The threshold that decides most Pathfinder proposals
Excellence must reach 4.0 out of 5. On the official descriptor ladder, 4 means “addresses the criterion very well, but a small number of shortcomings are present”. A 3 — “addresses the criterion well, but a number of shortcomings are present” — is a good score in ordinary language and a rejection here. Half the ranking weight sits on the criterion with the least room.
Impact adds a second trap: its threshold is 3.5, the only half-mark threshold in any programme we review. The difference between “a number of shortcomings” and “a small number” is the difference between rejected and ranked. And because every criterion must clear its own threshold, a superb Excellence score cannot rescue a 2.5 in Implementation — the section applicants habitually write last.
Pathfinder is not an early Accelerator
It is the opposite end of the pipeline, and reviews built on downstream habits do real damage.
No traction, no business model, no TRL gate
Pathfinder funds research towards a proof of principle at TRL 1–4. There is no TRL entry requirement and none needs to be stated. Asking for customers, revenue, letters of intent or TRL 5–6 evidence would push you to damage a correctly scoped proposal, so those tests are blocked from reaching a Pathfinder review.
High risk is required, not penalised
The scheme exists to fund research that might not work. The review never scores ambition down as a defect. What it does test is unmanaged risk — ambition without interim decision gates, without fallbacks, or without a stated meaning for a negative result.
Impact means the technology, not your deliverables
Section 2 asks about the transformative effect of the technology your project makes conceivable — not about what the project itself produces. This is the exact inverse of Accelerator drafting, and it is one of the most common ways a strong scientific proposal loses Impact.
If you are in the wrong scheme, we say so
A validated result that needs maturing belongs in EIC Transition; a product with customers and a financing gap belongs in the Accelerator. That is not an eligibility failure — it is an Excellence failure, because the breakthrough is behind you rather than ahead. Where the evidence points that way, the report says it early and plainly.
Open and Challenges, in one review
Same criteria, same thresholds, same weights. Four things differ, and the review switches all of them when you pick your variant.
Page limit — 22 or 30
Part B sections 1–3 are capped at 22 A4 pages for Open and 30 for Challenges. Content past the cap is blanked out before evaluators see it, so an overrun does not look untidy — it silently deletes an argument you believe you made. The review names the specific content that would be lost.
Consortium rule
Open needs at least three independent legal entities from three different countries. Challenges permits single applicants, provided they are not mid-caps or larger companies. In both, affiliated entities do not count towards the minimum — a consortium built from one organisation and two of its affiliates does not satisfy the rule.
The portfolio work package
Challenges requires a portfolio activities work package of at least 10 person-months, participating in a roadmap steered by an EIC Programme Manager. Open has none, and the review raises no portfolio findings against an Open proposal.
The portfolio step
Under Challenges, an evaluation committee composes a balanced portfolio from proposals that clear the thresholds, so a high score does not guarantee selection. Portfolio fit is arguable in your text, and most applicants leave it silent. We report it as an arguable strength or an unaddressed dimension — never as a probability.
What the review actually checks
Drawn from Pathfinder’s own criteria and sub-aspects. Each produces a located, evidenced finding with a concrete correction.
- Whether the long-term vision names a radically new technology rather than an application area, a problem or a market — and whether the work plan ever references it again.
- Whether the science-towards-technology breakthrough gives a mechanism, or only promises an outcome.
- Whether the state of the art names competing approaches and the groups pursuing them, cited and dated — or reviews only your own prior work.
- Whether the critical scientific uncertainty is named, and whether it is scheduled early enough that a negative result can still redirect the project.
- Whether the proof of principle is defined as a measurable end state: what is demonstrated, under what conditions, measured how, against what threshold.
- Whether each claimed discipline survives the deletion test — remove it, and does the breakthrough become unreachable? — and whether the consortium and the effort table actually contain it.
- Whether impact is claimed for the envisioned technology across economy, environment and society, with a mechanism rather than adjectives.
- Whether milestones are decision points with a threshold and a stated consequence on both branches, or dated activities.
- Whether the risk table carries real scientific risks with owners and triggers, or generic entries about partner withdrawal.
- Whether the lump-sum workbook, the work-package table and the summary of staff effort reconcile.
No paired corpus of Pathfinder proposals and their official Evaluation Summary Reports is publicly available. Scores here are calibrated against the official descriptor ladder, the native thresholds and conservative evidence rules. They are a simulation, not a predictor of your funding line — and Pathfinder Open funded roughly one proposal in fifty in recent rounds, so the report never lets a threshold pass read as good news on its own.
What you upload, and what comes back
What to upload
Required: your EIC Pathfinder Part B (PDF, DOCX, TXT or MD).
- Tell us the variant: Pathfinder Open or Pathfinder Challenges. It sets the page limit, the consortium rule and whether portfolio checks apply.
- Optional: the challenge guide for your topic, or supporting material you want cross-checked.
What comes back
A free score preview by email within 5–8 minutes, with your two highest-priority fixes.
- The complete report adds the simulated criterion scores against each threshold, the ranked blockers, and the section-by-section register.
- Every finding carries its location, the official aspect it damages (1.1–1.4, 2.1–2.3, 3.1–3.2), a correction, a model-answer pattern and a re-review test.
Start with the free preview
See your simulated scores against the 4.0 / 3.5 / 3.0 thresholds and the two corrections that matter most, before deciding whether the full report is worth €150. If it gives you nothing worth acting on, email us within 14 days and we refund it in full.
Questions
What are the Pathfinder thresholds and weights again?
Excellence 4.0/5, Impact 3.5/5 and Quality and efficiency of the implementation 3.0/5, scored in half-point steps. Each must be met independently — there is no compensating total. The weights of 50%, 30% and 20% rank proposals that already clear every threshold; they never decide whether a proposal qualifies, and no weighted figure is a pass mark.
Does it work for Pathfinder Challenges?
Yes. You pick the variant at upload and the review switches the page limit to 30 pages, applies the Challenges consortium rule, requires the portfolio activities work package with its 10 person-month floor, and tests whether you address the named challenge’s expected outcomes and argue your portfolio fit.
I have not stated a TRL. Is that a problem?
No. Pathfinder has no TRL gate and none needs to be stated. The review will not invent a minimum or a target. It does flag two things: a TRL claim above 4, which suggests a Transition-stage result submitted to the wrong scheme, and an undefined proof-of-principle end point, which is the real Pathfinder equivalent of a missing target.
Will it tell me my proposal is too risky?
No. High risk is a required characteristic of the scheme and is never scored down as a defect. The review tests whether the risk is managed: are there interim decision gates, is there a fallback at each, and is it stated what a negative result would establish?
What happens to my proposal?
Files are transmitted securely for transient processing in the EU. They are not retained and are not used to train models. The generated report may be stored for up to seven days so you can open it from your private email link.
Ready when you are
No account. No subscription. The preview costs nothing.
Review my EIC Pathfinder proposal