ERC panel choice is one of the highest-leverage decisions in the whole proposal, and one of the most poorly optimised by applicants. Pick the wrong panel and you are reviewed by experts who do not recognise the state of the art in your field — often the fastest route to a "B" or "C" score. This guide covers how to read panel scope descriptions, how to handle proposals that span panels, and when to use the secondary panel request.
Read the panel descriptor, not just the title
Each ERC panel has a published descriptor listing sub-topics and exemplar keywords. Before choosing, read the descriptor for your top two candidates side by side. The panel title can mislead — for example, "LS8 Environmental Biology, Ecology, and Evolution" covers microbial ecology and biogeochemistry, which a marine biologist might assume lives under "LS8" but actually maps better to "PE10 Earth System Science" depending on the specific question.
The proxy question: who will review this?
The most useful heuristic: list 5 researchers whose work yours is in dialogue with. In which panels would those researchers most plausibly sit? If 4 of 5 sit in one panel, that is your primary. If they are split, you have a panel-crossing proposal and need to pick the panel with the stronger framing for your specific angle.
When to use a secondary panel request
For interdisciplinary proposals, you can suggest a secondary panel at submission. Use this. Panel chairs use secondary suggestions when assigning external reviewers — it is a real signal, not a formality. A well-chosen secondary panel gets your proposal in front of at least one reviewer with the missing domain expertise, which often means the difference between "interesting idea, hard to judge" and "excellent execution in a specific field."
Common panel mistakes by domain
- •Computational biology: defaulting to LS panels when PE6 (Computer Science) has fewer applications and better reviewers for method-heavy work.
- •Biomedical engineering: picking PE8 when LS7 (Diagnostic Tools, Therapies, Public Health) is often the better fit for translational proposals.
- •Social science + data methods: under-using SH4 (Human Mind) or SH2 depending on whether the contribution is theoretical or empirical.
- •Neuroscience: LS5 covers neurobiology; SH4 covers cognition and behaviour — the exact framing of your research question determines which.
Check the success rate per panel in the most recent ERC statistics report. Some panels are structurally more competitive than others, which can inform strategy for borderline-fit proposals.