Data is growing in importance across all sectors, including agriculture. With the advent of new digital technologies and innovative business models, the amount of available data and potential use cases are increasing. Agricultural and Fintech innovators that recognize this trend are utilizing data in new ways, from creating digital platforms to delivering new, data-enabled services. Many are also exploring data partnerships, combining the power of multiple datasets to create greater impact for smallholder farmers.

In a new case study, ISF Advisors examined 33 engagements between Mercy Corps’ AgriFin program and 14 partners across four different countries. We found that around a quarter of the portfolio of AgriFin engagements—spanning various use cases—featured a strong data-sharing component. Underpinning many of these engagements are complex negotiations about how data sharing can unlock service delivery and enable different social and commercial outcomes for different players.

By analyzing these 33 engagements, we have uncovered lessons about the common barriers faced by data-sharing arrangements. In distilling these lessons, we hope to provide practical guidance and tools for overcoming these barriers to the broader ecosystem of actors involved in optimizing data sharing for agriculture.

The promise and role of data sharing in agriculture

Smallholder farmers’ remote location and lack of linkages to global markets have traditionally hindered robust data collection. But the penetration of mobile phones and other digital technologies into rural areas has made data collection and sharing significantly easier in recent years. New technologies such as drones, satellites, and sensors have also expanded data collection options.

Simultaneously, private companies are increasingly seeing farmers as potential customers and are turning to data to expand their understanding of this market segment. By entering into data-sharing partnerships, these companies can leverage combined datasets to develop new, tailored products for smallholder farmers, integrate farmer risk scoring, and optimize customer interactions. Some companies are even utilizing combined datasets to create entirely new business models where data itself is the commodity, sold directly to customers or to businesses that want to know more about their customers.

Despite the increasing use of data in agriculture, it is still in a nascent stage of development. At a foundational level, practitioners don’t even have a common data taxonomy to talk about what agriculture data is—making it difficult to learn from what others are doing. Additionally, many investments into new data-enabled platforms, models, services, and systems are still working out how to operate profitably at scale. Data-sharing engagements, while promising, tend to lack sophistication on several levels: 1) the types of data being shared (limited primarily to demographic data); 2) the format of data sharing (primarily static reports); 3) the level of analysis applied to the data (primarily simple analysis at the farmer level; and 4) the types of data-sharing agreements (primarily bilateral).

Barriers to effective data sharing

As might be expected in such complex arrangements involving a variety of actors, analysis of early data-sharing partnerships shows a range of barriers. Factors like the type of partners, type of data, use case for the data, and country where data is shared typically shape which barriers a partnership will face.

Looking at the AgriFin portfolio, we can see that the primary barriers to establishing a data-sharing agreement can be classified as:

• Cultural (e.g., the company’s leadership or established internal data-sharing practices);
• Capacity (e.g., the skills, technology, and experience each partner brings to the table);
• Commercial (e.g., example, the cost of data and how partners treat intellectual property and/or competition in the data space);
• Reputational (e.g., how partners think about the riskiness of sharing personal data); and
• Regulatory (e.g., national data policies and legal jusdictions).

Even for organizations that recognize the potential of data in agriculture, these barriers can prevent them from effectively assessing the business case for investment within different regulatory environments. Once a data-sharing agreement is created, the barriers tend to shift. For example, many organizations don’t have a dedicated team working on the engagement, which can lead to delays in sharing data. Data-sharing agreements also suffer when the partners don’t have on-the-ground staff members who understand how to work with disaggregated farmers.

Certain barriers are more apparent with certain types of partners. For instance, government and nonprofit actors often face skill and capacity barriers, while financial service providers often contend more with cultural and regulatory barriers. While each use case is unique, we have distilled a common taxonomy of reference barriers and a mapping of where they’re most likely to show up in different partnerships (see full case study for more detail).

Empowering effective data sharing