Dr Jacqui Poon coordinates business and product development and oversees partnership engagement. She holds a doctorate in plant pathology (2016) from the University of Cambridge and has a background in cassava post-harvest deterioration (MSc 2012). She believes that enabling smallholder farmers to access better markets makes a positive impact on their livelihoods.

David Godding manages tech development of our platform and develops algorithms for our data analytics. He is a PhD candidate in the Epidemiology and Modelling group in Plant Sciences at the University of Cambridge, where he models the spread of cassava diseases.

Dr Paul Bergen has a background in bacteriology and infectious diseases. He holds a doctorate from the University of Cambridge (2018). He sees that health, food security, and family income are intricately linked, so he believes that giving smallholder farmers the tools and data to reach better markets will greatly improve their quality of life.

Greg Wilsenach works on the system design and tech development of the platform. He is a PhD candidate in Computer Science at the University of Cambridge.

Kelvin Mei works on platform development. He’s currently a PhD candidate at Princeton University, carrying out his research in Theoretical Physics at CERN.