7 challenges the agriculture sector must address to unleash its data revolution

Farmers in Tanzania are trained to identify what crops would work well in an area based on gathered data on crop temperature and amount of rainfall. How can enthusiasts and specialists help unleash the data revolution in the agriculture sector? Photo by: Cecilia Schubert / CCAFS / CGIAR / CC BY-NC-SA

Open data enthusiasts gathered last week in Ottawa, Canada, as part of the International Open Data Conference, to reflect on what the data revolution means for agriculture and nutrition.

From filling data gaps —such as identifying the causes of food loss — to helping farmers with weather or market updates, to providing government agencies with key information regarding crop diseases, open data holds the promise of providing many solutions to problems related to food systems.

Experts and advocates reflected on the challenges that lie ahead in order to make that promise a reality. Below are seven main ways enthusiasts and specialists can help unleash the data revolution in the agriculture sector.

1. Open up data.

Open data doesn’t necessarily mean new data. A large part of the data revolution rather lies in bringing governments, donors and businesses to make existing data available.

“One of the biggest points open data can help with is the waste of billions of dollars that have been spent in data collection,” said Stanley Wood, senior program officer for data, evidence and learning at the Bill & Melinda Gates Foundation. He added that much of the results these efforts yielded were nowhere to be seen — pushing for data sets to be released would increase accountability of donor and implementing organizations.

One key aspect of the data revolution heard throughout the conference was how to exploit and formalize data sharing processes that already take place between organizations.

“Agriculture has a long history of data sharing,” said Matthew McNaughton, executive director at SlashRoots Foundation. “You get a lot of sharing among meteorological offices or between agriculture agencies in the Caribbean. The agencies and organizations we spoke to don’t perceive that as open data.”

2. Identify data users.

Policymakers, researchers, project implementers and farmers all have different needs when it comes to data. A data-driven agency will go beyond making information available and identify targeted users to make sure the data is available in formats that make sense to them.

“We have about 50 data sets in our portal, but [they] don’t translate into anything meaningful for the farmers,” noted Nkechi Okwuone, who manages open data for the state of Edo, Nigeria. In other words, availability doesn’t always translate into accessibility.

3. Bring intermediaries into the game.

Turning data into meaningful information, however, should not necessarily be the responsibility of the institution that owns the data. All panelists agreed there was a need for a middle person to take on the role of analyzing and packaging data sets into usable formats. Some of them saw room for a variety of actors such as entrepreneurs and developers to create new products, but also pressed the sector to create incentives for them to come on board.

“You have these smart people, these developers who don’t know anything about farming,” McNaughton said, adding these intermediaries need increased access to institutions, experts and farmers.

Steve Adler, chief information strategist at IBM, said the question of who should fill in the middle gap is a pressing one.

“Whether we do it or others do it, we’re replacing the government in those countries. That’s a concern for the stability of information,” he said. “I think there’s an opportunity to develop that role of the agricultural data scientist. If we don’t fill it, it will be filled by the private sector, in ways that we don’t always appreciate.”

4. Develop new tools for data collection.

With the help of new technology and platforms, data collection is bound to become more efficient and cheaper, and many believe it’s already time to devise new tools and methods to collect data. Wood called for the sector to be creative and think of innovative ways improve existing data collection efforts.

“One of them might be not to send people in Toyota Land Cruisers to collect information,” the Gates Foundation official suggested, “but actually have farmers with cell phones who would collect information.” He added a new model could involve compensating farmers to send data, which could be more cost-effective than conducting surveys.

5. Look beyond technology.

Participants identified several challenges to implementing the data revolution that technology alone cannot solve. One of them was the accessibility of data.

“In Jamaica, the average age of the farmer is 67,” McNaughton said. “If we’re really serious about improving access to information, we have to think of ways that are not technocentric.”

Trust was also identified as a major obstacle. Sander Janssen, who leads the Earth Observation and Environmental Informatics team at the Alterra research institute in Wageningen, the Netherlands, said building trust was crucial to get accurate data from participating farmers.

Issues related to data literacy, which were at the heart of the conference, were brought up during the session dedicated to agriculture and nutrition as well. Adler urged the audience to reflect on the power dynamics the data revolution could create.

“We have a significant challenge around the world of how we make this information relevant to small stakeholder farmers who may be disenfranchised, and in many ways disadvantaged by the proliferation of open data, specially in communities where a large amount of farmers are illiterate and don't have access to information at all,” he warned.

6. Foster cross-sector collaboration.

Open data often serves as a common denominator of organizations, businesses and individuals with vastly different missions and skills. While the question of how to bring a variety of actors to collaborate was raised throughout the conference, agriculture actually has a head start on other sectors.

The Global Open Data for Agriculture and Nutrition initiative currently has 120 members from the private, public and nongovernmental sectors who reflect on ways to encourage data collaboration and build high-level policy. Together with the Open Data Institute, GODAN released a discussion paper during the conference, highlighting concrete examples of how open data has been used effectively in the agriculture and nutrition sectors.

7. Address the need for disaggregated data.

While open data experts and enthusiasts have addressed some of the most pressing issues brought upon by the data revolution, certain questions were left unanswered. Among them is the need for the agriculture sector to dedicate part of its open data efforts to crosscutting themes, such as human rights, for instance in the case of land grabs, or gender.

“It’s one of those cases where the data is way ahead of the open data,” Wood said of the latter. “All of the surveys that we fund, and probably other funders as well ... focus on the gender disaggregation of all of these activities. Those data sets are only just becoming more accessible and available.” He cited a recent report, “Leveling the Field,” released by the World Bank and ONE Campaign as an example.

Check out more insights and analysis for global development leaders like you, and sign up as an Executive Member to receive the information you need for your organization to thrive.

About the author

  • Flavie portrait

    Flavie Halais

    Flavie Halais is a contributor based in Montreal who covers cities and international social issues. In 2013-2014, Flavie was an Aga Khan Foundation Canada International Fellow, reporting for Nation Media Group in Nairobi, Kenya. She’s also reported from Rwanda, Brazil and Colombia.