Running a Mock Evaluation of Your Horizon Europe Proposal: Reviewers, Scoring and Timing

The EU Funding & Tenders Portal itself recommends it: use the standard evaluation forms published in the Portal reference documents, and ideally ask a disinterested colleague to carry out a mock evaluation before you submit. Most consortia do something they call a review. Far fewer run an evaluation: an independent assessment against the criteria that will actually be applied, producing judgements the team has to act on.

A mock evaluation is not an official European Commission evaluation, and its scores predict nothing. Its value is that it exposes, while there is still time, what an expert reading only the submitted text is likely to see. Proofreading and compliance checks are separate tasks; for the content checks themselves, see our proposal readiness checks.

If your proposal has already been evaluated and rejected, see our guide to deciding whether to revise, rebuild or resubmit.

When to schedule it

Run the mock evaluation when a complete draft exists, including the impact pathway, the work plan tables and an approximate budget, but early enough to change substance. In practice that usually means three to four weeks before the deadline. This is practical judgement, not a rule.

Too early, reviewers assess an outline and their comments describe gaps the team already knows about. Too late, the only feasible corrections are cosmetic, and structural findings, such as a missing capability or an implausible effort profile, cannot be acted on. Leave at least one full revision cycle after the review, plus time for partners to approve the changes.

Selecting competent and independent reviewers

The point is distance from the text. A reviewer who helped write the proposal reads it through the authors’ intentions, which is exactly what real experts cannot do.

  • Two or three reviewers are a practical recommendation, not a Horizon Europe requirement. The number loosely mirrors the official process, in which proposals are normally assessed by several experts individually before a consensus discussion. One reviewer gives one perspective; disagreement between reviewers is often the most useful output.
  • Mix the profiles. One reviewer with domain expertise for Excellence, one with experience of exploitation or industrial uptake for Impact, and one who has managed EU projects for Implementation.
  • Independent of the writing team. Colleagues from another department, a partner organisation’s staff who were not involved, or an external reviewer.
  • Mind confidentiality. Use confidentiality terms, and do not discuss the proposal with anyone who might act as an expert evaluator for the same call: the Portal warns this can create a conflict of interest with consequences for both sides.

An external consultant running a mock evaluation is not acting as an official evaluator, whatever their background. Treat their comments as informed judgement, not as an indication of the outcome.

Preparing the review pack and the criteria

Reviewers should receive what real experts receive, and nothing more:

  • Part B as it will be submitted, within the page limit; text beyond the limit is not visible to real evaluators, so it should not be visible to mock reviewers either;
  • the relevant Part A information, such as participant descriptions, researchers and the budget summary;
  • the full topic text, including expected outcomes, scope and any specific conditions;
  • the evaluation form and criteria that apply to this call.

That last item matters. The standard Research and Innovation Action and Innovation Action form scores Excellence, Impact and Implementation from 0 to 5, with half-marks allowed, an individual threshold of 3 and an overall threshold of 10. Coordination and Support Actions use different aspects under the same headings; the first stage of a two-stage call assesses only Excellence and Impact, with a threshold of 4 each; and ERC, EIC, MSCA, partnership calls and individual topics can apply different criteria, thresholds or weightings. Give reviewers the rules of the actual call, not a generic Horizon Europe model.

Brief reviewers the way the Commission briefs its experts: assess the proposal as submitted, not its potential if changes were made, and reflect shortcomings in the judgement rather than rewriting the text.

Assessing Excellence, Impact and Implementation

Ask reviewers to work through each criterion’s official aspects and give a short written justification before any score.

Useful prompts:

Criterion

Questions for reviewers

Excellence

Are the objectives clear, measurable and pertinent to the topic? Is the advance beyond the state of the art shown, not asserted? Is the methodology sound, including open science, and the gender dimension where relevant?

Impact

Is there a credible pathway from results to the expected outcomes in the topic text? Are target groups and barriers to uptake specific? Are dissemination, exploitation and communication measures suited to those pathways?

Implementation

Does the work plan deliver the objectives? Do effort and resources match the tasks? Are critical risks realistic, with mitigation? Does each participant have a valid role and adequate resources?

If reviewers score at all, use the official scale and its definitions: “a number of shortcomings” for a 3, “a small number” for a 4. Treat the scores as a way to force a judgement, not as a forecast.

Running the review discussion

Collect individual assessments first, without discussion, so that reviewers do not anchor on each other. Then hold a consensus discussion, moderated by someone who did not write the proposal.

  • Agreement between reviewers on a weakness is a strong signal; treat it as a priority.
  • Disagreement usually means the text allows two readings. Ask what each reviewer understood. The fix is often clarity, not new content.
  • A single dissenting expert may still be right, particularly on domain questions. Record the point rather than outvoting it.
  • Authors defend, reviewers explain. Writers may ask clarifying questions, but the session is not a rebuttal. A real evaluator will not hear the explanation either.

Turning comments into owned corrective actions

Convert every substantive comment into a row with a named owner. The example below is illustrative, for a fictional proposal.

Criterion

Identified weakness

Evidence in the draft

Proposed correction

Responsible

Internal deadline

Excellence

Advance over the state of the art asserted, not shown

Section 1.1 cites no benchmark

Add comparison with two named current approaches and target values

WP2 lead

Day 5

Impact

Target groups too broad

“Industry and society” in 2.1

Name the user segments and the decision-makers who adopt

Exploitation lead

Day 7

Impact

No barrier to uptake identified

2.1 has no barriers paragraph

Describe regulatory and user-behaviour barriers and mitigation

Coordinator

Day 7

Implementation

Effort in WP4 does not match tasks

6 person-months for three pilot sites

Recalculate effort with pilot partners; adjust budget

WP4 lead

Day 10

Implementation

Single point of failure

One named data engineer

Name a second team member in Part A and WP3

Partner A

Day 10

Prioritising changes when the deadline is close

With limited time, rank corrections by consequence, not by the number of comments:

  1. Threshold risks: anything a reviewer believes could keep a criterion below the applicable threshold, or that touches a topic condition.
  2. Structural issues that can still be fixed: effort, resources, a role or a risk measure, where the change is feasible before partners must approve the final version.
  3. Clarity and consistency: contradictions between sections, tables and the budget, which are cheap to fix and visible to evaluators.
  4. Style: last, and only if time remains.

If a structural finding cannot be fixed in time, reduce the claim it undermines rather than leaving a promise the work plan does not support.

When an external review is worth considering

An external review is most useful when nobody inside the consortium can read the proposal at a distance, when partners disagree about whether the proposal is ready, when the coordinator is new to the programme, or when the call uses criteria the team has not worked with before. It should deliver written assessments per criterion, a consolidated weakness list and a prioritised correction plan. It cannot predict a score or guarantee funding.

For organisations considering external support with proposal preparation and review, see Nexuswelt’s proposal development and support service. Any specific review scope and fee should be agreed in advance.

Sources

Leave A Comment