Precision Ag for smallholder: reaching almost 500,000 via phones and 9 million via TV in Kenya


The majority of farmers in Sub Saharan Africa practice rain-fed argiclture. They do not use irrigation so are at the mercy of the rains. PlantVillage farmers in Busia County, Kenya experienced significant hardship at the start of the long rains in 2019 (March). The twin cyclones of Idai and Kenneth which hit Mozambique (March 14th and April 25th) removed moisture from the atmosphere above Kenya and Uganda reducing the volume of rain that fell. Increased numbers of cyclones is a consequence due to higher sea surface temperatures. These cyclones bring immediate devestation to farms in  their path. Or when they affect rainfall patterns as happened in 2019. Cyclones are also playing a major role in  the current  locust upsurge.

In 2019 many of the farmers planting maize simply had no germination of  the seed and had to re-sow seeds. For extremely poor farmers this was a considerable hardship. In some cases the second seeds sown had failed. 

Some farmers who had an earlier start so had some growth of maize saw their crop stop growing.  You can Roseline Akoichi, one of our lead farmers discuss this in the video.  



What was very clear was that farmers needed hyperlocal advice to enable climate change adaptation. PlantVillage therefore partnered with Mediae Company who produce a very popular TV show for farmers, Shamba Shape Up,  to deliver climate change advice. And with their offshoot company, iShamba to send messages out via SMS. This work was enabled by PlantVillage's excellent and ongoing collaboration with Mercy Corps AgriFin. 


Climate Smart advice reaching 9 million people weekly

Shamba Shape Up is East Africa’s leading agricultural TV program. The program is a reality make-over show filmed on smallholder farms addressing the problems of the farmers to increase production and encouraging farming as a business in Kenya and Tanzania. It’s broadcast on national TV stations in Kenya and Tanzania and on DSTV and a number of free to air digital channels reaching over 12 million viewers each week in both countries in English and Swahili over half the year. 

PlantVillage used data from CHIRPS and the Climate Hazard Center to create maps together with the Kenya Meteorological Service that could be shown each week. Here is one of the videos that are going out 

What Shamba Shape up does so well is recording the changes a farmer can do on their farm. They invite experts like extension officers, Universities,  PlantVillage, CGIAR and FAO to visit the farm and show farmers the best approaches. People watching this at home learn very well when they see other farmers following the advice. Here is a show the features Climate Smart Agriculture advice from FAO 

What is also nice is that Shamha Shape Up have a hotline farmers can join for weekly alerts or to request more information. This is iShamba which we discuss next 

Shamba Shape Up: hyperlocal advice for >500,000

The iShamba hotline has roughly 500,000 users across Kenya. It is not possible to get a GPS link from a feature phone but when farmers sign up iShamba asks them their nearest primary school and what sub-ward. This enables us to get within a few kilometers. 

Using the CHIRPS data set and detailed discussions with Dr. Chris Funk of the Climate Hazards Group we were able to use the historical and forecasted precipitation data at the spatial resolution of 5km to deliver three key numbers to each farmer.  

  • How much rain was fallen during historical Short Rain seasons in similar years
  • How much rain has fallen so far in this Short Rain season 
  • How much is expected for the next 10 days. 

In the Long Rains of 2022 we sent 8.8million SMS messages

We are working with iShamba to also collect the timing of planting and the type of crop planted. With this, we can be very specific and provide advice on how much water the crop needs at any point in the cycle. We will combine this with a risk prediction on pests for that crop and information on the soil conditions. We are developing advanced approaches in machine learning which we will detail on a separate page soon. 

We also are following up with telephone calls to ask farmers what they think. The early feedback is excellent. Here is an example reply 

                              I was able to plan my short Rain season, when to plant and spray


But for now, it is clear hyper-local data is possible for the smallholder farmer at scale. And it is cheap!  Sending 350,000 messages a week costs just $1,800.




Heart Heart icon