Steve Mellgren, Senior Solutions Architect with SAS’ AI Ethics, Governance and Social Impact Division
A farmer working a small plot does not have much room for a bad decision. Water may be limited, labour has to be used carefully, and money spent on seed or fertiliser cannot simply be written off if a crop performs poorly. That is why better information can make such a difference. It cannot remove the uncertainty that comes with farming, particularly as weather patterns become harder to rely on, but it can help farmers make decisions with more evidence behind them.
This year’s World Food Day theme, Innovate today. Nourish tomorrow., reflects the need for innovation that reaches the people who produce food and strengthen the systems around them. The Food and Agriculture Organization of the United Nations (FAO) places particular emphasis on innovation that makes agrifood systems more inclusive, resilient, and sustainable. For me, a recent SAS Data for Good project with South African micro-farmers shows what that can look like in practice.
A small plot leaves little room for guesswork
Working with dataDecisions.ai and The Dream, SAS analysed information collected from micro-farms near the Cradle of Humankind. The farmers were producing food on small plots, often with limited access to technology and reliable markets.
The information already available told us quite a lot. We had records of the crops farmers were growing, when produce went to market, and what it sold for. We could also work with the size of individual plots and compare how crops performed.
Analytics allowed us to turn that information into insights. We looked at the seasons in which different crops were planted and whether that aligned with the best growing periods for those crops. Because we knew the size of the plots and the value of the produce sold, we could also see differences in the return farmers were achieving from similar areas of land.
The analysis helped bring out patterns that were difficult to see in the raw records, giving farmers more useful information to work with when making planting and production decisions.
When the weather becomes less predictable
That becomes particularly relevant when farmers face changing weather conditions. El Niño is now firmly established and the World Meteorological Organization expects it to strengthen further, with the event likely to persist into early 2027. Its effects will not be the same everywhere. Some regions may face greater drought risk while others experience heavier rainfall, which is precisely why local information remains so important.
In our micro-farmer project, we were able to bring precipitation information into the analysis and compare it with crop performance. We did not build a long-term climate model or attempt to forecast an El Niño event. The project was limited in scope and used the data available to us.
This project showed where this approach could go next. If a farmer has several seasons of crop information and can see how particular crops performed during a drier or wetter period, that history becomes useful when similar conditions are expected again. It provides another piece of evidence when deciding what to plant and how to use scarce resources. For a micro-farmer, reducing some of that uncertainty can be valuable in itself.
Growing the crop is only part of the job
There was another part of the project that I found just as important as crop performance. The partner documented how produce was grown as it moved from planting towards market. That record created greater visibility into the production process and helped farmers participate in a more coordinated route to market.
Without that connection, a small farmer may grow produce and still have no certainty that it will sell. Something as simple as relying on a roadside stand can leave food unsold until it is no longer usable.
Better data can help connect the production decision with what happens afterwards. Farmers have a clearer idea of which crops have performed well and what those crops have fetched at market. The wider system also gains more visibility into where the produce came from and how it was grown.
That can improve the farmer’s income prospects while reducing the chance that food produced with scarce resources simply goes to waste.
Making small farmers visible
What stayed with me from the project was how much more visible the farmers became once the information around their work was brought together. Micro-farmers should not be treated simply as entries in a dataset. They are participants in local food systems, and for many households, farming can also provide an important source of income.
Large-scale commercial agriculture will continue to play an essential role in food production. Local resilience also depends on smaller producers being able to participate, particularly where they are producing food close to the communities where it will be consumed.
The lesson from the micro-farmer is fairly practical. Useful information was already being collected. Analytics helped reveal insights that were difficult to see from the raw records alone.
Innovation does not have to mean putting expensive technology on every small plot. Sometimes it starts by making better use of information that already exists and putting what we learn back into the hands of people making the decisions. As climate conditions become less predictable, that is a useful way to think about nourishing tomorrow.