Data Management Plan

Why Plan Data Management?

Planning how to manage research data is a crucial step in any scientific project. A Data Management Plan (DMP) is not simply an administrative requirement; it is above all a strategic tool that supports researchers in anticipating challenges and making informed decisions about their data throughout the research process.

One of the main reasons for planning data management is the need to reduce risks. Data may be lost, corrupted, or rendered unusable if they are not stored securely, backed up regularly, and properly documented. Issues of ownership and user rights can also create difficulties if they are only addressed at the end of a project. Sensitive information, such as personal data or health records, presents additional challenges: confidentiality must be guaranteed, ethical standards respected, and regulations such as the General Data Protection Regulation (GDPR) strictly applied. A DMP allows researchers to think through these issues in advance and to adopt strategies that prevent rather than react to problems.

Planning also offers significant benefits for the scientific process itself. By defining in advance how data will be collected, organised, stored, and described, researchers ensure that the data produced will be of high quality, consistent, and trustworthy. Careful planning guarantees that the information remains usable during the project and that it will not lose value once the project is completed. Thinking ahead about preservation ensures that data will remain accessible and understandable years later, whether for verification, reuse in new studies, or integration into larger datasets.

For PhD students, who are often managing research projects for the first time, a DMP is particularly valuable. It brings structure to the research workflow, clarifying what kinds of data will be generated, how they will be documented, and under what conditions they may be shared. This preparation makes later stages easier, such as writing articles, completing the thesis, or preparing datasets for deposit in a repository. Well-managed data are also more visible: when shared with proper documentation and persistent identifiers, they can be cited, reused, and recognised as a valuable scientific contribution. In this sense, a DMP is not only a planning tool but also a means of enhancing the researcher’s visibility and impact.

Finally, data management planning is closely linked to the principles of Open Science. It ensures that research data can be made “as open as possible, as closed as necessary” and that they respect the FAIR principles of being findable, accessible, interoperable, and reusable. Many funders, including the European Commission in Horizon Europe projects, require researchers to produce and update a DMP as a condition of funding. This obligation reflects the conviction that well-planned data management maximises the scientific, educational, and societal value of research.

✅ In short, a DMP is both a form of protection—against risks of data loss, non-compliance, or ethical breaches—and a source of opportunity, by increasing the quality, visibility, and long-term usefulness of research outputs. It is a living document that should accompany the project from beginning to end, guiding decisions at every stage of the research cycle.

What Is A Data Management Plan (DMP)?

A Data Management Plan, or DMP, is a structured document that describes how research data will be handled during and after a project. It covers the entire lifecycle of data: how they are collected or reused, how they are processed and stored during the project, how they will be secured, shared, or preserved afterwards, and what legal or ethical considerations apply. A good DMP is not only a technical description but also a strategic plan that demonstrates how the researcher ensures the quality, integrity, and long-term value of the data.

For PhD students, the DMP is more than a compliance exercise. It is a practical tool that helps structure the research workflow and anticipate challenges. By reflecting on data management early on, you are better prepared for writing publications, securing sensitive datasets, meeting regulatory obligations, and making your results available in a way that maximises visibility and reuse. In the context of Marie Skłodowska-Curie Actions (MSCA), each doctoral researcher is expected to produce and regularly update a DMP as the project progresses.

💡 In practice, a DMP should be considered a living document. It is first drafted at the beginning of a project, then revised and updated as the research evolves, new data are generated, or circumstances change. It is not intended to be a rigid form, but rather a flexible instrument that guides the researcher, provides transparency to collaborators and funders, and ensures that data remain usable and valuable long after the project ends.

Structure And Content Of A DMP

Although each template has its own structure, a Data Management Plan generally follows a common logic. It describes, step by step, what kinds of data will be produced or reused, how they will be handled, and under what conditions they will be accessible in the long term. Below are the main topics usually included in a DMP, with examples of what to document in each case.

Data Types And Volume

The first step is to describe the nature of the data you will collect or generate. This may include experimental measurements, survey responses, interviews, simulation outputs, images, code, or textual corpora. You should specify their formats—for instance, whether they are stored as CSV, TIFF, FASTA, or MATLAB files—and whether these are open or proprietary formats. Estimating the expected size of your datasets in megabytes, gigabytes, or terabytes is also important to plan storage needs, backups, and repository limits.

Ethical And Legal Issues

This section addresses sensitive aspects such as personal data, health information, or confidential material. You should explain whether informed consent is required, how anonymity or pseudonymisation will be ensured, and what measures will be taken to comply with the GDPR. It is also necessary to mention intellectual property issues, copyright restrictions, or contractual obligations. For example, if you reuse datasets from external sources, you must check the licence conditions and explain how you will respect them.

Documentation And Metadata

To make your data understandable and reusable, you must describe how they will be documented. A DMP should indicate whether you will provide README files, data dictionaries, or codebooks, and whether you will follow disciplinary standards for metadata (e.g. Dublin Core, DataCite, Darwin Core, TEI). The goal is to ensure that someone else—even outside your field—can interpret your datasets and understand how they were generated or processed.

Storage And Security During The Project

Here you should explain where and how the data will be stored while the project is ongoing. This includes describing the physical or virtual infrastructure (university servers, cloud services, encrypted drives), backup strategies (frequency, number of copies, storage locations), and access control mechanisms. If the data include sensitive or personal information, additional measures such as restricted access, encryption, or password-protected folders must be detailed.

