Name: Emiel Veersma
Title: Data Scientist
Country: Rwanda
Organization: One Acre Fund
One Acre Fund is a nonprofit organization that supplies smallholder farmers in East Africa with asset-based financing and agriculture training services to reduce hunger and poverty. Headquartered in Kakamega, Kenya, the organization works with farmers in rural villages throughout Kenya, Rwanda, Burundi, Tanzania, Uganda, Malawi, Nigeria, Zambia, Ethiopia, and India. One Acre Fund actively serves more than 1 million farmer families.
One Acre Fund offers smallholder farmers an asset-based loan that includes: 1) distribution of seeds and fertilizer; 2) financing for farm inputs; 3) training on agriculture techniques; and 4) market facilitation to maximize profits. Each service bundle is around US$80 in value and includes crop insurance to mitigate the risks of drought and disease. To receive the One Acre Fund loan and training, farmers must join a village group that is supported by a local One Acre Fund field officer. Field officers meet regularly with the farmer groups to coordinate delivery of farm inputs, administer trainings, and to collect repayments.
One Acre Fund offers a flexible repayment system: farmers may pay back their loans in any increment at any time during the growing season. Beyond their core program model, One Acre Fund also offers smallholder farmers opportunities to purchase additional products and services on credit. These include solar lights and reusable sanitary pads.
Awards Categories:
- Data Science for Good
- Most Impactful Ikigai Story
Business Challenge:
One Acre Fund is providing seeds and fertilizers to help almost 2 million farmers in East Africa grow more crops and increase their revenue. Farmers order planting inputs before the rainy season, and may pay back their loans in any increment at any time during the growing season.
Historically One Acre Fund communicated through local field officers with the farmers. Every field officer acted as an intermediary, managing 200 farmers each - hence it was very difficult for them to visit every farmer in person regularly, especially as they are typically scattered across rural areas. We were therefore looking for a more scalable way to address this issue, especially to support loan repayment collection, which had the potential to hinder our growth.
We wanted to meet farmers on the channels they were already active on, to maximize deliverability and impact. The strategy we devised is to leverage SMS to nudge those who hadn’t repaid during the season. A farmer would receive up to three SMSes every two weeks to remind them to stay on schedule.
In February 2022, we started sending SMS and were quickly able to see that it improved the farmers repayments. This convinces us that measures like these can help us to scale to increase the number of farmers we can serve while keeping the cost of field officers more manageable.
Business Solution:
We used Dataiku across the entire project - from setting triggers for repayment reminder SMSes to farmers, to A/B testing on messaging content and timing, and deeper performance analyses.
First, we clustered all of our clients in Kenya in small groups, then separated them into different buckets to make sure distribution is appropriate. To the first bucket, we sent out an SMS with a custom message; to the second bucket, a generic SMS; and the third group didn’t receive any SMS.
With Dataiku we created a list of farmers that were eligible for receiving an SMS that day. This would trigger a batch process at the side of Infobip, our SMS service. We’ve tested this during two months to assess the difference in repayment. Those who received the SMS paid back to a greater extent: 42% for group 1, 40% for group 2. This difference may seem small, but due to the scale of the SMS project, it resulted in $400,000 USD extra credit repaid during the pilot period.
SMSes are quite expensive for our organization, as sending SMS is costly. To reach out to 1 million farmers on a weekly basis, one would pay hundreds of thousands of dollars - which makes it especially important that we were able to easily track and communicate on impact against the third group in Dataiku, with quick visual graphs.
After this proof of concept, we conducted a new pilot to check if it would make sense to send fewer SMS. It turns out that sending the extra SMS increased the impact - and again, Dataiku made it easy to analyze ROI and ensure it was worth the cost.
In the coming weeks we'll set up an uplift analysis to target farmers who will benefit the most from receiving these SMS, enabling us to increase our ROI vs target farmers.
Business Area Enhanced: Marketing/Sales/Customer Relationship Management
Use Case Stage: Built & Functional
Value Generated:
For One Acre Fund, increasing the repayment rate is fundamental to ensure solvency and growth of our operations. With this project, we were able to test ways to do so at a controlled cost - very quickly attaining a maturity level in data analysis which ensures ROI on these new initiatives.
This is pivotal to enable us as a nonprofit to continue and expand our operations to better support the farmers. On their side, repaying enables them to join the program year after year and grow their output to increase revenue.
Specific Value Brought by Dataiku:
Dataiku enabled us to conduct this project in two months, starting with a quick pilot with limited resources, and growing impact from there.
The visual interface enables our team to upskill and accomplish much more than would be possible with code-only tools - from connecting to the data to analysis and communicating insights. During the pilot, we had a bug which triggered uncontrolled SMSes to send out to the farmers. With Dataiku we were able to create a user interface for the business.
Now the field manager, who supports numerous field officers, is able to trigger, monitor and abort jobs. Teammates in the field are actually able to test scenarios to send to their own mobile, then launch at a broader scale - Dataiku lets the business be in charge!
Value Type:
- Increase revenue
- Reduce cost
