Our approach is one of discovery, aiming to drive effective behavior change through repeated testing and model optimization using cutting-edge research techniques. We rely on A/B testing and extensive piloting to upgrade content continuously and to improve the farmer user experience.
Our research efforts also include scoping studies and analysis of third-party data. We present some of our most recent evidence from the field below.
Evidence from Kenya
We use randomized controlled trials to measure the impact of mobile phone-based advisory services for farmers. Our initial evidence, some done before the formation of PxD, provides a proof of concept that mobile phone-based extension services can improve outcomes for farmers.
Research shows that sugar cane farmers working with Mumias Sugar Company who received agricultural advice via SMS experienced yield gains of 11.5 percent. By reporting delays in input deliveries through the phone system, interruptions in fertilizer distribution were reduced by 21.6 percent. In contrast to this trial, a follow-up trial had no significant impact on yields.
Yield increase for sugar cane
Decrease in input delivery delays
Evidence from India
Farmers listen to more advice when asked for call preference in advance
A/B test (Gujarat, India)
A customized mobile phone-based extension service for coffee growers
Project report (Karnataka, India)
Using mobile phones to complement soil health card distribution
In-depth analysis (Gujarat, India)
Research shows that the reach of traditional in-person agricultural extension services in India is far from universal. We also find that the distribution of physical “soil health cards,” a priority of several governments in the region, could be made more effective by complementing with ICT approaches, including audio and video messages that explain the contents and recommendations to farmers.
Evidence from Ethiopia
Evaluation of a nationwide government voice-based agricultural advisory service
Project report (Ethiopia)
The Agricultural Transformation Agency (ATA), a government agency in Ethiopia, commissioned PxD to conduct a diagnostic assessment of its interactive voice response (IVR) mobile advisory platform. Since 2014, this “Farmers’ Hotline” has reached more than 4 million farmers with agricultural advisory information on 21 high priority crops such as maize, barley, and teff. PxD conducted rigorous analysis of how farmers use the service and provided ATA with a set of actionable recommendations on how to improve the user acquisition, content, user experience, and user retention of the service. PxD is now working with ATA to implement the highest priority recommendations and iteratively improving the service design using A/B tests.
Evidence from Bangladesh
Do branded input recommendations affect farmer trust in advice?
A/B test (Bangladesh)
Realizing the potential of digital development: The case of agricultural advice
A paper by Raissa Fabregas, PxD co-founder Michael Kremer and Frank Shilbach, published in Science, including a meta-analysis of randomized evaluations of mobile phone-based agricultural extension services.
Using Satellites and Phones to Evaluate and Promote Agricultural Technology Adoption: Evidence from Smallholder Farms in India
A draft paper by Shawn Cole, PxD co-founder, Tomoko Harigaya, PxD’s Chief Economist, Grady Killeen, PxD Research Assistant, and Aparna Krishna, Indian Institute of Technology