Turning FAIR principles into practice: a starting guide for researchers

by Ioana Maria Cortea — Last updated on June 20, 2025 — Reading time: 12 min

Highlights

» FAIR is a flexible framework, not a rigid standard.
» Data FAIRification transforms unstructured data into reusable, interoperable research assets.
» Metadata, identifiers, and licensing are the foundation of reusable research data.
» FAIR ≠ Open: Accessibility can include restrictions—as long as they’re transparent.
» Trustworthy repositories and data stewardship are essential to making FAIR sustainable.
» A rich ecosystem of tools, standards, and resources are available to support every step of the FAIR data journey.

Building a FAIR research ecosystem

In the era of open science, producing data is no longer the end of the research process. For data to live beyond the lab, serve new purposes, and be truly useful to others, it must be more than just stored. It must be FAIR.

First introduced in 2016, the FAIR Data Principles are now widely adopted across scientific disciplines, and research infrastructures. FAIR stands for Findable, Accessible, Interoperable, and Reusable—a global framework for improving the quality, visibility, and long-term value of research data. These principles aren’t just technical checkboxes. They’re part of a shift in research culture: toward collaboration, transparency, and sustainability. But putting FAIR into practice isn’t always simple. It takes tools, know-how, and time.

This post is intended as a beginner-friendly guide to the world of FAIR data, covering not only its core principles but also the broader ecosystem—from international frameworks to key European and global initiatives that support FAIR-enabling practices. Whether you’re a researcher, data steward, or repository manager, this resource is designed to support and guide you through this dynamic and evolving landscape.

What is data FAIRification—and why does it matter?

The FAIR principles aim to enhance the value of data by ensuring it can be easily located, understood, and reused by both humans and machines. When research data is FAIR, it can:

Help peers understand the research data
Enable collaboration and streamline data sharing
Boost the visibility and citation of your work
Strengthen transparency, reliability, and reproducibility
Protect against data loss and wasted effort

Today, FAIR data is more than a best practice—it’s often a requirement. Many journals, funders, and infrastructure providers now expect researchers to demonstrate the FAIRness of their data as a condition for publication or funding. The FAIR principles provide a powerful framework for improving the way we handle research data. They offer clear guidance for making data—and the metadata and infrastructure that support it—Findable, Accessible, Interoperable, and Reusable (FAIR). But how do we get there from the messy, inconsistent, or siloed data many researchers start with?

That’s where data FAIRification comes in. FAIRification refers to the practical steps involved in transforming unstructured or poorly documented data into FAIR-aligned assets. These steps can include data processing, cleaning, integration, and documentation—all tailored to make data compliant with the FAIR principles. Crucially, FAIRness isn’t all-or-nothing—it can be assessed at multiple levels of detail: from a single dataset or spreadsheet to entire repositories, data lakes, and platforms. The FAIR principles are also discipline-agnostic.

Depending on the nature and quality of your data, FAIRification may involve a wide range of activities. These might include:

Analyzing and assessing data readiness
Designing semantic models and metadata structures
Choosing appropriate data licenses
Developing strategies for data linking and provenance
Deploying tools and repositories to make data accessible

What FAIR is not…

Understanding what FAIR is not is just as important as understanding what it is. According to GO FAIR:

FAIR is not equal to open: Data can be FAIR even if it is not openly accessible, as long as access conditions are clearly stated and metadata is available.
FAIR is not a standard: Rather than being a rigid standard, FAIR is a set of guiding principles meant to be interpreted and implemented in context.
FAIR is not automatic: FAIRification is an intentional and often iterative process, not something that happens by simply uploading data.
FAIR is not only for humans: FAIR supports both human and machine access, which requires metadata to be structured and machine-readable.
FAIR is not static: It evolves alongside technologies and community practices and should be treated as a dynamic goal.

