Humanizing a Product

A Revelation

Koteshwaramma and her husband Venkateshwar Rao are farmers from Kopparu village in the Guntur district of Andhra Pradesh. Like many farmers around their village and district, they have been cultivating chilli on their one-acre rented agricultural land. 

To give an overview, chilli cultivation has become very input intensive over the years, which means that the risks involved in growing chilli crops have also increased. Over the years, climate change, unpredictable and extreme weather events, have made chilli crops increasingly susceptible to new viruses and pest infestations. In fact, there have been many cases of virus wiping out entire chilli fields of farmers, forcing them to go for re-plantation. 

Without adequate information about changing climatic conditions, unpredictable markets, and higher costs of production, Koteshwaramma and Venkateshwar were faced with a dilemma on whether to plough on. In 2019-2020, with the support from the Andhra Pradesh Government DoAC and RySS Community Extension functionaries, they adopted climate resilient natural farming methods to cultivate chilli crops. Some of these practices include using locally available resources to apply them to chilli crops, spraying botanical decoctions, extracts, and a wide variety of intercrops in chilli to help contain virus and pest infestations. 

Within a year of their transition to natural farming, Koteshwaramma and Venkateshwar Rao have seen significant success with 25 quintal of red chilli yields, and minimal farming expenses. Their success has become an inspiration for many farmers in and around their communities, and also across the state. There have been media articles written about this farming couple, and Koteshwaramma was also awarded as the ‘Best Farmer’ in the chilli crop and natural farming category. 

While Koteshwaramma and Venkateshwar Rao have continued to practice natural farming methods of chilli cultivation, the Digital Green team has followed them over a period of the cropping cycle to document their best practices in video form. This has now become a 12-part video series of Package of Practices. When we ask Koteshwaramma about her success in chilli cultivation, she promptly says that it has been due to the timely advisory and messages that she received on the climate resilient, natural farming methods of chilli cultivation. She also highlights that the support that she received from the RySS extension functionaries is complemented by the video dissemination that she attended to enhance her knowledge on her practice on natural farming.

There has also been a great response from farmers across the state about the DoAC-RySS and Digital Green Green Chilli Package of Practice video series. Featuring a progressive woman farmer, Koteshwaramma has moved many to adopt natural farming practices and learn more about it.

After the pandemic hit, as in-person video screenings became less frequent, the videos that would otherwise be shared within group settings were shortened and shared with farmers via Digital Green’s Whatsapp Chatbot service. 

Koteshwaramma and Venkateshwar Rao are just one example of many smallholder farmers across the state and the country that are dependent on their farming practices as a source of their livelihoods. Timely advisories and the proper dissemination of information and knowledge is imperative in empowering farmers to be more resilient in their farming practices. You may ask, how does this targeted advisory happen given the multitude of challenges that farmers may face across different geographical locations and weather conditions. More importantly, what is the larger story behind this direct impact on farmers day-to-day? The response has to be innovative, and farmer-focused technology that values a human-centered approach.

A Human-Centered Approach

As an organization working in the space of leveraging technology to empower smallholder farmers, who are amongst the poorest and most vulnerable groups, we have been on the quest to focus on solutions that directly impact farmers. With the continuing success of the community-based video approach, we have learned that a human-mediated approach to these videos, featuring farmers themselves as ‘actors’ in the video, and having a screening and discussion in a group setting, leads to higher chances of adopting the practices demonstrated on video. While the video screenings are participatory, we also have to think about what comes after this knowledge is shared? Establishing a direct line of communication and feedback tailored to meet the needs of each individual farmer is imperative. Chatbot fills that gap with one-on-one communication with a farmer who gets to choose which advisory they would like to receive based on the stage of their cropping cycle.

