Experiment (select + to view abstract)	Abstract	Start date	Location	Partner_organization	Research tags	Status	Agricultural_season	Experiment_type	Approved by RM	Public
Monsoon Forecast Dissemination 2024	Climate change is exacerbating weather uncertainty, and there is strong evidence that farmers adjust their behavior and investment decisions in response to accurate weather forecasts. In 2024, PxD, in collaboration with DIL and the Ministry of Agriculture and Farmers’ Welfare (MoA&FW), launched a large-scale pilot to deliver critical weather information to farmers across five Indian states via SMS. The pilot provided two key types of forecasts: Monsoon Onset (MO), which offered forecasted dates for the start of the monsoon season, and Total Rainfall (TR), which projected cumulative rainfall expected over the June to September monsoon period, with data sourced from the Indian Meteorological Department (IMD). The primary objective of the pilot was to scale up dissemination while maximizing learning to inform future expansion efforts. The pilot aimed to monitor dissemination effectiveness and evaluate the impact of providing accurate forecasts on farmer productivity and welfare. Specifically, the learning agenda focused on assessing the effectiveness of dissemination; understanding farmers’ reactions to the forecast, including comprehension, trust, and belief updates; exploring potential spillovers from peer-to-peer learning; gauging demand for forecasts; analyzing how forecasts influence farmer behavior and agricultural activity; and examining the broader effects on farmer welfare, profits, consumption, and household finance.		India	Development Innovation Lab - University of Chicago, India Ministry of Agriculture	Agricultural management advice, Communication technology, Social learning, Weather information	Ongoing	Kharif	Impact Evaluation
iCare Weather Project Karnataka	PxD partnered with the Coffee Board of India to deliver Coffee Krishi Taranga (CKT), a voice-based agricultural advisory service for coffee farmers in Karnataka. In addition, PxD collaborated with the Climate Forecast Applications Network (CFAN) to provide highly localized, sub-block-level 5-day cumulative rainfall forecasts suited to coffee farmers' contexts. Weather uncertainty amplifies agricultural production risk. Weather forecasts are a technology that can help farmers better manage this risk — particularly with improving forecast skill and increasing weather variability due to climate change. Accurate, granular weather forecasts at seasonal, medium-range, and short-range timescales can help farmers form better expectations about upcoming weather. These expectations could translate into optimal action when farmers are equipped to act on them. In this study, we focus on one strategy to help farmers translate beliefs into action — providing advisory on recommended actions in addition to a weather forecast through a voice-call-based service. The standard advisory service provides instructions on best practices for different agricultural activities. In separate calls, the weather forecast service provides weather forecasts alone or forecasts combined with recommended actions based on forecasted weather. The objective of this experiment is to assess whether providing agronomic advice together with weather forecasts improves farmers' engagement with the information and supports informed decision-making.		India, Karnataka, India	CFAN, Coffee Board of India, iCare	Communication technology, Message timing and frequency, Weather information	Completed	Kharif, Rabi	A/B test
Asset Collateralized Loans: Non-linear Repayments	Rainfall and temperature variability induced by climate change pose a substantial economic risk to smallholder dairy farmers in developing countries (Tadesse and Dereje 2018). Rainwater harvesting tanks may help farmers adapt to climate uncertainty. Previous work has found that Asset Collateralized Loans (ACLs) help farmers purchase water tanks in Kenya (Jack et al, 2023). This study aims to validate past work and evaluate the impact of ACLs for water tanks on economic and household outcomes among dairy farmers in Kenya, with a focus on climate resilience. The study will also test an alternative loan design in which required monthly payments are a function of milk income, which may be better adapted to farmers with seasonally variable income.		Kenya	Development Innovation Lab - University of Chicago, LEPESA SACCO, Lessos Dairy Cooperative, Sirikwa Dairy Cooperative	Agricultural management advice, Input recommendations, Service design, Social learning	Ongoing	_Multiple seasons	Impact Evaluation