FAIR data management gaps (and solutions)

As the global research ecosystem shifts toward open science, the role of FAIR data becomes more crucial—and more complex. The Horizon 2020 Programme Guidelines on FAIR Data introduced the concept that “data should be as open as possible and as closed as necessary.”

One of the persistent challenges in moving toward a truly FAIR-aligned research culture is the lack of clarity around what “open” really means in practice. Openness is not absolute—it exists on a spectrum. While some datasets can and should be fully open, others may need to be partially restricted or fully protected due to concerns such as privacy. FAIR helps address this tension by promoting transparency over availability. The key is that metadata—the information about the data—should always be openly accessible, even if the dataset itself is not. In this way, FAIR serves as a bridge between the ideals of open science and the real-world responsibilities of data stewardship.

To truly advance FAIR and open science together, several systemic changes are needed:

Clear guidelines for the assessment and selection of data worth preserving and sharing.
Formal recognition of data stewardship as an essential research function.
Investment in sustainable infrastructure, including trusted repositories, community standards, and long-term stewardship models.
Integration of research data into scholarly communication, ensuring researchers are credited and rewarded for sharing high-quality datasets.
Upskilling the ecosystem, from researchers to data stewards to domain-specific data professionals.

Recent findings from the EOSC-Pillar survey highlight additional practical gaps and unmet needs that researchers face in adopting FAIR practices. Respondents emphasized the continued importance of: support for FAIRification workflows, especially for non-expert users; guidance and training on implementing metadata standards and licensing; simple, user-friendly tools for assessing and improving FAIRness; and cross-institutional interoperability to bridge infrastructure and policy gaps between data providers and users. All of these insights underscore a critical point: while FAIR is a powerful framework, its success depends on coordinated, community-driven efforts to close these gaps.

How to make your data FAIR – practical steps for researchers

Making your data FAIR involves planning, documenting, and sharing it in a way that maximizes discoverability, accessibility, and reusability. A wide range of materials and resources are currently available to guide researchers in starting their FAIR data journey. Drawing on established sources such as FAIR-IMPACT, OpenAIRE, DataONE, and howtoFAIR.dk, we have outlined below a step-by-step approach to help you make your research data FAIR. These steps help ensure your data can be easily discovered, accessed, reused, and integrated into other research.

Plan FAIR from the start (FAIR by design)
- Include FAIR practices in your Data Management Plan (DMP).
- Use tools like ARGOS or DMPonline for structured planning.
Use Persistent Identifiers (PIDs)
- Assign DOIs to datasets, ORCID iDs to researchers, and RORs to institutions. PIDs make data citable, traceable, and discoverable.
Use metadata standards and controlled vocabularies
- Apply domain-relevant metadata standards (e.g., DataCite, Dublin Core).
- Use community standards, thesauri, or ontologies where possible. Well-structured metadata is key to making data findable and interoperable.
Use open and interoperable formats
- Choose non-proprietary, widely supported file formats like CSV or XML.
- Avoid obscure or proprietary file types.
License your data clearly
- Apply open, machine-readable licenses (e.g., CC-BY).
- Ensure reuse rights are transparent and legally sound.
Provide rich documentation
- Include README files, and information on methods and workflows. Help others understand and reuse your data effectively.
Link related resources
- Connect your data to publications, software, protocols, and contributor identifiers. Create a network of context and validation.
Choose a trusted repository
- Use (or build) repositories that support FAIR and TRUST principles.
- Prefer platforms with certifications like CoreTrustSeal or those following OpenAIRE guidelines.
Use FAIR assessment tools
- Self-assess your dataset using tools like F-UJI or FAIRshake. Such tools identify gaps in FAIRness and provide actionable guidance for improvement.

