How to Use CORDIS to Research Project Partners (Beyond the Search Box)

Use CORDIS as a partner evidence system, not a directory. Start from projects like yours, identify coordinators and repeat participants, then verify fit by role, topic, geography, and public outputs. The search box finds names, but project pages show working relationships.

Start from comparable projects, not organisation names

Open CORDIS with a narrow research intent. Do not begin with a famous institution or a broad keyword. Begin with a project type you can recognise: the challenge, target group, technology, region, method, or policy field that matches your proposal.

Use search to collect a first set of project pages. Then switch from keyword thinking to pattern reading. A useful project page tells you who coordinated, who participated, what the public summary promised, which countries appeared, and how the consortium described its work. That is more valuable than a generic organisation profile.

For Horizon Europe and older Horizon projects, CORDIS is the best public starting point because it connects project descriptions with participants. Across our 2014-2027 dataset, Horizon Europe accounts for 23278 projects and H2020 for 35389 projects. That scale is useful, but only if you filter hard.

Your first action is simple: save a small cluster of similar projects. Keep only pages where the work looks close enough that a partner could plausibly understand your proposal without a long explanation.

Read roles and patterns before you make a shortlist

A partner name is not enough. Read the role signals around it. Coordinators behave differently from dissemination partners, pilot hosts, research institutes, software builders, data providers, public authorities, and community organisations. CORDIS will not always label these roles cleanly, so infer them from the project summary, work package language when available, deliverables, and participant mix.

Look for three concrete signs:

• Relevant work, shown by project topics, objectives, and public outputs.
• Consortium behaviour, shown by repeated appearances with credible peers.
• Proposal fit, shown by country, sector, stakeholder access, and likely capacity.

Our consortium graph covers 83490 projects, 114731 organisations, and 461953 participation edges from 2014-2027. It shows why role reading matters. Many organisations appear once. A smaller group appears repeatedly, and 1717 organisations appear in five or more projects. That does not mean they are always the right partner. It means they have a public track record you can inspect.

When you shortlist, write down evidence, not impressions. “Participated in a similar project as pilot coordinator” is usable. “Looks strong” is not.

Move from one project page to a partner map

CORDIS becomes more useful when you move sideways. From one relevant project, open each participant. Then inspect other projects where the same organisation appears. You are building a small partner map, not browsing profiles.

Trace connections in two directions. First, follow organisations from project to project. This shows whether they work across themes or stay close to one domain. Second, follow recurring partners around them. This shows their trusted network, which can reveal specialist SMEs, universities, cities, associations, and NGOs that do not rank high in a simple search.

Pay close attention to coordinators. In our dataset, 25915 organisations have coordinated at least once. A past coordinator can bring proposal discipline, but may not want that role again. A former participant may be a better fit if your consortium already has a coordinator and needs delivery strength.

Also check third-country experience when relevant. The data shows 10219 organisations have taken part with a third-country role or connection. If your call involves international cooperation, this can help you avoid cold outreach to partners with no visible experience in cross-border grant work.

For teams that need this mapped faster, StrandsUnited can turn public project data into a focused partner intelligence workspace through our platform work.

Turn CORDIS evidence into an outreach file

The output of partner research should be an outreach file, not a folder of links. Create a short record for every candidate. Include the organisation name, country, project evidence, likely role, contact route, and the reason you think they fit. Add one sentence on risk, such as “strong research record, unclear access to end users” or “good local network, limited EU coordination evidence.”

Then write outreach that proves you did the work. Mention the specific project that led you to them. Name the role you think they could play. State what you already have in the consortium and what is missing. Avoid sending a generic partner request with a full proposal attached. Good partners receive many vague invitations.

CORDIS should also help you decide who not to contact. Skip organisations whose public work is too far from your call, whose role is unclear, or whose projects show a mismatch in scale. A smaller, evidence-backed shortlist will outperform a large spreadsheet.

Before you send messages, compare your shortlist with programme context. For Erasmus+, use the official Erasmus+ site alongside project evidence, because education, youth, sport, and training partnerships often need different role signals than research consortia.

Where this comes from

This guidance comes from our work with EU cooperation teams that need partner research to become decisions, not database browsing. At StrandsUnited, we maintain our grant intelligence engine with a consortium graph built from CORDIS and Erasmus+ open data, and we use it to study organisations, project histories, and collaboration paths.

The same thinking sits behind the platforms we operate, including Impactful and private community knowledge systems: structured evidence, clear source links, and workflows that help people act. If you are preparing a consortium, your next step is to choose a narrow project cluster, build an evidence-backed shortlist, and decide which role each candidate should play before outreach begins. You can also explore how we design and operate these systems on what we do.