BD4D Conversations About Data

January 14, 2026 | M Celine Takatsuno

As the Better Deal for Data team prepares for the v1.0 launch of the BD4D Standard and the BD4D Playbook, we are excited to share some of the steps we took to get there. This look at our Data Consultations is the first of the series.

For more than a year, in well over one hundred conversations, the Better Deal for Data team has been listening to, and learning from, social impact practitioners to better understand how data flows across communities and organizations.

During the first half of 2025, we focused these efforts primarily on gathering real-world data use cases from individuals and organizations working with data. Our intent in these Data Consultations was to learn about:

• types of data used and how it is gathered;
• the data stakeholders;
• reasons why this data is collected and/or shared, and any challenges in doing so; and
• effects of legal requirements related to their use of data.

We held ad-hoc data conversations, informal ‘fill-in-the-data-blank’ data sessions, and a set of semi-structured interviews with people and communities around the globe. While we had a concentration of participants representing agriculture, food, and farming, we also met with individuals working in conservation, education, philanthropy, health, social services, and tech-for-good.

Data Madlibs

Our fill-in-the-data-blank sessions were conducted adjacent to larger virtual and in-person events, and were inspired by the familiar Mad Libs^Ⓡ word game. These “Data Madlibs” provided a quick look at how participants…

• collect [types of data] by doing [data collection activities];
• want to use data to accomplish [specific social good] for [specific people or community]; or
• want to make sure that [specific bad things] don’t happen with the data collected from [specific people or community].

You can still participate in our Data Madlibs! If you’re a nonprofit working with data, we invite you to take a moment and fill out an online BD4D Data Madlibs too!

Data Interviews

While our Data Madlibs offered a quick glimpse across a breadth of ways that organizations think about data, our semi-structured data interviews gave us a much deeper look. These consultations were directly with people who manage data for their community, project, or organization, including two academic researchers, three practitioners who provide direct social services, and four who offer a platform for data collection or exchange. We also interviewed individuals whose data is collected and used by nonprofits, and representatives of membership associations representing more than a thousand farmers and ranchers across the Americas.

We followed a structured question-and-answer format to ensure consistency, and invited additional comments from participants once the questions were done. Our interview questions were as follows:

• What types of data are collected, or would you like to collect?
• From whom do you collect the data?
• What is the benefit or purpose of collecting the data?
• What mechanisms and methods are used to collect, store, and/or share the data?
• What challenges affect your ability to collect or share the data?
• With whom are the data shared? What agreements, if any, are made, with any of the data stakeholders?
• What regulations apply to data sharing in your region? How do they affect collection or sharing?

What We Learned About the Data

While most participants discussed data that they collected directly from individuals and communities, some pointed to data they used, but gathered indirectly. In these cases, the data fell into two broad categories:

• Observed data, which is either (a) collected or sensed via automated means, or (b) recorded by a nonprofit representative based on interactions with a client; and
• Partner or intermediary data, which includes data collected by a contracted third party.

Generally, data was was one of six types:

• Client personal information, which may include identifiable or other sensitive data;
• Case management, such as records of interactions, assessments, or interventions;
• Demographic data;
• Lived experience information related to social situations and well-being;
• Program performance metrics; and
• Employee or workforce data.

Agricultural data often included highly technical information, like characteristics related to crop, harvest, seed, and soil; livestock grazing, herd, and habitat; precise boundary, location, and mapping; water; and more.

In all areas, data might be collected for academic research; decision-making and management; funding and regulatory compliance; or program monitoring and evaluation. Several organizations are building data exchanges, which aggregate raw data and datasets from individual and institutional participants who then have access to a shared data platform. This data can then be used to identify efficiencies, improve outcomes or productivity, or to promote and provide opportunities for collaboration or markets.

Interestingly, in agriculture, numerous interviewees reported that land producers–the farmers and ranchers–often saw data sharing as transactional. They were willing to provide their data in exchange for a service or information that would enhance or improve their business or yield. For example, allowing a third-party consulting firm to use crop, land, and financial data in exchange for support with filing funding applications or government forms.

Challenges and Concerns

“We’d love to share more data if we knew it wouldn’t be used against us.”

– Leader of a nonprofit cooperative

While some of the challenges encountered by participants supported our initial assumptions, others were unexpected. Across all respondents, common challenges were often related to either data literacy, given that data policies are often embedded in lengthy, complicated legal agreements, or nonprofit capacity, as organizations don’t have the funding or the staff to dedicate to data governance.

Data collectors expressed concern about accuracy, completeness, and validity of self-reported data, citing “data fatigue” resulting from numerous forms and surveys asking for the same information, and lack of access and accessibility of digital tools for data collection. Environmental barriers presented a further challenge to those working with field data, including weather, remote terrain, and concern for personal safety.

However, it is trust, whether in people, organizations, governments, or systems, that was the single most common theme across all of our conversations. Many participants voiced a fear, based on their own or a community member’s experience, that data would ultimately be used to exploit them. Land producers and stewards might only be willing to provide what is minimally required to obtain funding, permits, or government designation, unless it is collected by someone with whom they held a trusted relationship. And, in human services nonprofits, the risk of losing individuals’ data, and the trust of those they serve, was described as paramount, making data minimization and limited sharing common practice. Said the director of a local youth program, “we collect only what we need, nothing more. And we don’t share (raw data) with anyone else”.

Takeaways and Impact

“Trust…is earned through shared goals and values.”

– Agricultural association project lead

Through our many conversations about data over the year, three important lessons emerged which have informed our approach and our future roadmap:

1. Context Matters.
   Organizations with highly specialized contexts may require more specific guidance on adopting the BD4D Standard. However, we found that even between these specialized contexts, similar uses of data exist. This has led us to broader categorization in the BD4D Playbook, for example, describing how BD4D can be applied to certification data and sensitive data rather than narrowly on agricultural land use, or health data.
2. One Size Fits Many, Not All.
   As a data manager at a health and human services NGO stated, “there was never a ‘one size fits all’ that governs data sharing, so we are looking for a universal, barebones standard.” Thus, in revising the BD4D Commitments, we sought to provide a straightforward floor for data governance that can easily be adapted across sectors and use cases.
3. No Surprises is Essential.
   Building trustworthiness and transparency are crucial for effective data stewardship, and an organization’s actions should meet the expectations of those they serve, and not surprise them. For adopters of BD4D, this means engaging individuals and communities in designing practices involving their data, and asking critical questions: Do constituents have a relationship with the organizations collecting and using their data? Do they know what is being done with their data, and who will benefit? Do they have a way to access it or remove it?

The lessons we’ve learned in our conversations about data now anchor the BD4D Standard in real-world use, and the trust that comes through shared goals and values. In shifting our focus away from creating model agreements and licenses, and towards supporting organizations with a way to declare and demonstrate their commitment to trustworthy data practices, we hope to seed a movement across the social sector. We are grateful to all those who have contributed to our work thus far, and look forward to growing this community of practice, as we co-create a better deal for data, for all.

We’re looking forward to the upcoming v1.0 release of the full BD4D Standard, and the BD4D Playbook to guide its real-world implementation. If you’d like to to be among the first to know when it does, make sure to sign up for updates here!

*Data Consultations were conducted by Jim Fruchterman, Celine Takatsuno, Steve Francis, Katy McKinney-Bock, Courtney Tiberio, Derek Caelin, and Amaya Webster.