Digitization or AI?: Bridging the Agricultural Extension Gap
In some Nigerian states, 1 extension agent serves more than 10,000 farm families.
The national average, according to National Agricultural Extension and Research Liaison Services (NAERLS), Ahmadu Bello University , is 1 agent to 6,400 farm families, meanwhile the FAO says the ideal is 1:400. The gap is 16X wide!
Here’s what the math looks like:
If an agent works 6 days a week, 8 hours a day, and spends just 4 hours with each farm family per visit, it will take that agent over 10 years to visit all 6,400 families once. That means a farmer who last received a visit in 2016 may still be waiting, using 2016 knowledge on 2026 climate.
And the cost of that gap is measurable.
A well-managed rice farm should yield 5–7 tonnes per hectare but the average Nigerian smallholder gets 2 tonnes. Not because the land is bad or the farmer isn’t working hard but because no one has shown them how.
Research shows that access to extension services alone can increase crop yields by 10–15%. For 38 million Nigerian farmers, that advice is slow in arriving, if it arrives at all.
Since being established in 2021, Extension Africa has been quietly working on this problem.
Through the Africa Agribusiness Extension Academy, we have trained over 4,000 extension agents using 2D and 3D animated content on agronomic skills covering tomato, cowpea, maize, soybean, rice, and cassava.
We know well and have seen what a well-trained digitally empowered agent can do.
Now we want to take that to a scale that was previously impossible by making the agent 10X smarter not by working harder but by working smarter with AI.
An AI-powered extension agent can answer questions in multiple local languages and deliver location and crop-specific recommendations to a farmer anywhere in Nigeria. AI tools can reach a woman farmer in Oyo via a simple IVR interaction without a smartphone or internet access.
Beyond advisory, extension agents also connect their farmers to certified input right there, in the rural community without a middleman. They collect the soil, yield, and behavioral data that agribusinesses have never had visibility into and facilitate market linkages between smallholders and global businesses who didn’t know each other existed.
Every agent becomes a distribution channel for knowledge, input, financial services, and technology adoption. Every interaction feeds the AI system making it sharper with time.
When you put AI in the hands of Nigeria’s existing 15,000 agents at a 10x productivity multiplier, those agents can cover the work that would otherwise require 95,000.
The same expertise but infinitely scalable.
We are not replacing the agent, we are giving them a superpower.