Kerry Harwin and Rikin Gandhi
Digital Green
A Rural Video-Based Social Network
for Farmer Training
Innovations Case Narrative:
Digital Green
Reality television programs and social media now offer a world stage for anyone
who aspires to become a star. The window these platforms provide into the lives
of others is inspiring people to pursue their dreams as they see peers dance, sing,
invent, and cook their way to fame. People living in rural communities in emerg-
ing parts of South Asia and sub-Saharan Africa learn and are motivated by their
peers in similar ways. Although they may not have access to the Internet or band-
width or even electricity, these individuals learn by observing their neighbors’
fields, by asking others about the crops they grow and how they grow them, O
inquiring about neighbors’ health issues and how they treat them. Development
agencies, from government departments to the NGOs that work with these rural
communities, are critical catalysts in this learning process.
The Digital Green approach, currently deployed in India, Ghana, E
Ethiopia, was founded in the belief that video can be a powerful tool to increase
the effectiveness of agricultural extension, but that its benefits cannot be fully real-
ized unless it is instituted through a process of localized content creation, facilitat-
ed dissemination, and institutionalization within broader extension processes.
Kerry Harwin is a New Delhi-based development consultant and journalist. His
professional background includes agricultural, energy, environmental policymaking
in the U.S. Senate, energy and climate policy advocacy with the Shakti Sustainable
Energy Foundation, and monitoring and evaluating with Digital Green. His articles
have appeared in publications including Foreign Policy, GQ, Times of India, IL
Hindu, and several Indian periodicals.
Rikin Gandhi is CEO of Digital Green. Rikin is a licensed private pilot, and he holds
patents for linguistic search algorithms that he helped develop at Oracle. Born and
raised in the U.S., Rikin ventured to rural India to start up a social enterprise to
develop biofuels. He then joined Microsoft Research in Bangalore as a researcher on
the Technology for Emerging Markets team that incubated Digital Green, a not-for-
profit organization supported by the Bill & Melinda Gates Foundation, the UK’s
Department for International Development, USAID, Google, and others.
© 2014 Kerry Harwin and Rikin Gandhi
innovazioni / volume 9, number 3/4
53
Scaricato da http://direct.mit.edu/itgg/article-pdf/9/3-4/53/705119/inov_a_00216.pdf by guest on 08 settembre 2023
Kerry Harwin and Rikin Gandhi
HUB-AND-SPOKE SYSTEM TO ENSURE LOCALIZATION
At Digital Green, we train development agency employees and people in the com-
munities where they work how to produce and distribute content, mainly in the
form of videos. These videos, which primarily feature information about better
farming techniques and nutrition practices, are shared in groups formed by moti-
vated and talented community members. Produced by and for the community, IL
videos spur an ecosystem of educational, entrepreneurial, and entertaining con-
tent, acting as a kind of village social networking platform.
Those of us with regular Internet access are interested in the information that
we see in our social media news feeds because it has been shared by people we
know and trust. In a similar way, when farmers assess the relevance and trustwor-
thiness of a Digital Green video, they consider not just its language but also factors
like the clothes the featured farmer is wearing and the type of dwelling she lives in
to determine whether that person is someone they identify with. Infatti, viewers
often ask the name of the individual featured in the video and the village she lives
In. Seeing is often believing for members of rural communities, especially women
who have a low level of literacy, for whom visual cues about a practice pertaining
to a person or a crop can be crucial in their determining its applicability.2
We use a hub-and-spoke approach in the production and distribution of con-
tent to mitigate the need to produce a video in every village while still ensuring a
high degree of localization. Roughly 80 percent of the videos an individual views
in her village are produced in the district where she resides; the other 20 per cento
might be from an adjacent district with a comparable linguistic, sociocultural, E
agro-ecological context. Localization also means ensuring that farmers have access
to all products or services (inputs) needed to convert the practices learned on the
Digital Green videos into concrete action on their farms.
VIDEO AS AN AGRICULTURAL TEACHING TOOL
Traditional agricultural extension methods rely on highly trained experts who go
out into the field to interact directly with farmers. Given the nearly 700 million
Indians who live in rural areas, most of whom rely on agriculture as their primary
source of income, the barriers to extension on a large scale are huge: there is not
sufficient human capital available, farmers grow a variety of crops, they may not
speak the same language as the mediator, and an inadequate transportation infra-
structure can make it difficult for extension agents to reach the rural communities.
