Digital Green is an organization that uses video to improve agricultural extension in India. Over 600 million people depend on agriculture in India but small landholders earn less than $2 per day. Traditional extension methods are not scalable enough. Digital Green produces instructional videos with local farmers and distributes them through community screenings with a local mediator. Early results found Digital Green was over 10 times more cost-effective per adoption than traditional extension. The system is being expanded through a structured process and analytics platform to improve operations over time.
2. Agriculture in India
600M agriculture-dependent lives
Majority small landholders (<3 acres)
<$2 a day ($750 a year)
Growing debts ($300 per year per farmer)
Earlier technology interventionâŠ
â Green revolution had mixed results
âą Increased yields, butâŠ
âą Led to rising input costs, declining soil
fertility
âą Due to excessive use of A farmer from Yellachavadi village,
fertilizers/pesticides outside of Bangalore
Indiscriminate use of technology partially
responsible for current agrarian crisis
2
3. Agricultural Systems?
farmer expert
Low literacy
in local lang
Expensive No unique ID
No bank account Credit card
credit
Poor roads
Market
Quantity
Poor quality buyers
control
Computing device and connectivity not enough!
4. Agriculture Extension
Dissemination of expert agriculture
information and technology to farmers
âTraining & Visitâ extension popularized
by the World Bank in 1970s
â Face-to-face interactions of
extension officers and farmers
100,000 extension officers in India
â Extension agent-to-farmer ratio is
1: 2,000
â 610,000 villages in India with
average 1,000-person population
Extension officer âcommutingâ between farms
4
5. Agricultural Social Networks
?
Main source of information about new technology and
farm practices over the past 365 days (India: NSSO 2005) 5
6. The Problem
How can the speed and effectiveness
of agriculture extension be improved
at a reasonable cost?
Extension officer on-field demonstration
6
7. Digital Video for Extension
Video providesâŠ
â Resource-savings: human, cost, time
â Accessibility for non-literate farmers 7
8. Early Experimentation
Early Experimentation
Parameters Varied
Background of actors in video, Types of content,
Background of actors in video, Types of content,
Location and timing field trying various combinations
Six months in of screening, Method of dissemination,
Over 200Degree of mediation,Background of mediator, etc. design
Degree surveys, ethnographic investigation, and iterative
days of of mediation, Background of mediator, etc.
8
9. Digital Green System
Participatory Content Production
Introduction to innovations
â Standard extension
procedure
Rough âstoryboardingâ
â Repetitive pattern; easy to
learn
â Minimize post-production
Local farmers on their own fields
â Reduce perception of
âteachersâ
â Promote âlocal starsâ
9
10. Digital Green System
Video Database
Online video
databasehttp://www.digitalgreen.org
>2,100 videos of 8-10 minutes each
Quality-control, minor video editing,
and metadata tagging
Indexed by type, topic, locale,
season, crop, etc.
Distributed via memory cards
10
11. Digital Green System
Mediated Instruction
Local mediator
â Performance-based honorarium
Human engagement
â Field questions, capture feedback,
encourage participation
â Balance genders
On-demand screenings
â Choice time and place
â Not âstand-aloneâ kiosk
Support and monitoring
â Daily metrics and feedback
â Official extension staff
11
12. Digital Green System
Structured Sequencing
Community Practices with Practices with
Assessment short-term longer-term
visible rewards visible rewards
Group Participation
Audience
Awareness
Season
Location
Time
12
13. Digital Green System
1. Participatory content production
2. Video database
3. Mediated instruction
4. Structured sequencing
13
14. Preliminary Evaluation
Experimental Set-Up
15-month study
21 villages in Karnataka: Classical GREEN (8)
â Language: Kannada
Expert Same as usual
â Crops: Ragi, banana, mulberry, coconut Digital Green (9)
Research Assistant
â Population: 50-80 households 3 sessions per week
â Irrigation: 10-20 households with access
Extension Cost:
Rs. 9,500 ($240) for TV/DVD
â Television: 15-20 households
Officer
per village
PC / camera costs shared
Extension officer shared
Metrics: Mediator salary
â Knowledge: Before-and-after
Local Mediator Local Mediator Local Mediator Accountability:
â Attendance: Farmers at each screening Daily metrics and feedback
Official extension staff
â Interest: Intent to take-up a practice
Poster Green(3)
â
FarmingAdoption: Number of households Farming up
