Why Digital Farming Tools Fail Smallholder Farmers in Africa

Why Digital Farming Tools Fail Smallholder Farmers in Africa

The precision agriculture startup had raised millions. Their app promised to tell farmers exactly when to plant, how much to water, and which fertilisers to use. They piloted it in Kenya. They got glowing testimonials from a handful of farmers. Then they scaled up, and the whole thing fell apart.

Here is what the data actually shows: across sub-Saharan Africa, smallholder farmers are not ignoring digital tools. They are using them. But not in the way Silicon Valley imagined. And the gap between what the technology promises and what it delivers is not a technical problem. It is a structural one.

Dennis Choruma and his colleagues at the International Maize and Wheat Improvement Center and the University of KwaZulu-Natal recently published the most comprehensive review yet of how digital agriculture actually works for smallholder farmers in sub-Saharan Africa (Choruma et al., 2024). They combed through 27 studies spanning multiple countries and years. What they found is not a story of failure. It is a story of mismatch.

What Farmers Actually Use vs What Tech Companies Sell

The authors found that digital agriculture in sub-Saharan Africa clusters into two main categories: digital extension services and digital marketing platforms (Choruma et al., 2024). That sounds broad, but it tells you something important.

Digital extension services are basically farming advice delivered through phones. Think text messages about pest outbreaks, voice recordings about soil health, or WhatsApp groups where farmers share tips. Digital marketing platforms are marketplaces where farmers can find buyers for their crops without going through middlemen.

Both of these are low tech by Silicon Valley standards. They do not require satellites, drones, or machine learning algorithms. They require a basic smartphone and a network signal.

But here is the paradox: the most sophisticated digital tools are the ones that get the most funding. Startups pitch AI powered crop disease detection using computer vision. They pitch drone based soil mapping. They pitch blockchain supply chain tracking. And these tools work beautifully in controlled conditions. But they fail in the field because they assume things that are not true.

The authors found that the primary technologies actually being used by smallholder farmers are not these high end systems. They are simple information services (Choruma et al., 2024). And even those face obstacles that no amount of code can solve.

The Three Barriers That Software Cannot Fix

Choruma and his team identified three major barriers to digital agriculture adoption in sub-Saharan Africa: limited internet connectivity, low digital literacy, and affordability issues (Choruma et al., 2024). Each of these sounds like a problem that can be solved with better design or more funding. But look closer.

Connectivity is not just about coverage

It is true that large parts of rural sub-Saharan Africa lack reliable internet access. But even where coverage exists, the quality is often terrible. A farmer might have a 4G signal in the morning and nothing by noon. The data might cost more per megabyte than the farmer earns in a day. The phone might be shared among multiple family members.

The authors documented that connectivity problems are not evenly distributed. They are worse for women, for farmers in remote areas, and for households with lower incomes (Choruma et al., 2024). This means digital tools do not just fail equally. They fail in ways that widen existing inequalities.

Digital literacy is not about being dumb

Here is where the tech industry gets it wrong. When they say "low digital literacy", they mean farmers do not know how to use the app. But the real problem is more subtle.

A farmer who has never used a smartphone before can learn to use one. The question is whether the app respects how they actually work. Many digital agriculture tools assume farmers can read. They assume farmers have consistent access to electricity to charge their phones. They assume farmers have time to sit through a tutorial.

The authors found that digital literacy is closely tied to education levels, age, and gender (Choruma et al., 2024). Older farmers and women are disproportionately excluded. But this is not because they are incapable. It is because the tools are not designed for their context.

Affordability is not just about price

The cheapest smartphone in Africa costs about $50. That is a significant investment for a farmer living on $2 a day. But even if the phone is free, the data costs money. The charging costs money. The time spent learning the app costs money.

The authors found that affordability is the most cited barrier across all studies (Choruma et al., 2024). But here is the twist: farmers are willing to pay for digital services if they see a clear return. The problem is that many digital tools do not deliver enough value to justify the cost.

The Gender Gap Nobody Is Talking About

This is the finding that should make every agtech investor uncomfortable. The authors found that gender disparities limit the equitable distribution of digitalisation benefits (Choruma et al., 2024). This is not a footnote. It is a systemic failure.

Women make up roughly half of the agricultural labour force in sub-Saharan Africa. But they are significantly less likely to own a smartphone, less likely to have access to credit, less likely to have formal land rights, and less likely to be targeted by digital agriculture programs.