Digital Green was conceived as a way to extend the reach of extension systems. Noi
identified video as a primary tool for communicating with farmers, based on a rig-
orous testing process that compared cost and adoption rates when farmers were
trained by experienced extension agents, viewed informational posters, listened to
radio programs, watched training videos with a facilitator to mediate the experi-
ence, and watched videos without a facilitator.
54
innovazioni / Digital Inclusion
Scaricato da http://direct.mit.edu/itgg/article-pdf/9/3-4/53/705119/inov_a_00216.pdf by guest on 08 settembre 2023
Digital Green
We found that facilitated video viewing can spur farmers to adopt new agri-
cultural practices for about one-tenth of the cost of traditional extension systems.3
We believe that much of the success of facilitated video stems from its ability to
demonstrate localization. Farmers featured in the videos are usually from the same
district as the viewers, and because the videos are shot in the farmers’ own fields,
the viewers can compare the condi-
tions in the videos to those in their own
fields. This builds trust in the informa-
tion shared.
The videos leverage homophily,
which is the human tendency to trust,
associate, and bond with those who are
similar to us. Farmers participating in
the Digital Green approach report that
viewing a practice on a video while
being told about it by a facilitator
boosts their recall, which improves the
effectiveness of video in inspiring
changes in rural behavior.
To build even deeper confidence,
We found that facilitated
video viewing can spur
farmers to adopt new
agricultural practices for
about one-tenth of the
cost of traditional
extension systems.
the facilitators, who typically live in the same village as the farmers viewing the
video, are on hand to ensure that viewers understand the processes being demon-
strated. These facilitators are a key element of the Digital Green approach, E
they often vouch for the local applicability of the practices taught in the videos.
Their similarity to the people viewing the videos also means that they are consid-
ered trustworthy. Equally importantly, they provide structure to the video screen-
ing process, which promotes engagement and learning. They ensure that all par-
ticipants are present and engaged, and before starting a video, they pause through-
out the screening process to ask questions, which helps to ensure that the partici-
pating farmers are grasping the information. Facilitators also help link farmers
with necessary inputs (such as seeds and fertilizers) or provide information about
markets.
CHOOSING CONTENT FOR VIDEOS
The topics for Digital Green videos are selected in two ways. The first is through
people participating in the Digital Green approach, who suggest topics they
believe could create value for them. Through COCO, our in-house data manage-
ment tool (discussed below), we are able to track all questions asked and com-
ments made by farmers viewing the videos, which allows us to identify needs at the
farmers’ level.
Video topics also are selected by agricultural experts who work with our field
partners and are familiar with the local context. These experts help to identify
knowledge gaps and promote content that addresses them. Per esempio, a farmer
innovazioni / volume 9, number 3/4
55
Scaricato da http://direct.mit.edu/itgg/article-pdf/9/3-4/53/705119/inov_a_00216.pdf by guest on 08 settembre 2023
Kerry Harwin and Rikin Gandhi
may know she needs a low-cost solution to a pest problem, and her questions
about that topic could guide our partners in creating new content. Or, farmers
may not be aware of an improved seed variety, such as drought-resistant rice, E
our agricultural experts can introduce them to it through a Digital Green video.
Once a video topic is identified, it is put into production by teams that operate
at the district level, as there is likely to be a greater degree of homogeneity among
farmers within a given district in terms of class, linguistico, and agro-climatic char-
acteristics. Members of the video production teams, who are typically high-per-
forming farmers or Digital Green partners’ frontline extension staff, are trained by
Digital Green staff, with support from our partners. Our training is increasingly
supplemented by facilitated video training modules—in a sense, we’re using the
Digital Green approach in our own training regime.
When a team gets ready to produce a video, they work with our subject-matter
specialists to create storyboards, which are then vetted by scientists working with
our partners. After the videos are shot and edited by the production team, they are
again vetted by subject matter specialists and cross-checked by Digital Green staff
to ensure their technical clarity and soundness before being disseminated in the
community.