Farming taking Same as Digital Green with local
each new farming practice or technology
Community Community Community mediator, but no TV/DVD
Mediator makes posters and holds
regular group sessions
Audio Green (1)
Same as Poster Green with
MP3 audio tracks from videos
14
15. Digital Green: Early Results
7 times more adoptions over classical extension
Sustained local presence
90
80
Mediation
70
Adoption Rate (%)
60
Repetition (and novelty) 50
Classic GREEN
40
30 Digital Green
Integration into existing extension Poster Green
20
operations
10 Audio Green
0
Social homophily between mediator,
Jun-07
Oct-07
Jan-08
Mar-08
Jun-08
May-07
May-08
Apr-07
Jul-07
Aug-07
Sep-07
Feb-08
Apr-08
Nov-07
Cumulative
Dec-07
actor, and farmer
Desire to be âon TVâ
Trust built from identities of farmers 15 months:
and villages in videos 13 villages, 3 nights a week, 1,000 regulars
15
16. Cost-Benefit
Cost (USD) Adoption (%) Cost/Adoption
System /Village/Year /Village/Year (USD)
Classical GREEN $840 11% $38.18
Digital Green $630 85% $3.70
Poster Green $490 59% $4.15
Note: Decreasing amortized cost of hardware with time and scale
Digital Green is at least 10 times more effective
per dollar spent than classical extension! 16
17. Incremental Adoptions,
Incremental Incomes
$250
$200
$150
In first 8 months, adoption of improved
practices increased the incomes of farmers by
$100
an average of $242!
$50
$0
Jun-10 Jul-10 Aug-10 Sep-10 Oct-10 Nov-10 Dec-10 Jan-11
Chili -
NurseryChili Line Beans Line Sowing Bitter Gourd Pest & Ginger Rot Potato Line Sowing, Tomato IntercroppingSeed Poultry Rearing
Raising, Sowing, Beans FertilizerRice Intensification
System of Application Management
_ Improved Onion
Improved
17
18.
19. Digital Green System
Network Effect
Viral Web 2.0 in the Web-less world
- Content ecosystem: education, entrepreneurship, entertainment
- Cost-realistic access: pico projectors, TVs, DVD players, and camcorders
Reinforce existing social networks to diffuse innovations through communities
Local âidolâ competitions to be a better farmer
3
1 19
2
20.
21. Digital Green System
Platform
Online Offline (no/low connectivity)
Browser-based input
Cloud-based central database Data stored in local database
Synchronized with local databases Synchronized when connectivity available
21
23. Digital Green System
Platform
Robust system to share, track, and analyse data to manage operations and
target interventions over time
Analytics dashboard built on top of a simple yet robust data
entry system that can toggle between online and offline connectivity modes
http://www.digitalgreen.org/
100,000
simultaneous
offline users
Offline mode 10x
faster than online
23
26. Non-Non-Profit Digital Green
Subsidize agriculture Digital Greenâs value to farmers is
established â viewers contribute Rs. 2-4
extension with ads? per screening.
Could DG also be supported by ads?
Advertisers get access to a distributed,
captive audience with demonstrated
interest in better agriculture.
Ads follow Digital Greenâs distribution
channels.
To do:
â Scale Digital Green
â Devise mechanism for ensuring
appropriate ads
â Quantify ad effectiveness
Digital Green DVD title screen â Quantify ad value to advertisers
33. Agricultural Systems?
farmer expert
Low literacy
in local lang
Expensive No unique ID
No bank account Credit card
credit
Poor roads
Market
Quantity
Poor quality buyers
control
Technology not enough!
34. Technology magnifies human intent and capability.
Technology itself requires support from
well-intentioned, competent people or organizations.
Successful technology interventions work as a part of
well-intentioned, competent organizations.
The number of people dependent on agriculture in India is declining. Still, today, there are about 60% of people whose income depends on agriculture to a large extent.Majority of the farmers are smallholders and cultivate in a land of less than 3 acres.Agriculture has been getting increased attention, primarily because of the farming crisis.Farming related suicides has been increasing and in the period from 97 to 2005 there have been 150,000 suicides. Green revolution, which was a major event in Indian agriculture history, where India converted to chemical based farming in 1960s focused on high yielding varieties of crops.While yields have increased, there has been criticism from many quarters about it. Farmers have been using fertlizers/pesticides indiscriminately leading to rising input costs and decreasing soil fertility. There has been tension between whether to produce market oriented crops vs promoting sustainable agriculture. Lots of opinions, research, debates..