Sharing And Preservation

One of the central elements of a DMP is the plan for making data available to others and ensuring their long-term preservation. You should specify what datasets will be shared, under what conditions, and where they will be deposited. Repositories may be thematic (e.g. GenBank, PANGAEA, ICPSR), institutional (e.g. Recherche Data Gouv), or generalist (e.g. Zenodo, Dryad, Figshare). You also need to clarify if any embargo will apply before data are released, what licences will be attached (e.g. CC BY, CC0, Etalab), and how persistent identifiers (such as DOIs) will be assigned. Preservation should go beyond simple storage and guarantee that the data will remain accessible in the future.

Responsibilities And Resources

Finally, a DMP must indicate who is responsible for each aspect of data management within the project. This may be the PhD student, the supervisor, or a data steward, depending on the context. You should also mention the resources required, whether financial (e.g. fees for storage or archiving services), technical (e.g. specialised software or hardware), or human (training, support staff). Clarifying these responsibilities avoids uncertainty and ensures that the plan is realistic.

✅In summary, the structure of a DMP follows the natural lifecycle of research data. It begins with identifying what data exist, then considers how they are documented, stored, and protected, and finally explains how they will be shared, preserved, and credited. A well-written DMP is both a roadmap for the researcher and a guarantee for the community that scientific results are transparent, reproducible, and valuable beyond the project’s lifetime.

Machine actionable DMP or maDMP, and tools

Writing a Data Management Plan does not mean starting from a blank page. A number of online tools have been developed to guide researchers through the process, offering ready-made templates that correspond to funder or institutional requirements. These tools not only save time but also ensure that all the relevant aspects of data management are covered in a structured way.

A recent development in this field is the emergence of machine-actionable Data Management Plans (maDMPs). Unlike traditional text documents, a maDMP is designed to be read not only by humans but also by computers. It uses structured fields and persistent identifiers—such as DOIs for datasets, ORCID IDs for researchers, and ROR IDs for institutions—so that the information can be automatically connected across systems. For example, if a dataset is deposited in a repository with a DOI, the maDMP can link directly to it; if a researcher updates their ORCID profile, the maDMP can retrieve this information without manual duplication. The goal is to embed the DMP into the wider research workflow, making it easier to update, monitor, and integrate with institutional or funder platforms.

Several tools are already available for researchers, each with its own strengths and communities of use. ARGOS, developed by OpenAIRE, provides Horizon Europe–compliant templates and supports collaborative editing. DMP OPIDoR, hosted by the French research infrastructure, offers multilingual templates and is widely used in France. The Data Stewardship Wizard, originally designed for life sciences, allows the creation of highly detailed DMPs with a modular approach. EasyDMP, developed by EUDAT, focuses on simplicity and usability for researchers who are new to data management planning. DMPonline is one of the oldest and most widely used platforms, supporting a large number of institutional templates. DaMaP, a German initiative, provides a central service for universities and research centres in Germany.

For doctoral researchers, these tools are not only a way to comply with funder obligations but also valuable aids in structuring the project. Instead of treating the DMP as a static document, these platforms encourage an iterative approach, with updates at different stages of the research. They also allow supervisors, collaborators, and even data stewards to contribute directly, ensuring that the plan reflects the whole project context.

🤖 The adoption of maDMPs and online planning tools represents a shift in research practice: the DMP becomes part of the digital research ecosystem, connected to repositories, metadata catalogues, and institutional systems. For PhD students, becoming familiar with these tools is an investment in future research careers, as funders and institutions increasingly expect data management planning to be integrated into everyday scientific practice.

DMP OPIDoR as Data Management Plan Tool

DMP Templates Required By Funders

Although many institutions and disciplines propose their own guidelines, the most widely used references in Europe are the Horizon Europe template and the Science Europe core requirements. Both are designed to ensure that researchers address the essential aspects of data management, while leaving some flexibility to adapt to different fields.

The Horizon Europe template is mandatory for all projects funded under the programme, including Marie Skłodowska-Curie Actions. It is structured around a series of questions that guide researchers in describing how data will be generated, documented, stored, shared, and preserved. It emphasises compliance with the FAIR principles—ensuring that data are findable, accessible, interoperable, and reusable—and requires information on standards, licences, repositories, and long-term preservation strategies. For PhD students involved in Horizon Europe projects, this means that preparing and regularly updating a DMP is not optional but a contractual obligation.

The Science Europe guidelines provide a more general framework that is now recognised across many European funders and institutions. They define six core topics that should appear in any DMP, regardless of discipline or project size:

• Data types and volume.
• Ethical and legal issues.
• Documentation and metadata.
• Storage and security during the project.
• Sharing and preservation.
• Responsibilities and resources.

These requirements are deliberately broad, allowing individual funders or institutions to add further detail or adapt the structure to their own needs. Many national funders in Europe have aligned their policies with these guidelines, which facilitates international projects and collaborations.

‼️For PhD students, becoming familiar with both the Horizon Europe template and the Science Europe framework is essential. Even if your project does not fall directly under Horizon Europe, these documents represent the current standard for what is expected in a high-quality DMP. They provide not only a compliance checklist but also a roadmap for responsible and efficient research practice.

Additional Resources :

Self-Assessment Quiz