West Africana by Hicham Chahidi – http://www.musicscreen.org/
SFX from http://www.freesoundslibrary.com
(Female robotic voice): New SMS. Reading…
Voice actor: [In mystic tone and voice]: The full moon has scared the monkeys…
Narrator: Without the proper context, that message makes no sense. But if you are a small-scale farmer living in Mbeere, Embu county, Kenya, it means only one thing…
On this episode of Built in Africa, we will be looking at how South African agritech startup, ITIKI uses artificial intelligence to simplify rainfall forecasts for small-scale farmers around Africa.
Most African economies rely on the activities of small-scale farmers who, according to the United Nations, produce about 80% of the continent’s food.
With nearly 95% of their planting activities dependent on rainfall, accurately predicting weather conditions is crucial for small-scale farmers who can lose everything to wrong cropping decisions. This is where ITIKI comes in.
Professor Muthoni Masinde: “The idea behind ITIKI is to produce relevant drought forecasts for small-scale farmers”
Narrator: That’s Professor Muthoni Masinde, Kenyan computer scientist and founder and CEO of South Africa based ITIKI.
Prof. Muthoni: “The reason why we use the keyword ‘relevant’ is because there’s so much information out there. If you look at your phone, there’s a forecast. If you turn on the TV, there’s a forecast. If you open a page in a newspaper, there’s a forecast. But for us, we realised that none of that is useful to the small-scale farmers, so we went about creating a solution that works for them.”
Narrator: What is novel about ITIKI’s model is that it combines indigenous knowledge with technology to guarantee accurate predictions, and deliver them in a relevant, yet inexpensive way.
Prof. Muthoni: “We relay that information to them in formats and semantics that relate to their context. So it’s an SMS in their local language and it does not tell them things like ‘near normal rainfall’ or ‘300 millimetres of rainfall’. Rather, it will tell them, ‘this season, you’ll get rainfall that is not enough to grow your usual maize, so you may consider planting millet or sorghum’. When it’s very near the cessation, when the rain is about to stop, a week before, it tells them, ‘Ish… you need to stop planting because, in the next 3 weeks, there’ll be no rain.
So things like those, they relate to them and that’s how we managed to get to their hearts.
Narrator: Remember that seemingly cryptic message from earlier? A typical SMS to ITIKI users reads just like that.
Prof. Muthoni: “The moon has scared the monkeys. So that means that in 3 days time, it will start raining, and this rain will be so short-lived that you have to hurry in your planting.”
Narrator: In case you are wondering, yes Prof. Muthoni grew up in one of such villages, so she has first-hand experience and understanding.
Prof. Muthoni: “It’s simple but it’s so rich. When you mention the moon scaring the monkeys, it immediately rings a bell in their minds because they have used that as a sign before. The only new information they have is exactly the number of days, which they may not be aware of, and the nature of the rain”.
Narrator: The groundwork for ITIKI was laid between 2011 and 2012, while Muthoni was doing her postgraduate research at the University of Cape Town, South Africa.
In 2013, with the help of a team of researchers that she led at the incubation hub of the Central University of Technology, Free State, South Africa, she experimented further with the models and algorithms that help ITIKI predict droughts.
While carrying out live experiments, in Mbeere, Embu county, Kenya, it became clear that ITIKI needed to be launched as a business.
Prof. Muthoni: “So my original idea was really not to do a business. My original idea was to help my people. By design, the tool was tested in my own village, with my mum, my sisters, my relatives, my neighbours. I decided I was going to do it for free, in the sense that I was willing to pay the ambassadors from my own personal salary”
Narrator: But there was a problem. Even though the product was free, many of the farmers expected to be paid to use it. It was almost like they did not appreciate the value of the solution.
Prof. Muthoni: “After 6 months, I knew this cannot work. I decided I would do it some other day, but as a business. From that time henceforth, I knew it’s going to be a business. Even if they pay half of the cost, let them pay it, then they’ll value it some more”
Narrator: In 2015, the product expanded to Mozambique and by 2017, it went fully commercial
Prof. Muthoni: “All this time, we were doing it in small bits, like 50 farmers. In Kenya, we had 200, Mozambique, 50, South Africa, 50. In 2017, we went fully commercial. So the real commercial ITIKI started in 2017”.
Narrator: Don’t be fooled by its simplicity. In the background, ITIKI is doing a lot of heavy lifting, using a neural network of AI algorithms to learn historical rainfall patterns so it can predict future rainfall with large-scale accuracy of about 98%.
But the farmers don’t need to know the technical details. In fact, to give them a sense of belonging and foster adoption, ITIKI involves them in the process.