LOCAL FACILITATORS
When paired with human facilitators, people in rural communities readily recog-
nize the value of the video content, as the facilitators put the practices demonstrat-
ed in the videos into the context of the farmers’ daily lives, provide follow-up sup-
port, and connect information with the inputs needed to take action. Even sophis-
ticated Internet users with broadband connections and access to abundant infor-
mation look to informed peers to help process their experiences. This kind of
guidance is especially crucial in rural communities where people have a low level
of self-efficacy and education.
The facilitators we work with are typically already engaged with our partners
as community health workers or agricultural extension agents, and they are core
constituents of our approach. We began by embedding our team members in the
offices of our partners to train the facilitators to handle a mobile projector and use
videos to stimulate interactive discussions.
This approach helped us appreciate the local context, but our ability to scale
was limited until we made our partners responsible for driving the Digital Green
activities themselves. We established regional offices to provide training and sup-
port to our partners and saw the rollout of our approach accelerate. Within the
next two years, we expect to increase the number of facilitators we work with from
5,000 A 10,000.
Facilitators typically work through self-help groups led by women. The groups
Digital Green works with tend to be previously formed, although in some loca-
tions it has provided video content to help form and strengthen groups. IL
groups meet regularly and often are involved in activities like group savings.
56
innovazioni / Digital Inclusion
Scaricato da http://direct.mit.edu/itgg/article-pdf/9/3-4/53/705119/inov_a_00216.pdf by guest on 08 settembre 2023
Digital Green
Working with groups that have strong bonds apart from the Digital Green inter-
vention provides a reliable venue for video dissemination. The women who attend
these groups often lack other formal sources of farming information; although
some information is available through newspaper, radio, and television, it often
lacks local specificity.
Facilitators select videos from a library of seasonally appropriate content, E
they also rescreen videos at the request of the farmers. It isn’t uncommon for
groups to view the same video multiple times, which helps them gain confidence
in their ability to implement the practices seen successfully.
We also have learned that, when working through front line village field work-
ers, it’s important that facilitators are screening videos in the most effective way
possible. Therefore, Digital Green instituted a multilevel monitoring system to
oversee and grade facilitators’ presentations. The monitors use a checklist of desir-
able behaviors to grade facilitators, such as maintaining eye contact with farmers,
asking follow-up questions, and maintaining proper attendance records. Digital
Green partners conduct periodic checks, which are verified by Digital Green staff.
We also periodically conduct large-scale quality audits using independent enu-
merators. This quality monitoring has allowed us to identify skill gaps rapidly, E
we have provided refresher training to facilitators to ensure the quality of media-
tion in all locations.
OUTREACH TO MARGINALIZED COMMUNITIES
Traditional agricultural extension reaches out primarily to male heads of house-
hold in farming communities, reflecting the view that the man of the house is the
primary farmer. Tuttavia, the bulk of agricultural labor done on small farms in
most of the developing world is in fact done by women. Another issue is that agri-
cultural extension programs have long been plagued by “elite capture,” which
means that powerful segments of rural society are able to monopolize extension
efforts. At Digital Green, we believe that one reason for our success is that we
reach out to women and other marginalized farmers; Infatti, women account for
79 percent of the people participating in Digital Green screenings. In keeping with
our strategy of leveraging homophily to increase the effectiveness of our approach,
most of our facilitators and the farmers featured in the videos are female. Digital
Green also is active in many communities with predominantly tribal populations,
particularly in the Indian states of Jharkhand and Odisha. By bringing these pop-
ulations together and featuring their peers as role models in the community, we
help to bolster their local social standing.
We’re unable to quantify this particular effect of our efforts, but Digital Green
program participants say that the knowledge they gain through our facilitated
screenings has bolstered their decisionmaking power within their households.
