Evidence demonstrating why research libraries and academic publishers should adopt open infrastructure (tools, systems, standards, protocols) has accrued more dramatically than ever in the first nine months of 2025. Social, political, and ethical pressures have elevated concerns about control, security, distribution, and exit strategies for knowledge assets and collections to the forefront of knowledge stewardship conversations. Even “open access” content that is held in closed and proprietary systems is at risk of myriad threats that we once considered more hypothetical than real, including radical “memory hole”-level dataset disappearance and deletion promoted by authoritarian leadership. Safeguarding content (e.g., activities around PubMed) and rescuing collections (e.g., public datasets) have become major expenses and activities, borne largely by volunteer networks.
So why is our field’s adoption of and investment in open research infrastructure still below that of proprietary software and approaches?
Answers to this question have previously hinged on safety, reliability, and predictability. Until recently, choosing “open” felt riskier. Skeptics expressed concerns that institutions without technical development staff could not safely host or integrate open software locally. They also cited the challenging and often painstaking processes that rely heavily on community decisions and volunteer energy to improve features and maintain development directions. Proprietary approaches, on the other hand, came with the benefit of “customerness,” in which the library or publisher paid someone to provide a service, outsourcing hosting, maintenance, and the responsibility to address issues that inevitably arose.
But now, the development and hosting challenges previously associated with “open” are changing. Perceptions of stability have dramatically increased as these open infrastructures and their communities of support have demonstrated their longevity and maturation over more than two decades. New players have also emerged, including skilled third-party IT shops that provide hosting, integration, and features-building services.“Open” carries less the staffing burdens (and stigma) of the past, and open solutions are increasingly recognized for being at least as stable, safe, and reliable as their proprietary counterparts. Still, as J. Robertson McIlwain wrote in 2023, “Despite anecdotal evidence that interest or discussion of open source in libraries is increasing relative to 10 years ago, that does not seem to have translated into significant adoption rates.”
Despite recognition of the advantages of open infrastructure, institutions still face challenges in identifying, procuring, and adopting open solutions. What information do adopters and investors need in order to find, procure, and use these open approaches to protect the knowledge assets they steward?
This year, a well-timed grant award from D//F has provided our team with resourcing to test what factors matter most in decision making for libraries, and what information might best equip those libraries that are not yet adopting open to widen their consideration to include these options.
Photo by AbsolutVision on Unsplash.
Our research so far: Evaluation involves values; values have themes
In our D//F-funded research project, “Measurement of Community Health Indicators” (MoCHI), we have sought to define specific themes that reflect the factors influencing decisions about technology adoption and investment. Our work began with identifying and interviewing a cross-section of open infrastructure stakeholders to better understand how they think about “open infrastructure,” and what strategies they deploy in order to decide whether or not to use, contribute to, or fund open infrastructures as part of their work (see Trust, transparency, and technology: Do community health frameworks shape open infrastructure decisions? and our podcast appearance on CHAOSScast for more details).
Based on our interview analysis, we have identified 11 key themes, which can be briefly summarized as follows:
- Affordability. Stakeholders who considered using open infrastructure often cited the challenges they faced in understanding the “real costs” associated with development and hosting, which is akin to pricing information for a proprietary product. They also discussed wanting more stories from user organizations about implementation and the costs incurred, as well as third-party options.
- Data ownership, portability, and control. Interviewees often discussed the growing appeal of having technical infrastructure options that did not entrap or hold their collections hostage in “black box” approaches. Here, too, stakeholders indicated an interest in knowing more about an infrastructure’s support for standards (e.g., metadata, security, authentication, metrics) and how peer institutions had successfully transitioned between systems, including migrating from specific proprietary platforms to open platforms.
- Fiscal security. Some of the stakeholders we talked with expressed concern about the longevity and financial resilience of open infrastructures. While this may include financial documentation, more often, interviewees seemed to want indications of current financial and labor investment and a reliable signal of how likely those investments were to continue into the future.
- Longevity and embeddedness. Related to “fiscal security,” stakeholders repeatedly signaled their desire for evidence that an infrastructure is likely to thrive and how connected it is to the ecosystem as a whole. This could include lists of users (e.g., labs, and institutions), tool and system integrations, and partnerships with other infrastructures. It might also include concrete signals of adoption and investment levels within the ecosystem.
- Loyalty and benevolence. Trust was a significant theme in our interviews. Interviewees sought evidence that the open infrastructure had the user community's best interests in mind. This could include word-of-mouth reputation. This can also be demonstrated through statements, policies, or other documentation that declares the OI is governed by and for the benefit of the community, prioritizing community interests.
- Policy and regulatory compliance. In a few cases, interviewees mentioned the need to comply with specific institutional requirements related to security standards, accessibility, specific business forms, or locations of incorporation. For example, an institution may require all technology providers to be based in a specific country or region or to be operated as non-profit organizations.
- Sense of community and belonging. OI adopters are not just choosing a software, they are joining a community of practice. The success of their implementation often depends on their ability to develop collaborative relationships and make their voice heard within the community of other adopters and maintainers. As such, interviewees frequently referenced a desire to know that open infrastructures were populated by friendly, welcoming people whose values, interests, and goals aligned with their own. They wanted to see vibrancy in different modes of participation. They also sought evidence that open infrastructure communities could both incorporate and resolve inevitable differences in a manner that fostered a communal sense of inclusion.
- Support and technical training. Interviewees expressed interest in knowing whether the OI provided support services, such as technical/development support, as well as training, either directly or through third-party service providers. They also indicated a strong desire to be able to quickly identify trusted service providers that currently work with a given open infrastructure.
- Technical requirements. Open solutions need to be fit for purpose. Adopters seek technical solutions that meet their specific requirements, including the presence of features and functionalities that align with their needs, integrations with other software in their workflow, and compatibility with in-house technical skills or familiarity. A list or description of technical information and capabilities was essential, as it was frequently cited as a first-order decision-making criterion.
- Transparent governance. Interviewees said they looked for evidence of how decisions and choices were made and by whom. Even when interviewees stated they lacked the capacity to participate in governance themselves, they sought to understand how decision-making worked, and documentation or discussion lists/forums were frequently the sources they consulted.
- Usage and adoption. While evidence of usage and adoption was mentioned as part of "Longevity and embeddedness," interviewees frequently cited this as a standalone "want." Being able to understand what organizations are using an OI, and in what contexts, suggested to evaluators how appropriate the OI was for their own needs — and provided guidance on who to talk with to gain further intel.
Now, our project team’s next challenge is to find the sweet spot — figuring out how to present the information decisionmakers most need, without providing too much information for them to digest.
Until the end of 2025, we will be hosting workshops with three open infrastructure stakeholder groups — adopters and end users, open infrastructure communities, and funders and "buyers" — in order to further develop possible ways of raising the visibility of these themes to make it easier for people to evaluate and choose open infrastructures. If you are interested in participating in an upcoming focus group, please send us an e-mail at research@investinopen.org. To stay up-to-date on our research, sign up for our newsletter and follow us on LinkedIn, Mastodon, or Bluesky.