There is no guaranteed formula for getting an ERC Starting Grant. The competition is strong, the proposals are evaluated by scientists, and the scientific judgement cannot be reduced to a checklist.
What can be improved systematically is the competitiveness of the proposal: the importance of the question, the visibility of the ambition, the clarity of the scientific logic, the strength of the PI independence case, the quality of Part I and Part II, the use of figures, the handling of uncertainty and the way the programme is defended under review.
At Arvacore, the objective is not to manufacture a research idea for the applicant. It is to reduce avoidable weaknesses and increase the chance that an excellent scientific idea is understood as intended during evaluation.
How do you "win" an ERC grant?
Applicants often use the phrase how to win an ERC grant because the outcome is competitive and binary: the proposal is funded or it is not. The useful interpretation of that question is not "what is the trick?" It is what makes a proposal sufficiently important, clear, ambitious and credible that reviewers can justify ranking it among the strongest submissions?
Even the ERC currently uses the phrase in an official applicant video titled Writing a winning ERC proposal: What makes evaluators say yes. The existence of that video does not imply a guaranteed formula. It reinforces that proposal quality, scientific judgement and the way the idea is communicated can be improved deliberately.
This guide focuses on those controllable elements.
1. Start with a question worth building a five-year programme around
A Starting Grant can provide up to EUR 1.5 million for up to five years, with additional funding possible for justified eligible costs under the current rules.
That scale of public funding should support more than a collection of technically competent experiments.
The proposal needs a scientific question large enough to justify sustained attention and precise enough to support a sequence of decisions.
A useful first test is:
If we answer this question, what changes in the field?
If the answer is only that a specification improves, the scientific ambition may still need work. If the answer changes interpretation, capability, mechanism or the direction of future research, you may be closer to ERC scale.
Read: How ambitious should an ERC Starting Grant be? ->
2. Create the reason to care before you explain the method
Strong proposals often produce a clear reaction early: this matters.
That reaction should come from the scientific vision, not from adjectives.
The reader should understand:
- the field-level problem;
- the unresolved scientific limitation;
- the central question;
- the proposition or hypothesis;
- the strongest consequence if the programme succeeds.
Only then does the detailed method become interesting.
This is why I treat an ERC proposal as a vision proposal with rigorous implementation, rather than a methods paper with an ambitious introduction.
3. Keep genuine scientific uncertainty
A frontier-research proposal benefits from a central proposition that can fail.
If the preferred result is almost guaranteed, the programme may be technically difficult without being scientifically adventurous enough.
The aim is not to create artificial risk. It is to preserve genuine uncertainty in the scientific answer while controlling avoidable execution risk.
A strong proposal can state:
- what outcome would support the central proposition;
- what would only partially support it;
- what would contradict it;
- what each of those outcomes would teach us.
This is scientific optionality: more than one plausible outcome can move the field forward.
Read: Scientific risk in ERC Starting Grants ->
4. Make Part I / B1 carry the scientific vision
Under the current structure, Part I is limited to five pages. At Step 1, panel members evaluate it together with the CV and Track Record.
The five pages therefore need to establish the scientific case without relying on Part II to rescue the argument.
Part I should make the following visible:
- what the field currently knows;
- what remains unresolved;
- why the unresolved question matters;
- the new scientific idea;
- the main objectives;
- the overall research strategy;
- the strongest scientific consequence.
My rule is simple:
Part I should create the need to read Part II.
Read: ERC Part I / B1 guide ->
5. Use figures only when they clarify the science
A conceptual figure can be exceptionally valuable in a five-page document because it compresses the scientific logic.
It can show:
- current understanding versus the unresolved problem;
- the proposed mechanism;
- how objectives connect;
- the outcome space;
- how the programme changes direction under different results.
The figure needs to earn its page space. Decorative graphics are expensive in a constrained proposal.
Read: How to use figures in an ERC proposal ->
6. Use Part II / B2 to create confidence in execution
Part II is where methodology, work plan, risk assessment and mitigating measures are developed in detail.
It should answer a different question from Part I.
