Research administrators managing multi-funder portfolios face a recurring headache every grant cycle: no two major funders ask for a data management plan in the same way. A single investigator with an NIH R01, an NSF collaborative award, and a Horizon Europe consortium grant may need three structurally different documents that all attempt to answer the same underlying question — how will research data be generated, described, preserved, and shared? Building a reusable data management plan template that maps cleanly onto each funder’s requirements is now one of the most practical efficiency gains available to a research office.
The stakes have risen. NIH’s 2023 Data Management and Sharing Policy is now actively enforced through award terms and conditions, UKRI’s open access policy has tightened expectations around data underlying publications, and Horizon Europe continues to treat the data management plan as a living deliverable rather than a one-off proposal attachment. Administrators who still draft a fresh plan from scratch for every submission are absorbing avoidable cost. A well-designed crosswalk — and a template built from it — turns a compliance burden into a repeatable process.
The Funder Crosswalk: NIH, NSF, and Horizon Europe Data Management Plan Requirements Compared
The three funders diverge on format, timing, and philosophy, even though all three now anchor their expectations in FAIR (Findable, Accessible, Interoperable, Reusable) data principles in substance if not always in name.
- NIH: The Data Management and Sharing Plan is submitted as a distinct attachment at the time of proposal, is not subject to the page-limit rules that apply to the research strategy, and is expected to address six elements — data type, related tools and software, standards applied, oversight of data sharing, and preservation and access timelines, including where data will be deposited. NIH review does not score the plan competitively but the awarded terms make compliance a condition of funding, and lack of an approved plan can hold up an award.
- NSF: The data management plan is a mandatory two-page supplementary document across all directorates, required since NSF’s foundational 2011 data sharing policy. NSF is comparatively brief on prescribed sections but expects coverage of the types of data produced, standards for metadata, provisions for access and sharing, and policies for reuse and redistribution. Reviewers do weigh the plan as part of the intellectual merit and broader impacts criteria, which makes NSF’s version more consequential to scoring than NIH’s.
- Horizon Europe: The DMP is not typically required at proposal stage for most calls; instead it is a formal deliverable due within the first six months of a funded project and is explicitly framed as a “living document” to be updated at least once more during the project lifecycle, often at mid-term and final reporting. Horizon Europe’s template, aligned with its open science policy, requires explicit narrative on FAIR compliance for each dataset, plus details on cost, responsibilities, and security, including whether data will be open by default or requires a documented exception.
The practical consequence for administrators is that the same investigator’s data description work has to be repackaged three times: NIH wants it compact and attached at submission, NSF wants it capped at two pages and reviewer-facing, and Horizon Europe wants it detailed, iterative, and post-award. A shared template only works if it separates the stable content — data types, standards, repositories, roles — from the funder-specific packaging around it.
Where UKRI and Clinical Trial Plans Diverge Further
Multi-funder portfolios rarely stop at the “big three.” Two further categories complicate the picture for UK-facing and clinical research offices.
A UKRI data management plan follows UKRI’s Common Principles on Data Policy, but implementation is devolved to the individual research councils (MRC, BBSRC, ESRC, and others), each of which has its own template and level of prescriptiveness. This is a different model from Horizon Europe’s single harmonised template, and it means a UKRI-funded co-investigator on a Horizon Europe project may technically owe two structurally distinct plans for the same dataset. UKRI’s broader push on open access — extended in recent policy updates to cover monographs and underlying data alongside journal articles — has raised the profile of the DMP as a compliance artefact rather than an administrative afterthought.
A clinical data management plan is a different instrument entirely, and administrators should not conflate the two. Where a funder DMP addresses data stewardship at the study or grant level, a clinical data management plan operationalises data collection, validation, cleaning, and quality control for a specific clinical trial, typically governed by Good Clinical Practice (GCP) principles and referenced in trial protocols. ICMJE’s data-sharing statement requirement for clinical trial registration adds a further, related but non-identical obligation: a public statement, at registration, of whether and how individual patient data will be shared after publication. A portfolio that includes clinical trials therefore needs both a funder-facing DMP and a trial-level clinical data management plan, cross-referenced but not merged.
Building a Template Structure That Works Across Portfolios
A functional cross-funder template separates content into modular blocks that can be recombined per submission rather than rewritten. A workable structure includes:
- Data inventory: types, formats, and estimated volumes of data to be generated or reused, written once and reused across all funder versions.
- Standards and metadata: discipline-specific metadata schemas and file formats, referencing recognised community standards where they exist.
- Storage and security during the project: active storage, backup, and access-control arrangements, particularly relevant to Horizon Europe’s security section and to clinical trial data governance.
- Preservation and repository: the named repository (disciplinary, institutional, or generalist, such as those indexed by DataCite) and expected retention period.
- Access and reuse conditions: licensing terms, embargo periods, and any restrictions arising from participant consent, commercial sensitivity, or export control.
- Roles and responsibilities: named individuals accountable for each stage, which Horizon Europe expects explicitly and NIH and NSF increasingly expect implicitly through institutional data stewardship policies.
From this modular base, administrators can generate NIH’s compact attachment, NSF’s two-page version, and Horizon Europe’s fuller living document by adjusting emphasis and length rather than starting over. Tools such as DMPonline (maintained by the Digital Curation Centre) and DMPTool already offer funder-specific templates built on broadly this logic, and reviewing existing data management plan examples published through these platforms is a faster route to a working draft than starting from a blank page. The discipline is in maintaining the underlying data inventory as the single source of truth and treating each funder’s version as an export, not an independent document.
What This Means for Research Administrators
For research offices supporting investigators across NIH, NSF, Horizon Europe, and UKRI portfolios simultaneously, the crosswalk approach changes three things in practice. First, pre-award staff can build a standing “data profile” per investigator or dataset at the proposal-development stage, rather than waiting for each funder’s specific form to trigger the work. Second, post-award compliance monitoring becomes more tractable: Horizon Europe’s requirement for plan updates at mid-term and final reporting, and NIH’s enforcement of the terms attached at award, both depend on someone tracking which version is current and when the next revision is due. Third, offices supporting clinical research need to keep the clinical data management plan and the funder DMP as separate but cross-referenced documents, since conflating them risks under-specifying either the trial-level quality controls or the funder-level FAIR compliance narrative.
The administrative overhead of multi-funder compliance is not going away — if anything, the direction of travel among NIH, NSF, UKRI, and Horizon Europe is toward more explicit, more frequently updated, and more publicly scrutinised data plans. Institutions that invest now in a modular, crosswalk-based template will spend less time reconciling funder idiosyncrasies later, and will be better positioned as additional funders and national mandates converge, however unevenly, on the same underlying FAIR data commitments.