Key resources to guide your FAIR journey

Issue	Resource
Writing a data management plan (DMP)	DMPonline Horizon Europe DMP template BELSPO template DCC DMP How-to Guide
Data organization and documentation	CESSDA Data Management Expert Guide DataONE webinars
Selecting persistent identifiers (PIDs)	DCC Persistent Identifiers Briefing Paper Guidelines for PID Managers PIDs. Information Sheet for Data Stewards PIDs. Information Sheet for Researchers
Selecting a metadata standard	FAIRsharing.org Flowchart to find and select a metadata standard RDA Metadata Standards Catalog
Ensuring interoperability	EOSC Interoperability Framework Cross-Domain Interoperability Framework (CDIF) RDA Recommendations & Outputs Catalogue Semantic Interoperability. Information Sheet for Researchers Semantic Interoperability. Information Sheet for Data Stewards
Data licensing and access rights	DCC License Research Data How-to Guide Creative Commons
Data repositories and sharing	zenodo figshare Datavserse OSF
FAIR data policy	FAIRsFAIR FAIR-enabling data policy checklist
FAIR-assessment tools	FAIR AWARE F-UJI FOOPS!
Training and community support	GOFAIR How to FAIR DCC How-to Guides OpenAIRE Guides for researchers FAIR-IMPACT implementation workshops FAIRsFAIR Implementation stories

Major initiatives supporting FAIR data adoption

Leading Europe’s shift toward a more open and data-driven research culture is the European Open Science Cloud (EOSC). Originally developed as a central portal for discovering and accessing research data services across Europe, EOSC has evolved into a federated, collaborative ecosystem. Today, the EOSC EU Node plays a central role in connecting European data infrastructures and making research outputs more accessible and reusable across borders and disciplines. It supports a virtual environment where researchers can find, access, and reuse data and services with ease. A central component of this effort is the EOSC Interoperability Framework (EOSC-IF), which sets out the foundational principles for ensuring that systems, services, and data across the EOSC landscape are interoperable—not only technically, but also legally, organizationally, and semantically. This framework plays a crucial role in enabling seamless collaboration and data reuse.

The vision of a FAIR-enabled research ecosystem has been advanced in recent years through a range of international and cross-disciplinary initiatives, many of which originate or are coordinated through the Research Data Alliance (RDA). As a global, community-driven organization, the RDA has developed a robust portfolio of peer-reviewed recommendations and practical outputs that serve as the foundation for implementing FAIR practices. These include: metadata standards and catalogues, data citation guidelines, repository assessment criteria, and the FAIR Data Maturity Model, among others.

A number of project across Europe and beyond—some completed, others still ongoing—have also significantly contributed to shaping the FAIR data landscape:

FAIR-IMPACT: Delivers frameworks, guidance, and tools that help research communities implement FAIR practices, assess maturity levels, and align with EOSC recommendations.
FAIRsFAIR: Offers essential infrastructure components, including policy frameworks, metadata schemas, and comprehensive training programs.
FIDELIS: Supports repository managers in assessing and improving their alignment with FAIR and TRUST principles.
WorldFAIR: Facilitates international collaboration to implement FAIR practices across a wide range of scientific disciplines, emphasizing global inclusivity.

Collectively, these initiatives have delivered a diverse portfolio of tools, frameworks, and capacity-building resources that empower research communities, institutions, and infrastructure providers to implement FAIR principles within specific contexts and domains.

How to cite this resource

Cortea, I. M. (2025). Turning FAIR principles into practice: a starting guide for researchers. Zenodo. https://doi.org/10.5281/zenodo.15704641

Associated metadata

Title	Turning FAIR principles into practice: a starting guide for researchers
Author(s)	Ioana Maria Cortea
Affiliation	National Institute for Research and Development in Optoelectronics – INOE 2000
Resource type	Blog article
Keywords	FAIR data principles, FAIR data, FAIR implementation, research data management, data curation, data FARification, data management plan, FAIR by design, starting guide
Target audience	researchers, data curators, data stewards, students
License	CC0 1.0 Universal
Persistent Identifier (PID)	DOI: 10.5281/zenodo.15704641
File download	https://doi.org/10.5281/zenodo.15704641

INFRA-ART BLOG