Although our workshops are largely focused on farming decisions, we believe that
women who are better informed will have more productive farms. We also hope
that our efforts will also help to promote gender equality in rural India.
innovazioni / volume 9, number 3/4
57
Scaricato da http://direct.mit.edu/itgg/article-pdf/9/3-4/53/705119/inov_a_00216.pdf by guest on 08 settembre 2023
Kerry Harwin and Rikin Gandhi
Figura 1. Digital Green Analytics Dashboards
DATA TRACKING AND ITERATION
We constantly track and use data and feedback to inform the production of our
videos and target their distribution more effectively. Digital Green’s Connect
Online, Connect Offline, affectionately referred to as COCO, is an open-source
data management tool that allows us to monitor which farmers are attending
viewings, which videos they’ve viewed, which they’ve asked questions about or
expressed interest in, and which agricultural practices they’ve adopted after view-
ing one of our videos. These data are important, as they allow us to understand
58
innovazioni / Digital Inclusion
Scaricato da http://direct.mit.edu/itgg/article-pdf/9/3-4/53/705119/inov_a_00216.pdf by guest on 08 settembre 2023
Digital Green
both the successes and the challenges faced by each region, group, facilitator, E
farmer.
Working in areas with limited electricity and telecommunications infrastruc-
ture can complicate technology-based solutions, and relying on paper-based
forms that are entered into COCO at a central hub can slow our access to timely
dati. To improve the timing and accu-
racy of our data reporting, we have
worked to make COCO simple to use,
even by those with limited computer
literacy, and have developed a mobile
version. To enable uninterrupted data
collection in areas with intermittent
Internet connectivity, COCO has been
built using
technologies enabled
through the HTML5 standard. IL
capabilities of this architecture make
COCO accessible across browsers,
whether a tablet, laptop, or smart-
phone. The tool also serves as a generic
library, which enables the open-source
community to use other programs and
organizations to collect data, both online and offline. We are currently deploying
COCO on the mobile phones of facilitators across our project sites, and we have
seen a fall in reporting time where it has been implemented.
“When we see videos and
do it, then we can believe
in it. Belief comes by
doing, not just by seeing.
So after seeing the video
we have to practice it.
Then we can believe.”
COCO feeds data into Digital Green’s analytics dashboard (see Figure 1),
which provides a detailed snapshot of the progress of our programs.4 This allows
us to monitor the pace at which presentations are carried out, attendance at those
presentations, and adoption rates, whether by country, state, village, group,
farmer, or practice, as well as other project data, such as types of practice or par-
ticipants’ gender. Inoltre, these data are publicly available, like all Digital Green
videos, so other organizations working on changing agricultural and health behav-
iors are able to learn from our successes and failures by viewing the adoption rates
of various types of practices and videos.
COMMUNITY USAGE AND FEEDBACK
The Digital Green intervention has reached about 420,000 farming households
through 296,000 video screenings, and has induced 167,000 farmers to adopt at
least one new practice. We estimate that the adoption of Digital Green practices
will increase farmer incomes (yield) by 20 percent or increase their input costs by
15 per cento. In one early study of an NGO that adopted the Digital Green
approach, we found that a typical farmer who changes just one practice due to the
Digital Green intervention will earn an additional $175 annually. We are currently
innovazioni / volume 9, number 3/4
59
Scaricato da http://direct.mit.edu/itgg/article-pdf/9/3-4/53/705119/inov_a_00216.pdf by guest on 08 settembre 2023
Kerry Harwin and Rikin Gandhi
conducting a large-scale randomized control trial to evaluate well-being, income,
and productivity measures among our participating farmers.
“Previously we weren’t even able to grow 800 kilos of grain,” remarks Sugna
Bai, a farmer in Madhya Pradesh. But after watching a video on organic pest con-
trol methods for wheat cultivation, her yields increased substantially. “Now we
harvest 3,000 A 3,500 kilos of wheat. When we see videos and do it, then we can
believe in it. Belief comes by doing, not just by seeing. So after seeing the video we
have to practice it. Then we can believe.”
