Geo Gecko is a Ugandan based company delivering geo-intelligence around the world using big data, satellite imagery and drones. Space in Africa had a chat with Bernard Wright, the Founder and Managing Director of the company, to discuss its operations and the company’s contributions to sectoral development in Africa.
Please walk me through what the company does.
We are Geo Gecko. We are helping agribusinesses optimise their operations to buy from and better support smallholders. They use our crop forecasts to plan their imports, exports and operations. We do this using satellite imagery, machine learning and data from Fieldy, the free smallholder monitoring service we are rolling out across the continent now for large organisations supporting >100 farms. Users get crop and weather data per month, per field without cost (ever), limit (we’re monitoring 100K fields already), or catch (no sneaky upgrades, restricted features etc.).
We have many large agribusinesses, breweries, UN, NGOs, ag-tech, fintech organisations using it to tailor their support to farmers and integrate it into their models for credit profiling. We should have 1,500,000 fields registered within the next few months, all being monitored for free across the continent. We are built to scale, and we have an office in Kampala and a sister company in Dublin, Ireland.
Any particular institution that you are working with, and what challenges are coming up?
Our biggest challenges are clouds obstructing clear satellite shots and ensuring the data quality coming from the ground. The primary challenge we are addressing is to figure out how satellites can effectively monitor and quantify smallholder agricultural activity. It is a tech that is well proven in Europe and the US. We are refining that model for smallholder farming which is typified by small, dispersed and less connected farms.
With crops requiring different metrics to define their health accurately, how does this translate to the monitoring tool?
This is all based on our users’ ground and feedback, especially the agronomists we are working really closely with. It will be a constant effort to figure out what an NDVI value of 0.6 means for maize in the various landscapes and farming practices our partners are working in.
In a 2018 interview with Disrupt Africa, you mentioned that you are working with customers who work with networks of smallholder farmers. However, when it came to smallholders, the data is not perfect. Why is this, and has it improved?
Yep, satellite data is not perfect at identifying what is happening within a smallholder field, especially if you are looking at 100,000 simultaneously. That is the same also in Europe. It is good, though, when comparing stats like crop health, rainfall and soil moisture, especially when you look at it over time. We are putting massive efforts into improving the accuracy and understanding which cases we can achieve higher accuracy and others where it will be less effective, e.g. coffee grown under shade trees and intercropped. Again, working closely with our users will help us improve our analysis.
How is the general uptake of the service?
With Fieldy, our free smallholder monitoring, there is massive interest, to be honest. Plenty of organisations have been flirting with this tech for a decade. However, the cost and technical barrier to adopting it have been high, especially when the use cases are few. So, when we come to them with an entirely free service to provide them with crop/weather stats for an unlimited number of farmers every month, they get very excited… then they ask what the catch is. This is when we explain it will always be free, and we will generate revenue from crop forecasts trained on their data.
I think it is great that organisations can tailor their support to each smallholder. However, I get excited when our partners integrate our data into their value chain digitisation products or fintech credit profiles. They can access all the data we generate for their fields via an API.
Geographically, how far is the reach in regards to the smallholder monitoring tool?
We are already offering it across Africa. We have seen many demands in Ghana, Nigeria, Zambia, Malawi, Mozambique, Uganda, Kenya, and Tanzania. We are open to other regions, too if the learning opportunity is high. We are doing some work in Thailand now and have had a few requests for South America.
The effort of deploying to another geography is zero as our model involves us working closely with partners on the ground.
Going back to your operations, Geo Gecko serves three continents across the globe. Which ones are they, and what informed your decision to settle in Uganda as the base for your operations?
We have been working with organisations primarily in Africa and have done some work in Asia and South America. Africa will remain our main focus for the foreseeable future.
I started working in Uganda in 2005, researching the conflict in the North and stayed running geospatial projects with NGOs and the UN. After a few more years of responding to emergencies with the UN worldwide, I saw the potential in the disruption of the satellite sector (higher frequency, lower cost, higher resolution), cloud computing and machine learning. I knew Uganda, I liked Uganda, I saw the demand and little competition, so I co-founded Geo Gecko then.
We spent the past eight years building an outstanding tech team that are all Ugandan and 66% female. They have been the driving force behind our cloud infrastructure and our algorithms upon which we can scale our service.
Who are your primary satellite data providers?
We primarily use European Space Agency imagery because of the geographic-temporal coverage. Much research has been built on it. We are a Planet partner, too, and we will use that as a premium service for folks who want higher frequency and resolution.
I understand you are internally funded, are there any plans to raise external capital?
Yes, we are internally funded. We have gotten some grants from the European Space Agency and the Irish Department of Agriculture. We are beginning to raise a seed round now. We see much interest as we have a micro impact by making it easier to work with smallholders, have a macro impact by stabilising commodity prices, and our business model is sustainable… we are ticking many SDGs.
Did you have any challenges?
Yep, its cloudy in Africa and the optical satellites cannot penetrate that. We are tackling using the higher frequency Planet imagery and experimenting with Synthetic Aperture Radar (SAR) which does penetrate clouds. However, it is heavier to use, and there is less research about its applicability to crop health.
How did the pandemic affect your business?
It dried up some funding streams, but it did give us the chance to focus intently on Fieldy and our crop forecasts, so it worked out well. It also gave us the chance to set up a sister company in Ireland to build better relations with the European Space Agency and research. We have integrated an ERP into our work a few years back to allow more flexible working for our teams.
What partnerships have you formed with academia and universities locally?
We like Makerere University in Kampala and have recruited a lot of our team from there. We have also been working with IITA (International Institute for Tropical Agriculture). They are great because they bring academic rigour to testing our models and results.
What can we look forward to in the coming year?
We are constantly adding new features to Fieldy, registering more fields. Our data is being integrated into more services, like credit profiling and insurance.
I am excited about the crop forecasts that we are working on. We are making them faster, cheaper and more accurate to fill that gap in market supply data for agribusinesses, NGOs, civil society and government.