Agricultural extension is essentially the dissemination of expert agricultural information and technologies to farmers. Agricultural extension was popularized by the World Bank during the 1960s and 1970s in the form of the âTraining & Visitâ system. Today, India still has over 100,000 civil-service extension officers. This represents the second largest extension force in the world, but India has an even vaster population of farmers. Indeed, there is only 1 extension officer for every 2,000 farmers.
In 2003, the Government of India sponsored a National Sample Survey to understand the sources of information farmers were relying upon for new technology and farm practices. They discovered that the formal channels of extension â including, the âTraining and Visitâ-style extension and the governmentâs broadcast media programs â were reaching a small proportion of farm households. Instead, they found that farmers primarily relied on the informal channels of information diffusion that existed by âword of mouthâ in their own village communities.
Inspired by Digital StudyHallâs efforts in improving rural education, we began with the premise that digital video is a technology that can be taken to the last-mile and provide significant resource-savings â particularly, since the hardware has become so affordable. A one-to-one demonstration between an extension officer and a farmer could now be digitally captured and shown to many farmers who could easily relate to a visual media.
I spent over 200 days in the field working with a grassroots-level NGO, called GREEN Foundation, which follows the classic âTraining & Visitâ-based approach for agricultural extension. Through iterative design, we experimented with various parameters of a digital video-based extension model.For example, we considered the background of the âactorsâ that are featured in videos. On one hand, an agricultural expert can present highly-quality content. But on the other, we found that farmers were not receptive to being âlecturedâ to by outsiders of a very different socioeconomic demographic. Instead, they preferred to watch a fellow farmer share his or her experiences in a manner similar to the informal social networks that they were used to interacting with. On distributing these videos, we initially experimented with setting up a TV in front of a public square next to a stack of DVDs that farmers could choose to access as they pleased. Though there was an initial novelty, the community quickly became disinterested as they couldnât understand its purpose. We found that they needed a human mediator who could use the âvirtualâ on-screen demonstrations to engage farmers in a sustained learning and adoption process.
The process of producing the content starts by following the existing field activities of extension agents. The mere presence of the camera improves the productivity of an extension agent and a farmerâs interaction and rough âstory boardingâ helps reduce the need for post-production. Those extension agents or farmers that are featured in the videos know that they will be seen by others in their respective communities as role models. We want to stimulate this creation of âlocal starsâ. The actors featured in the videos are seen not as outside âteachersâ, but as local farmers who are adopting the practices or technologies within the constraints of local resources on their own fields.
We have been building a repository of this video content with our field partners for the last 1.5 years. The repository currently includes over 250 videos that average 8 minutes in length. The videos are reviewed by the experts of our partner organizations in content hubs, like Bangalore, and minor video editing and meta tagging is used to upload the content to a video database. Though these videos are mirrored online at www.digitalgreen.org, that is not how farmers actually view them. Villages are sent DVDs.
These DVDs are received by local village mediators. These mediators were chosen on the basis of local language literacy for record-keeping purposes and equitably balance genders to engage both men and women farmers. Indeed, these mediators mainly serve to engage the community by pausing and rewinding the videos, fielding questions, and encouraging group participation. Unlike broadcast programs or standalone kiosks, the mediators take the shared TV and DVD players to farmers at their choice time and place and serve as a feedback mechanism for farmers. Though weâve seen many mediators become resource persons in their local communities, its also important to note that they are supported and monitored by the official extension system.
Of course, a single screening is not enough. Weâve found that over time we can build trust and sustained participation in communities only by using a model for structuring the sequence of interventions. We begin the process by using the local mediator to help assess some general characteristics of his or her community, such as its socioeconomic and agroecological resources and needs. By seeding locally relevant content, an informal farmers group is created using the video programs as the focus. Then, since there is span of agricultural practices and technologies that have varying time-horizons for investment and reward, the mediators first showcase those practices that offer visible short-term rewards to gain traction in a community before showcasing a practice, like mulching, which has a longer gestation period to offering a gain in soil fertility.
