Data Management Plan Generator

How to Use the Data Management Plan Generator

1. 1. Pick the template your funder expects. Generic follows the DCC checklist; NSF follows the Research.gov form; Horizon Europe follows the Commission's FAIR template. You can switch at any time — your answers stay.
2. 2. Answer in short, concrete sentences. Name formats, volumes, repositories, licences and people. One line per point; each line becomes a paragraph.
3. 3. Watch the plan check. Each section shows whether it has its required answers, and unanswered questions stay in the draft in brackets. Fields a template does not use are listed so nothing you typed vanishes silently.
4. 4. Copy it where it has to go. Copy for Word or Google Docs, download text or Markdown, or — for NSF — paste each answer into the matching box of the Research.gov form.

Four Templates, One Set of Answers

The DCC checklist from the UK's Digital Curation Centre distils what funders and institutions ask for, which is why so many university templates read like it. Free planning tools such as DMPTool in the US and DMPonline in the UK and Europe host funders' own templates; if yours is there, use it, and paste in the answers drafted here.

What Changed at NSF in 2026

For years an NSF proposal carried a supplementary document of no more than two pages that could describe five things: the types of data and other products, the data and metadata standards, access and sharing policies, policies for re-use and derivatives, and plans for archiving. Policy Notice NSF 26-202, issued in January 2026, replaced that upload with a Data Management and Sharing Plan built in a Research.gov webform, live from 27 April 2026. The form is generated from the proposal's primary directorate and asks the same set of questions for each of up to four data or research product categories.

The notice also tightened sharing: data supporting NSF-funded publications must be shared at the time of publication, and any exception has to be described and justified in the plan. A plan saying no detailed plan is needed is still allowed, but now needs a clear justification. Directorates publish their own guidance, so check yours before you fill in the form.

What a Specific Answer Looks Like

Reviewers read dozens of plans that promise data "will be shared as appropriate." A useful answer names the repository, the licence, the timing and the reason for any restriction. For a study of bedroom heat and sleep in older adults, the sharing section of the generic template could read:

Common Mistakes in Data Management Plans

• "Available on request." That leaves sharing to chance, and NSF now expects data behind a publication to be shared when the paper appears. Name a repository and a date.
• Sharing data your consent form never mentioned. If participants were not told their de-identified data might be shared, you may not be able to share it. Write the plan and the consent form together.
• No README or codebook. A spreadsheet without variable definitions is barely reusable, even when it is open.
• Proprietary formats only. Keep a copy in an open format — CSV rather than a statistics package's own file type — for long-term access.
• Forgetting the cost. Preparing data for deposit takes time, and some repositories charge. Budget it, or explain who covers it.
• Nobody in charge. Name the person responsible for backups, for the deposit, and for keeping the plan up to date.

Frequently Asked Questions

What is a data management plan?

A data management plan (DMP) is a short document describing what data a project will collect or create, how it will be documented, stored and protected during the research, and how it will be shared and preserved afterwards. Most major funders require one with the proposal or early in the grant, and many universities expect one for theses.

What should a data management plan include?

At minimum: the types, formats and volume of data; the documentation and metadata; ethics, consent and intellectual property; storage, backup and security during the project; which data will be kept, where and for how long; how and when data will be shared and any restrictions; and who is responsible and what it will cost. Those are the eight headings of the DCC checklist.

Does NSF still require a two-page data management plan?

No. For proposals submitted from 27 April 2026, NSF collects the Data Management and Sharing Plan through a webform in Research.gov rather than as a two-page PDF. The form asks about access, standards, provenance, archiving, timeline, availability and accountability for each data or research product category. Choose the NSF DMSP template here to draft answers in that order.

When is the Horizon Europe data management plan due?

In Horizon Europe the DMP is a project deliverable rather than part of the proposal, usually due by month 6 and updated during the project as decisions change. The Commission's template is recommended but not mandatory; whatever format you use has to meet the research data management obligations of Article 17 of the Grant Agreement.

What does FAIR data mean?

FAIR stands for Findable, Accessible, Interoperable and Reusable, from the FAIR Guiding Principles published in Scientific Data in 2016. In practice: give data a persistent identifier and rich metadata, deposit it where people and machines can retrieve it, use shared formats and vocabularies, and attach a clear licence and documentation. The Horizon Europe template is organised around these four ideas.

Which repository should I choose for my data?

Prefer a trusted repository in your discipline, such as ICPSR for social science or GenBank for sequence data, because its users are the people most likely to reuse your work. Where none fits, general repositories such as Zenodo, Dryad, Figshare, OSF or Harvard Dataverse assign DOIs and are widely accepted. Check whether your funder or institution names a preferred one.