Part I asks:
Is this an important and original scientific idea worth pursuing?
Part II asks:
Can this programme generate rigorous and interpretable evidence within the proposed team, resources and constraints?
The two parts should be complementary rather than repetitive.
Read: ERC Part II / B2 guide ->
7. Make scientific independence unmistakable
The proposal is not only judged as a project. The ERC also evaluates the Principal Investigator in conjunction with the proposed research.
The PI should look ready to establish or lead an independent research programme.
Ask:
- Is the central question clearly mine?
- Does the programme move beyond simply continuing my former supervisor's work?
- Can I explain my scientific contribution rather than only my publication list?
- Does my trajectory make the proposal credible?
- Will the ERC allow me to build a recognisable direction and group?
Remaining in the same institution is not automatically a weakness. Intellectual ownership matters more than geography, but the proposal should make that ownership visible.
Read: ERC CV, Track Record and independence ->
8. Build a research group around the question
The team should follow the scientific programme.
Depending on field, country, salary levels and equipment needs, the right structure will vary. In some programmes, two or three PhD researchers and one or two postdocs may be a reasonable starting model. In others, a smaller or more specialist structure is better.
Experienced postdocs can provide continuity, specialist expertise and reduce the amount of direct training the PI must provide across several doctoral projects.
The important point is to build a PI-led research group, not simply hire people to execute disconnected tasks.
Read: How to build an ERC Starting Grant team ->
9. Use preliminary evidence to establish plausibility, not certainty
ERC guidance states that preliminary results may be included when useful, but they are not mandatory.
That leaves room for scientific judgement.
My view is:
- no preliminary data can be acceptable if the question is compelling and the programme is credible;
- some evidence can help demonstrate access, feasibility of a critical measurement or an unexplained observation;
- too much evidence can make the central question look already answered.
The ideal evidence often says:
We have seen enough to know the question is real, but not enough to know the answer.
Read: ERC preliminary results ->
10. Get external challenge before the final submission
A mature draft should be read by people who are not emotionally attached to it.
Ask them to identify:
- the central question;
- the strongest reason to fund the work;
- the biggest unresolved concern;
- where the logic becomes difficult to follow;
- whether the figures clarify the argument;
- whether the proposed risk is scientific or merely operational;
- whether the PI independence case is convincing.
Language editing is useful. Scientific challenge is more important.
11. Prepare for evaluation, not only submission
Before submission, read the proposal as if you were a panel member with limited time and a large stack of strong applications.
At Step 1, can you understand the scientific reason to continue after Part I?
At Step 2, does the complete proposal withstand specialist scrutiny?
If invited to interview, can the PI defend the scientific assumptions, programme architecture, team and budget without hiding behind prepared slides?
Read: How ERC evaluation works ->
Read: ERC Starting Grant interview guide ->
12. Start early enough to change the idea, not only the prose
The most valuable proposal changes often happen before the text is polished.
A weak objective may need to disappear. A central question may need to be reframed. A figure may reveal that the programme contains two competing narratives. A reviewer may identify that the strongest scientific consequence is hidden on page four.
Leave enough time for conceptual revision.
If every review happens in the final week, most feedback can only change wording.
How Arvacore works to improve proposal competitiveness
No consultancy can guarantee an ERC award. The final result depends on scientific quality, evaluation and competition within the call.
Arvacore works on the parts that can be improved before evaluation:
- scientific concept and ambition review;
- central-question stress testing;
- Part I / B1 architecture;
- Part II / B2 coherence;
- conceptual figures;
- scientific risk and alternative outcomes;
- CV, Track Record and independence narrative;
- reviewer-style red-team feedback;
- interview preparation.
The science remains the PI's. The role of proposal support is to help make the scientific vision, logic and evidence easier to understand and harder to misread.
Explore ERC grant consulting ->
Start with the ERC proposal scorecard ->
If you have not started writing yet
Begin with the ERC Start Here guide and the blank-page method.
Official source
For current rules, eligibility, call documents and evaluation information, use the official ERC Starting Grant page.