Chatbot presents a unique opportunity to employ human-centered design that follows a hybrid model of communication to complement community videos and anything beyond, with regards to knowledge sharing. This has been a groundbreaking intervention to share timely and targeted advisories with farmers in regional languages at the click of a button and free of cost. The cycle of farmers’ interaction with the bot after video-based dissemination goes like this – farmer gathers knowledge in group setting, and engages in a participatory discussion. After leaving this meeting, a similar communication is sent to farmers via Chatbot which could be a shortened version of the video shared earlier; this helps tremendously with knowledge recall. At a frequency of two to three days, messages related to the cropping stage that the farmer is at, and key actionables to be performed on the crop are shared on WhatsApp.

This has been a continuous learning process for years before the launch of Chatbot. Erica Arya, Head of Product at Digital Green states – “a few years ago, a similar line of communication was set up through the IVR medium, with supplementary messages to reiterate key actionables that farmers needed to perform. We saw that there were a large number of listeners and thus, it informed subsequent steps towards launching a Chatbot on WhatsApp.”

Solutions at a ‘Click’

As the use of smartphones increases in rural areas, the doors to developing digital solutions as effective tools for information exchange have also opened herewith. Our strive to find dynamic solutions on smartphones that can reach farmers organically dates back to 2018-19. Our primary vision has always been to create an application that serves the purpose of adding value. So, instead of investing on a new platform, we thought of leveraging a platform that our users are already familiar with, like WhatsApp. This means that half the battle was already won.

In an immersion visit in Karnataka in 2018, the Digital Green Product team spent days observing farmers’ daily practices, and routines. The team also ran social experiments on Whatsapp groups to study the patterns on how effectively a farmer responds to texts on Whatsapp, how often they use their mobile phones, at what time, what do they click or what they do not respond to.

With a great response from extension agents at first, it gave us the confidence to move to more direct-to-farmer solutions which led to the launch of our first Chatbot in Jharkhand to share crop advisories with farmers. Our initial goal was 500 users, and we started with 467 users who were onboarded. Continued engagement and responses allowed us to take the leap and scale it up to Andhra Pradesh and Telangana as well.

During the pandemic, with the restrictions posed on in-person video disseminations, digital solutions such as Chatbot and IVR took precedence in ensuring the continued dissemination of information directly to the farmers, and were further explored with regards to expansion and higher reach. This was essentially the right space and time to leverage Whatsapp as a medium to connect directly with farmers.

The power of an AI-enabled Chatbot on a platform like Whatsapp simulates a human-mediated, conversational mode of communication with our farmers which can yield great results once tapped. The familiarity and ease of using Whatsapp has also shown through with 94% of users responding to the bot once they have joined the service, in Jharkhand. The two-way communication also makes personal feedback an important source of information, and in Jharkhand we have seen over 52% of users responding to these feedback questions when asked. We have also seen that although phones were generally in possession of men, 89% of the male users receiving advisory also shared the information with the women of the households.

For any tech innovation to be farmer-focused, the agency of the farmer is of utmost importance. Farmers can interact with this chatbot only after they have given consent which can be gathered through multiple channels such as Whatsapp, IVR, SMS or even on paper. Trust building also plays a crucial role, and hence partners’ extension systems are significant in onboarding users to the service.

Once the consent is given, the users can start interacting with the bot. Users, in this case, farmers can communicate with the bot using a chat interface or their voice just like they would converse with another person; the chatbot then decodes the words, or voice notes sent to them to provide a pre-set answer. Our learnings show that an intervention such as this meets the needs of the farmers, and the community finds it easy and convenient.

A Way Forward

The caveat is that while onboarding is a direct process, getting continued responses from farmers is still a challenge. At the response stage, we lose about 30% of users, and this also accounts for the aspect of farmers losing interest on Chatbot because it has been designed in a way that it only responds to certain queries based on its presets. With the recent launch of Voicebot, as farmers are able to use their voice instead of text, it garners a quicker response and there has been higher levels of engagement and retention. 