These field trials allowed to converge on the Digital Green system which includes four unique components that Iâll briefly outline: (1) a participatory process for content production, (2) a video-based database, (3) locally mediated instruction, and (4) a model for sequencing content to build trust and sustained participation within farming communities.
After we developed Digital Green, we began a 9-month study to evaluate its effectiveness as an approach for agricultural extension. In 8 villages, we followed GREEN Foundationâs extension officers as they adhered to the classic Training & Visit model. In 8 other villages, we used the full Digital Green system that I have described, which includes a local mediator who is provided a shared TV and DVD player to conduct a 3 video screening each week in his or her community. Youâll notice that Digital Green still incurs the costs of being supported by the existing extension system, but its reach is extended by the presence of local mediators at the village level. Finally, as part of a preliminary study, called Poster Green, we selected four villages in which we follow the similar approach to Digital Green with a local mediator but without a TV and DVD player. Mediators in Poster Green villages hold the same frequency of regular group sessions in their communities and use posters and discussions instead of videos to communicate to their audiences.To compare these three modes of agricultural extension, we measured what knowledge was transferred to farmers, whether farmersâ attendance and interest sustained over time, and for our short-duration study, how many practices farmers ultimately adopted on their own fields. The 20 villages that were selected for this evaluation are roughly 2-hours outside of our MSR India office in southeastern Karnataka. The communities are comprised mostly of subsistence farmers that grow the staple crop, ragi, and the communities are comprised of between 50-80 households. Though most qualify for the governmentâs âbelow the poverty lineâ schemes, one-quarter have access to irrigation facilities and one-quarter have TVs in their own homes.
Early indications show that Digital Green system has increased the adoption rate 7-fold over classical extension. Over the last 9 months, weâve also seen that farmerâs participation has sustained even with the frequency of 3 screenings per week in each village. I should note that these figures only count whether a farmer has adopted at least one new practice during a particularly period. Also, there are a variety of practices that a farmer could adopt, but they were sequenced uniformly across each group of villages on a calendar-basis. The key factors that resulted in the substantial gain of the Digital Green system over the classical approach include the sustained presence of a local mediator who can engage the community. The on-demand nature of video technology offers the capacity for repetition to ensure that concepts are grasped as well as the novelty that is introduced by showcasing new farmers adopting practices. We have found that some farmers are incentivized to adopt practices just to be featured âon TVâ. This helps us reduce the perceived disconnect between experts and farmers, and allows farmers whose first questions are often âWho is this person in this video?â and âWhich village is he or she from?â to authenticate the viability of the content.
Now, letâs consider the per village costs of the Classical GREEN, Digital Green, and Poster Green models. Digital Green is supported by the same Classical GREEN extension system and adds the costs of village mediators, TVs and DVD players, a video camera, and a PC. Surprisingly, though, Digital Green costs less than Classical GREEN on a per village basis because each higher-paid extension agent essentially reaches more villages. In terms of benefit, as we saw in the previous slide, Digital Green can potentially increase annual adoption rates by over 7-fold. And, as a result of the lower per village cost and higher adoption rates, weâve found that Digital Green could be 10 times more effective per dollar spent than classical extension!
In addition to the cost-benefit calculations, weâve also seen some interesting side-effects of Digital Green. Weâve found that we can take concepts of Web 2.0 to the Web-less world where the infrastructure has yet to reach the last mile. We can build an ecosystem of content around farmer education, entrepreneurship, and entertainment by just using cost-realistic technologies like TVs, DVD players, and camcorders. We can take the existing social networks that farmers use for channeling information and expose them into public view through video. In some cases, weâve found that we can stimulate a sort-of âFarmer Idolâ competition where farmers want to be seen as the best farmer in their community and are incentivized to appear âon TVâ.
Over the last 2 years, the tribal communities that weâre working with in five Indian states have produced over 400 short videos that are by the farmers, of the farmers, and for the farmers. The content spans a variety of topics and genres and includes step-by-step demonstrations, testimonials, and interviews. Some farmers even compete to appear âon TVâ in a âFarmer Idolâ sort-of program to be seen as the best farmer and generates motivational âcurrencyâ doing so. The first two questions that farmers often ask when they watch these videos is âWhat is the name of the farmer in this video?â and âWhich village is he or she from?â to authenticate that the content comes from a source that they can relate with before considering a change in their behaviour.