The studies reviewed by Choruma and his team showed that even when women have access to digital tools, they use them differently. They are more likely to use phones for communication than for financial transactions. They are less likely to adopt digital extension services. They face social barriers that men do not, like restrictions on mobility or expectations about household labour.

The result is a digital divide that reinforces an analog one. The tools that are supposed to empower smallholder farmers are actually empowering the ones who were already better off.

What Actually Works: The Surprising Success Stories

Not all digital agriculture fails. The authors found several cases where digital tools genuinely improved outcomes for smallholder farmers (Choruma et al., 2024). The pattern is revealing.

Simple information services work

The most successful interventions were not flashy. They were text message services that told farmers when to plant, what to plant, and where to sell. One study found that farmers who received regular SMS updates about market prices earned significantly more than those who did not.

The key insight: these services did not require farmers to change their behaviour. They just gave them better information to make decisions they were already making.

Voice based services reach the excluded

For farmers who cannot read, voice based services are a game changer. Several studies found that voice messages about pest management or weather forecasts were adopted at much higher rates than text based alternatives.

The authors noted that voice services work because they do not require literacy, they can be delivered in local languages, and they fit into existing oral knowledge sharing practices (Choruma et al., 2024).

Group based adoption reduces risk

When digital tools are introduced through farmer cooperatives or women's groups, adoption rates are higher. The authors found that group based approaches help with affordability (shared devices), digital literacy (peer learning), and trust (someone to ask when things go wrong).

The pattern is clear: digital tools work best when they are embedded in existing social structures, not when they try to replace them.

What the Research Does Not Prove

The authors are careful to note the limitations of their review. They only analysed 27 studies, which is a small sample given the diversity of sub-Saharan Africa. Most of the studies focused on East and Southern Africa, with very few from West or Central Africa (Choruma et al., 2024).

This means the findings might not apply everywhere. A digital tool that works in Kenya might fail in Nigeria for reasons that have nothing to do with the technology.

The review also cannot tell us whether digital agriculture is actually increasing overall agricultural productivity in sub-Saharan Africa. The studies are too small, too short, and too focused on specific interventions to answer that question.

And here is the biggest open question: do digital tools actually improve long term outcomes for the poorest farmers? The studies suggest they help farmers who already have some resources. But for farmers who are truly marginalised, the benefits are unclear.

Why the Tech Industry Keeps Getting It Wrong

Every few months, a new agtech startup announces a pilot program in Africa. They promise to transform smallholder agriculture. They raise millions of dollars. They launch. They fail. And then they do it again.

Why does this keep happening?

The authors found that most digital agriculture initiatives are designed in labs, not in fields (Choruma et al., 2024). They are built by engineers who have never spent a day farming in Africa. They assume that farmers want the same things that Silicon Valley wants: efficiency, optimisation, scale.

But smallholder farmers are not trying to optimise a single metric. They are trying to manage risk. They are trying to feed their families. They are trying to survive a bad season. A tool that increases yields by 20 percent in a good year is useless if it costs money in a bad year.

The authors also found that many digital tools fail because they do not account for the complexity of smallholder farming systems (Choruma et al., 2024). A farmer in Malawi might grow six different crops on the same plot. They might have livestock. They might work off farm. They might rely on family labour that is not available at certain times of year.

A digital tool that assumes monoculture or standardised inputs will fail in this context.

What This Actually Means

The research by Choruma and his team points to concrete actions that would actually help smallholder farmers benefit from digital agriculture. Here is what the evidence supports:

• ▸Stop building for the average farmer. There is no average farmer. Digital tools need to be adaptable to different crops, different contexts, and different levels of literacy. One size fits all is a design failure.

• ▸Invest in infrastructure before apps. Connectivity, electricity, and affordable devices are prerequisites. Without them, no amount of app design will matter. Fund the pipes, not the software.

• ▸Design for women first. If a digital tool works for a woman farmer with limited literacy, limited mobility, and limited time, it will work for everyone. If it only works for men with smartphones and education, it will widen inequality.

• ▸Embed tools in existing institutions. Farmer cooperatives, women's groups, and extension services are already there. Digital tools should support them, not bypass them.

• ▸Measure what matters. Yield increases are nice. But the real question is whether digital tools improve household income, food security, and resilience. Those are harder to measure, but they are what farmers actually care about.

The promise of digital agriculture is real. The authors found that farmers who use digital tools do get better access to information, better market prices, and better decision making capabilities (Choruma et al., 2024). But the promise is not automatic. It requires design that respects context, investment that prioritises infrastructure, and a willingness to admit that the hardest problems are not technical.