In the agricultural value chain, farmers need correct information and recommendations from agri-experts on best practices, and the know-how on remedies for crop protection, weather information, suitable time for harvesting, etc. Advisories directly to farmers, through Chatbot, have been truly farmer-centered and they have gotten tailored advisories on a wide-range of topics such as natural farming, pest control, that are relatable because of the issues that they face frequently. With the timely delivery of such information, farmers have been able to build on their knowledge of crop cultivation as well as increase their production, and subsequently have become more resilient.

Delivering customized and climate smart dairy cattle advisory services in Ethiopia

This blog outlines the Digital Agriculture Advisory Services (DAAS) project’s approach to climate smart livestock advisories. The project is implemented by Digital Green, Precision Development, and the International Food Policy Research Institute (IFPRI), in partnership with the Government of Ethiopia’s Ministry or Agriculture and Agricultural Transformation Agency, and is funded by the Bill and Melinda Gates Foundation and the Foreign, Commonwealth and Development Office.

Tsegaye Ayana, pictured on the left with her dairy cows, lives in the Amhara region of Ethiopia with her husband and seven children. In addition to her cows, Tsegaye also raises oxen and sheep and farms maize. She credits extension services with teaching her how to better care for her animals, and notes that “I learned how to feed dairy cows to increase their milk production. My family works hard and we are able to earn more income from dairy activities. There are customers who buy milk from us and with this, I earn a 6,500 ETB per month income.”

Tsegaye is one of many small-scale producers turning toward livestock as a growing source of income. In Ethiopia the population of cattle, the most popular type of livestock, increased by ~50% from 2004 – 2015 and the total number of livestock owning households increased by over 5 million during that same period. As an important source of income, nutrition, and household resilience, livestock ownership is expected to continue growing as small-scale producers aim to diversify their incomes in response to a changing climate which increasingly affects their ability to rely on annual crop productionThese trends represent significant economic opportunities for farmers and show that livestock production should be a priority investment area for governments, non-governmental organizations, and agribusiness actors; however, there is another important aspect to consider – emissions. Globally, livestock production contributes an estimated 14.5% of total GHG emissions, primarily through methane emissions accounting for an equivalent of 7.1 gigatonnes of CO2-eq per annum. In Sub-Saharan Africa, on a per unit basis, milk production is a high contributor to GHG emissions primarily due to enteric fermentation and lack of manure management.

Simple solutions come with complex implementation challenges

There are many practices which can reduce emissions and improve productivity in small-scale livestock production systems. Improving general calf and cow health, enhancing genetic stock through artificial insemination, and providing more digestible fodder all contribute to reduced emissions. Some climate-smart practices can also greatly improve livestock productivity, which is an immediate co-benefit seen by the farmers themselves. Additionally,  in accordance with the DAAS project’s climate and nature strategy, we also consider the natural resource conservation or nature improvement value of the practices promoted. Within the DAAS project we have focused our dairy value chain efforts on some of the most impactful dairy husbandry practices, found within the Ethiopian Ministry of Agriculture’s package of practices, for reducing GHG emissions which also offer clear and immediate co-benefits to the farmer. Additionally, some of the practices promoted lead to clear natural resource conservation outcomes. The table below outlines the key dairy husbandry practices promoted within the DAAS project, the mitigation pathway and/or nature positive value, and the co-benefits seen by the farmers.

Although we know the practices in the table above are effective, there are complex implementation challenges which must be addressed in order to effectively promote them. The predominant environment where most small-scale livestock production happens is highly variable. In Ethiopia, despite the large  number of government Development Agents delivering crop and livestock training services, the reality is that localized conditions, variability in access to goods and services, and language and structural barriers to information dissemination and uptake make the delivery of these solutions difficult. There is a need to develop information pathways which complement the ability of the extension system to provide more tailored advisory services at a larger scale while capturing data on uptake and results of those interventions.*Hypothesis within the DAAS project