Andhra Pradesh Coffee A/B Test	PxD has operated the Krishi Tarang service in Gujarat since 2016 to provide free agriculture information via mobile phones using a two-way IVR platform with "outbound" push calls and an "inbound" hotline service. PxD developed advisory content for users of our coffee advisory service, Coffee Krishi Taranga (CKT), a voice-based advisory platform developed in collaboration with the Coffee Board of India. Identifying that increasing weather variability due to climate change has increased the importance of timely agronomic information to aid farmers in decision-making. In this study, we focus on coffee farmers in Andhra Pradesh. In this study, we want to see the impact of the CKT service on priority practice adoption. The farmers are stratified by block into a control arm, where they will receive access to the hotline, and a treatment arm, where they will receive access to the outbound voice-call-based advisory service in addition to the hotline and SMS dissemination. Results are awaited.		India	Coffee Board of India, Walmart	Agricultural management advice, Communication technology	Ongoing	Kharif, Rabi	A/B test
Bayer - GeoPotato Evaluation	mPower and PAD partnered with Bayer to test branded SMS alerts in the GeoPotato advisory service, which delivered real-time fungicide recommendations between 2018 and 2019. The partnership aimed to explore sustainable models for delivering digital advisory services by testing how branded content affects farmers’ behavior and perceptions. The main purpose of this trial was to assess whether Bayer-branded GeoPotato alerts influence farmers’ brand recognition, fungicide usage, and trust in the service, as well as their willingness to adopt recommendations. Farmers were randomly assigned to receive either standard GeoPotato alerts or alerts that included Bayer branding. The study measured outcomes related to brand recall, fungicide use, and trust in the service. The results showed that branded messages led to an 18% increase in the usage of Bayer products without undermining farmers’ trust in the GeoPotato service or their willingness to adopt its recommendations. However, the branded alerts did not significantly impact farmers’ ability to recall the Bayer brand or recognize its products. These findings support the potential of partnerships with private input suppliers to sustain digital advisory services while boosting product uptake.	2018	Bangladesh	mPower	Extension agents, Input markets, Social learning	Completed	_N/A	A/B test
Experiment 112: Livestock Menu Arrangement	PxD is partnering with Ethiopia’s Agricultural Transformation Agency to help improve the effectiveness of their existing voice-based service through continuous iterations and experimentations, as well as making suggestions for improvement and customization potential. Ethiopia’s mobile advisory service, the 8028 hotline, has received more than 52 million calls from over 5.9 million farmers (April 2022 data) and represents the first in Africa to be maintained by a government entity at such a large scale. The livestock content was prepared and launched through 8028 in May 2020. The livestock content and the menu were structured as dairy (first menu), fattening (second menu), small-scale poultry (third menu), improved household poultry (fourth menu) and apiculture. Based on the analysis that PAD has done throughout, callers are likely choosing based on the arrangement of the main menu items. Dairy gets the most hits, followed by fattening, then small-scale poultry, improved household poultry, and apiculture. Similarly, the distribution of hits is more proportionate when the menu choices are fewer. For instance, the fattening menu has two submenus: cattle fattening and sheep and goat fattening. Despite the fact that sheep and goat fattening appears second, it gets more hits compared to cattle fattening.		Ethiopia	Ethiopian ATA	Communication technology	Completed	_N/A	A/B test
MoA-INFO LR2021 Registration A/B Test	PAD operates the MoA-INFO platform in collaboration with Kenya’s Ministry of Agriculture to provide free agricultural recommendations to farmers via SMS. In this A/B test, we test a redesigned registration process into the platform to see whether inviting users to opt into the advisory message series first, before asking registration questions, affect the likelihood of registration completion and platform engagement behavior. Results show that the treatment group that was asked the first question to opt-in to preferred crops had a higher rate of crop series opt-ins but provided less profile information (e.g., name, location) and engaged slightly less with the platform after registration. Overall, switching the registration order improved specific registration metrics but reduced others, indicating trade-offs in registration design choices.	2021	Kenya	Kenya Ministry of Agriculture	Message framing	Completed	Long Rains	A/B test
