Most Horizon Europe projects treat the Data Management Plan (DMP) as a checkbox. Something to submit early and forget.

If you are wondering what we are talking about, the DMP is a document that explains how your project will handle data during and after its lifecycle. It covers how data is collected, stored, shared, and protected.

In Horizon Europe, you usually submit the first version within the first six months of the project as a deliverable. After that, you are expected to update it as the project evolves.

However, beneficiaries often forget about this document until the next deliverable report is due and then find themselves scrambling through the generated data to draft a last-minute update.

This approach creates problems later. In fact, a weak DMP at the start rarely affects your evaluation score. But during implementation, it can slow you down, create confusion, and introduce compliance risks.

Based on what we see across the numerous projects we have supported over the years, these are the 5 most common mistakes to avoid and how to fix them:

1. Treating the DMP as a one-off document
   Many teams submit the DMP and never look at it again. The DMP is meant to be a living document. Your data evolves and so your plan should too.
   How to fix it:
   Plan updates every 3-4 months, not only before deliverable submission. Treat the DMP as part of your project management, not just a one-off task.

2. Using generic content
   Copy-paste text is easy to spot. Reviewers and project officers notice immediately (as they are the ones providing the templates).
   How to fix it:
   Tie your DMP directly to:
   – project work packages
   – specific datasets
   – actual methodologies
   Ask yourself a simple question: Could this DMP fit any project? If so, it’s not good enough.

3. Not identifying datasets clearly

Vague statements like “data will be generated” do not help. You need to define what kind of data you expect, in which format, and roughly how much.
How to fix it:
For each dataset, specify:
– type (i.e. experimental data, survey results, simulation outputs)
– format (i.e. CSV, images, code, etc.)
– estimated volume
If you can’t define it clearly, it will give the impression that you’ll not be able to manage it either.

4. Mentioning FAIR principles without applying them

Many DMPs refer to FAIR principles (Findable, Accessible, Interoperable, Reusable) but stop there, instead reviewers want to see practical implementation choices.
How to fix it:
Be concrete by answering these questions:
– Findable: Will datasets have DOIs or persistent identifiers?
– Accessible: Where will they be stored? Under what conditions?
– Interoperable: Are you using standard formats?
– Reusable: Is metadata sufficient for others to understand the data?
Expectations are clear: explain how your data will actually meet FAIR principles. For reference to the EC policy, see the Horizon Europe “Open Science FAQ”.

5. Ignoring GDPR and sensitive data issues

If your project handles personal or sensitive data, this needs attention from the start. It should not be left for later. This is particularly important in projects with user data or health-related information.
How to fix it:
In your DMP make sure to include:
– legal basis for data processing
– data protection safeguards
– anonymization or pseudonymization application
If personal data is involved, this cannot be an afterthought. For reference to the GDPR Regulation, click here.

What matters most

A strong DMP does not need to be long or complex. It needs to be clear, realistic, and specific to the project.

The Funding Institution looks for consistency, awareness of risks, and a practical approach. That is often more valuable than polished but generic language.

A good DMP supports the whole project. It helps with compliance, coordination, and later use of the data.
If you want support with your DMP or other Horizon Europe project needs, get in touch at [email protected]