To bolster our data management system, we have instituted a series of in-
depth perceptual workshops with our partners and participants. In these work-
shops, we consult with farmers, mediators, and partner staff to enrich our quanti-
tative data through detailed discussions of what does and doesn’t work in the
intervention. Farmers have reported that they overwhelmingly prefer video to
other types of agricultural extension, due to several factors:
(cid:2)(cid:1)The ability to view videos repeatedly aids recall of the practices taught
(cid:2)(cid:1) The opportunity to discuss featured practices with facilitators makes farmers
more comfortable adopting them
(cid:2)(cid:1) Illiterate farmers report that they are more likely to participate in video-led
extension than other forms
(cid:2)(cid:1)Allowing farmers’ family members to attend screenings increases practice recall,
as multiple members of a household can reinforce one another’s memory
(cid:2)(cid:1) Videos that include shots of packaging for required inputs increases farmers’
ability to ensure that they are buying the appropriate type and quantity of inputs
from agricultural dealers
INSTITUTIONALIZATION
Community facilitators and forums that nurture peer-to-peer sharing, such as
women’s self-help groups, engender a level of trust and understanding by flipping
the traditional top-down process of content production and delivery. Broadcast
television and mobile services have great capacity to scale to the masses, but view-
ers’ ability to translate the information these services provide into actions in their
fields or households can be limited. We have found that a decentralized content
production and distribution process that is integrated into existing public, private,
and civil society development efforts can be more efficient than approaches that
rely on high-cost, high-quality, one-size-fits all content.
Recording a one-to-one demonstration of a new practice on video and taking
it to many improves the efficiency of an extension service, but the process of pro-
ducing and distributing videos can improve the quality and systemization of our
partners’ development efforts more broadly. Videos become an artifact of the
information that our collaborators are already sharing through costly face-to-face
means, like demonstrations and group trainings, and bring consistency to their
knowledge bases. Feedback and usage data recorded at each video screening help
target content and guide our partners’ interventions to address the needs and
60
innovazioni / Digital Inclusion
Scaricato da http://direct.mit.edu/itgg/article-pdf/9/3-4/53/705119/inov_a_00216.pdf by guest on 08 settembre 2023
Digital Green
interests of individuals in a community more efficiently. The Digital Green video
screenings also introduce a dimension of knowledge-sharing to self-help groups
that are involved in other activities, such as microcredit and savings. Inoltre,
the content, dati, and feedback hosted on Digital Green’s website foster sharing
and collective learning among our partners. With nearly one million views of the
videos posted online in the last year, our open-access platform stimulates account-
ability and transparency between partners and external audiences, like researchers
and the general public, who in turn can contribute knowledge and leverage the
data shared by our growing grassroots network of communities to support their
own work.
Digital Green also creates tools that help our partners become more respon-
sive to farmers’ needs. As we’ve grown, we’ve discovered that many extension
organizations want to institute processes for bottom-up content development, Ma
they lack the capacity to do so. To address this, we and our partners have conduct-
ed stakeholder workshops in which we consult with farmers, facilitators, and part-
ner staff about content. We and our partners have also used our COCO monitor-
ing data to conduct bottleneck analyses that help identify gaps in our outreach to
farmers. By helping our partners incorporate consultative processes, we are sup-
porting a model of rural development that’s grounded in farmers’ needs and
desires. And we don’t do this only because with think it has intrinsic value; we
know that farmers are more likely to adopt our solutions when they solve the
problems they consider most pressing.
Over the last five years, Digital Green has engaged more than 380,000 farmers
In 5,000 villages across India, Ethiopia, and Ghana. Allo stesso tempo, technology
and content are only good at magnifying human intent and capability. Digital
Green has been successful because it has developed its videos using local content
attained through a bottom-up selection, which ensures that the messages in our
videos are accessible to all farmers, and we have been institutionalizing our
approach within established extension systems. When a video sparks the curiosity
of people in rural communities, it can help them take small steps toward improv-
ing their lives and of those around them.
1 “Agricultural extension” is a general term meaning the application of scientific research and new
knowledge to agricultural practices through farmer education.
2 Available at http://www.worldbank.org/content/dam/Worldbank/Feature%20Story/Africa/afr-
tanguy-bernard.pdf.
3 R. Gandhi et al., “Digital Green: Participatory Video and Mediated Instruction for Agricultural
Extension.” Information Technologies & International Development, 2009.
4 The analytics dashboard is available at http://analytics.digitalgreen.org.
innovazioni / volume 9, number 3/4
61
Scaricato da http://direct.mit.edu/itgg/article-pdf/9/3-4/53/705119/inov_a_00216.pdf by guest on 08 settembre 2023