Ama Krushi livestock speaker phone nudge	Different livestock activities are usually performed by different people within a household. In order to make sure that livestock advisories reach all the members of the households, particularly women in the family, the team is exploring if it is possible to nudge farmers to use their speakerphones to jointly listen to the advisories with other household members, particularly spouses of the Ama Krushi livestock users.		India, Odisha, India	FARD	Communication technology, Gender, Message framing, Social learning	Completed	Rabi	A/B test
Name	PxD has operated the Krishi Tarang service in Gujarat since 2016 to provide free agriculture information via mobile phones using a two-way IVR platform with "outbound" push calls and an "inbound" hotline service. The A/B test aims to assess the effect of including the recipient’s name in advisory push calls on user engagement.		Gujarat, India, India	J-PAL	Communication technology, Message framing, Message narration	Completed	Kharif	A/B test
Frequency A/B test	PAD operates Krishi Tarang (KT), a two-way IVR service through which farmers receive weekly advisory messages and have access to an inbound hotline service to access previous content and ask questions. This A/B test tested whether increasing call frequency with shorter messages or allowing farmers to choose their preferred frequency would improve overall engagement. Farmers were randomly assigned to receive either standard weekly calls, twice-weekly shorter calls, or to select their preferred frequency. The A/B test ran for ten weeks. The results suggest that both higher frequency and preference-based approaches increased listening rates, with stronger effects observed when farmers could choose their call frequency.	2017	Gujarat, India	J-PAL	Message timing and frequency	Completed	Kharif	A/B test
Training method A/B test	Analysing different methods of training - one-on-one training (remote v/s field) v/s group and its effect on farmer engagement	2016	Gujarat, India	J-PAL	Social learning	Completed	Rabi	A/B test
Call length A/B test	PAD operates Krishi Tarang (KT), a two-way IVR service through which farmers receive weekly advisory messages and have access to an inbound hotline service to access previous content and ask questions. KT push calls range from 60-140 seconds and cover one to four topics. Farmers who receive messages do not know the total length of the message, or the lengths of various topics being discussed. Farmers may hang up at different points during the call, which could be an arbitrary decision as they do not know how much content remains and whether it is relevant to them. This experiment is designed to test whether the decision to listen or not is a function of whether or not they are interested in the advice. A total of 6,000 cotton farmers were randomly assigned into three groups to vary whether they were provided total call length and topic length information. Over 8 weeks, each group received two weekly advisory calls, with the order of topics randomized. In the final week, all calls included both a relevant topic (pesticide use) and a less relevant one (lemon cultivation). Results show that call completion rates increased when relevant content was presented second, supporting the hypothesis that listening is driven by perceived relevance. Providing call length information also improved engagement, with both treatment groups showing significantly higher completion rates compared to the control. These findings suggest that call structure and transparency about content length can enhance IVR message engagement.	2018	Gujarat, India	J-PAL	Message framing	Completed	Kharif	A/B test
Cropping Series Invite Messages	PxD operates the MoA-INFO platform in collaboration with Kenya’s Ministry of Agriculture to provide free agricultural recommendations to farmers via SMS. This experiment aimed to improve cropping series engagement by refining MOA-INFO's cropping series invitation messages.	2019	Kenya	Kenya Ministry of Agriculture	Message framing, Message narration	Completed	Short Rains	A/B test
Crop diversification	PAD is working with the West Bengal Accelerated Development of Minor Irrigation Project (WBADMI) to implement a mobile phone-based extension system, Krishi Katha, for smallholder farmers belonging to the Water User Associations (WUAs) formed under the ADMI project across West Bengal. The objective of this A/B test is to nudge WUA farmers in West Bengal through IVR and SMS messages to grow more profitable crops like Brinjal, Tomato, Okra, Gourds, Cucumber and Black Gram on a portion of their land during Kharif 2020.	2020	India, West Bengal, India	ADMI	Agricultural management advice, Message framing	Discontinued	Kharif	A/B test