Learning from farmers, agro-dealers, NGOs and agricultural scientists
We interviewed 67 rice farmers across four rice-growing regions of Myanmar. Most of these farmers expressed concern about the declining fertility of their soils. They knew they had to apply more fertilizer but they didn’t know if the yield increase would justify the high interest loan required to buy the fertilizer. For this reason we resolved to provide an expected return on investment (with an associated risk factor) for the app’s site-specific fertilizer recommendations. We plan to calibrate predictive models using farmer field trials and refine these calibrations through machine learning. This will enable farmers to make better-informed decisions as well as secure loans for fertilizer purchases.
We bounced our prototype off input suppliers, rice processors and contract farmers. They were very interested in the app but two of them expressed concern about the variable quality of fertilizers. One input supplier explained he had to stop buying one brand of urea (a type of nitrogen fertilizer) because it turned to liquid within days of arriving at the store! This has important implications for our smartphone app. Our fertilizer recommendations and estimated returns on investment will be misleading if the farmer buys poor quality fertilizer. To address this we will only recommend proven fertilizer brands. Advertising specific brands of fertilizer may also create opportunities to generate revenue.
One Acre Fund (NGO)
We also connected with One Acre Fund’s Myanmar team - a massive trip highlight! The team have delivered improved inputs and agronomic advice to smallholder farmers in Myanmar for the last two-and-a-half years. They have encountered a lot of soil variability so they really liked how our app’s fertilizer recommendations will be site-specific. They confirmed that most smallholder farmers have access to smartphones but also warned that some older farmers are illiterate. The team at One Acre Fund advised us on how to address this challenge. They learnt to give farmers picture-based fertilizer recommendations to support farmers that cannot read very well. We are keen to integrate this clever approach into our smartphone app. The One Acre Fund team also gave us a game of basketball as well as some advice on how to avoid food poisoning in rural Myanmar (we wished we talked to them earlier!)
We have consulted multiple agronomists, plant scientists and soil scientists – most notably Dr Achim Dobermann (co-creator of the technology we are digitizing). Achim generously gave us a lot of his time, knowledge and encouragement. He confirmed that calibrating different fertilizer recommendations for every individual rice variety grown in Myanmar would be unnecessary. This was a huge relief as there are a ridiculous number of rice varieties grown in Myanmar! Achim explained that we only need to differentiate fertilizer recommendations between traditional rice varieties and improved rice varieties. We are planning to return to Myanmar in early July to initiate farmer field trials, beta tests and learn more lessons. We’ll keep you posted!