Utilizing technology to customize solutions and overcome barriers

We believe that digital solutions can help overcome implementation and dissemination challenges. Within the DAAS project we are working closely with the Ethiopian Ministry of Agriculture and the Agricultural Transformation Agency to build digital solutions which enhance the ability of the extension system to deliver customized advisory services to small-scale producers. Depending on the value chain or solution being developed, the level of customization can vary from simply ensuring that a recorded phone message is delivered in the right language for a geographic area, to precisely calculating the fertilizer needs for wheat production at the village level. Any level of customization depends on the successful utilization of data and dissemination of information; we are addressing these requirements by developing FarmStack, an open source protocol which facilitates the secure exchange of data to inform advisories, and strengthening digital dissemination channels, such as the ATA’s 8028 IVR push line, to complement the existing government extension package. We are also building a farmer registry and work management application which will be deployed at the Development Agent level via tablets. In short, we are bringing together disparate data sets and filling the existing gaps in those data sets to develop more informed advisory content and then delivering that content via digital channels.

Delivering customized, climate smart advisories to millions of farmers

Within the livestock sector, we’re piloting a number of customized advisories based on the practices outlined in the table above. One key channel is a government managed phone information service (called 8028) which farmers can call from any phone in Ethiopia. Through the 8028 service we’ve been able to reach a significant number of small-scale livestock producers with a variety of content and we are continuing to pilot new use cases through these channels. As shown in the chart below, there is relatively high demand for dairy and beef cattle content from farmers using the 8028 hotline. We are also using this service to “push” advisories directly to farmers via recorded voice messages, which allows for a much higher degree of customization by targeting specific farmers based on predefined criteria such as location, type of livestock, or availability of services in their area.

Number of users accessing livestock content via the Government of Ethiopia’s 8028 service from July – September 2021

The push call advisory interventions are tailored to the individual farmer’s language and location and focus on the following:

  1. Artificial insemination: This is targeted at AI technicians and dairy farmers in their area in order to: (i) Improve AI technicians efficiency through reminders of the key steps of AI service provision; (ii) increase farmers demand for AI services; and (iii) inform farmers how to identify when the proper time is for AI services (i.e. heat identification);
  2. Calf and cow management: This promotes key practices for improving calf and cow mortality rates during pregnancy and after giving birth; and
  3. Location based content: This use case will utilize location data to tailor general messages about feeding practices which are timed to the seasonal availability of feed stocks and local conditions and can greatly reduce enteric fermentation. For example if a farmer in Oromia has heifers and sells milk then the feed advisory to optimize the health and reduce the emissions from the cows in March may be different than another farmer who is raising steers for plowing in SNNPR.

We are actively developing a methodology for measuring the adoption of practices promoted via the 8028 channel which will support our ability to quantify the economic and climate related impacts of these advisories. Early phone survey data indicates that farmers who use the inbound 8028 line often understand and retain the information they learn, but further studies are needed to definitively determine the adoption outcomes from this service. When the final iteration of the adoption methodology is complete and tested, we will release that and update these materials accordingly.

A pathway to scaling climate smart dairy cattle practices in Ethiopia

The DAAS project’s approach to reaching scale, with a target of 3.5 million farmers across all value chains by 2024, is threefold:

  1. Proof of concept: Develop customized use cases to demonstrate the value of data integration and customization
  2. System strengthening: Improve digital channels and tools, including video enabled extension, IVR technology, and digitizing extension processes
  3. Catalyzing the ecosystem: Build an ecosystem of actors utilizing FarmStack to share and combine data to further developed future advisories

What we aim to achieve here is to demonstrate how data sharing and digital tools can complement and strengthen existing extension systems in a scalable and repeatable way which can grow with the dairy sector. By prioritizing climate-smart advisory content and targeting the dairy value chain we are supporting farmers to increase their incomes while mitigating emissions from increased dairy production in Ethiopia.

Making the Government Adoption of Social Innovations Work – Articles

Diagnosing Web Data of ICTs to Provide Focused Assistance in Agricultural Adoptions – Articles