Prof. Muthoni: “Even as we speak, every Friday we speak to the champions on the ground. They feel it’s theirs and we’re just a facilitator. So they observe their own indicators, they take pictures using their phones, they describe them the way they have done in the past, and when the forecasts come, they say ‘oh this is exactly what we gave you.’But in the background, we do use science tools to do the actual prediction”.
Narrator: And with the help of champions and field agents, ITIKI is able to yield even more precise predictions on a micro-level.
Prof. Muthoni: “Our farmers provide micro-level to an accuracy of even 500 metres. You’ll find it’s raining here, where I am, and a kilometre from there, it’s not raining. So that gap, we fill that data gap using the indigenous knowledge and small handheld wireless sensor-based weather stations. So our model becomes highly accurate, compared to the national meteorological department ones ”
Narrator: From inception till date, ITIKI has received seed funding from the host university, a $500,000 USAID fund in 2017, and other investments from the government of South Africa.
Today, the startup has over 13,000 paying users in the 3 African countries, over 70% of which are based in Kenya. Considering the income level of the target market, this is quite a feat. Prof. Muthoni attributes their success to being people-centric
Prof. Muthoni: “In Kenya, we are almost hands off the operation. So, it runs around the local people, they knew that person before, he was already helping with that aspect of farming. So we anchor ourselves around him or we go through the church, from which they’ve already been receiving seeds. So when you come in, you’re not imposing or disrupting what they’re used to. You are just saying, ‘you know what? In addition to all these things, you’re doing, you can also have this.’
In South Africa, we did the same. The farmers are already in a group, they’re running a climate mitigation project. So instead of disrupting the nature of things, we visited a guy who’s already working with them, he had their names. Essentially, we just use the bottom-up approach and it’s magical because people already buy it before you sell it to them”
Narrator: It also helps that ITIKI offers flexible airtime-based subscription plans, allowing for seasonal or installment payments of as low as 150KSH (less than $2) per planting season.
Of course, until they attain critical mass, this pricing model is not necessarily sustainable. Not to mention how capital intensive their customer acquisition model can be. In addition, farmers within the same locality can decide to share SMSes, which will result in less revenue for ITIKI
To address these challenges, Muthoni says they’re shifting focus from B2C to B2B. Small-scale farmers remain the focus but bulk payments from large groups is the aim.
The idea is to engage off-takers, in whose best interest it is for the small-scale farmers to do well; like insurance companies, banks that give farmers loans and even the government”.
Prof. Muthoni: “So these companies add a service to their customers and this service is ITIKI service. We have 2 customers that we are discussing with and things are looking very good”
Narrator: Prof. Muthoni is certain that the startup’s financial model is sustainable and it will take a more refined form as soon as an effective payment collection system is sorted out.
Prof. Muthoni: “We already know the most lucrative and big customer for us is the government. Because the government has interest in the small-scale farmers being self-sustaining because then you don’t have to rescue them when they go hungry.”
Narrator: Although the big picture is to cover the entirety of Africa, in the next 18 months, ITIKI has immediate plans for three other African countries: one in East Africa, one in Southern Central Africa (either Zambia or Zimbabwe), and one in West Africa.
As the ITIKI team seems to be prepared for the coming months, the hope is that beyond attracting revenue, it will continue to tackle the challenge that inspired the idea at inception. Just recently, the startup began expanding capabilities to include other predictions like locust invasions.
They’re also looking to expand the product to include market prices so that farmers can know where to buy affordable farm inputs and where to sell their farm outputs at good prices.
Finally, Muthoni is confident that when smallholder farmers can maximise production by planting at the right time, they will be able to survive without depending on the government.
In the same vein, she believes the number of women represented in the market is also a plus that will help ITIKI with its focus.
Prof. Muthoni: “I think one of the things put out there is the fact that the small-scale farming sector is for women; they’re the ones tilling those small pieces of land. When we empower women, we are likely to change the face of Africa and that’s, for us, good. And especially their contribution to the continent’s GDP is strong.”
Thank you for listening to Built In Africa.
This script was adapted by Muyiwa Matuluko
Research and interview by Oluwanifemi Kolawole
Sound design by Oghenekaro Obrutu
This is a production of Techpoint Africa
I am Oluwanifemi Kolawole
Please subscribe, share and drop a review of this podcast, by searching for ‘Built in Africa’ on Apple Podcasts, Google Podcasts, Spotify or wherever you get your podcasts. You can also email us feedback at [email protected].
For ad placements: [email protected]
For more stories on startups and innovation in Africa, please visit Techpoint.africa