How Erasmus+ Proposal Evaluation Loses Points

A strong Erasmus+ idea can still receive an ordinary score because the proposal asks evaluators to make favourable assumptions. They will not. Erasmus+ proposal evaluation is an evidence test against the published award criteria, not a reward for a worthwhile cause, a respected consortium or a well-written narrative.

That distinction matters late in the drafting cycle. By submission, partners may have invested months in needs analysis, work package design, budgeting, stakeholder engagement and internal approvals. Once the form is submitted, none of that investment can repair an unstated link between an activity and an outcome, an uncosted delivery risk or a partner role that remains generic.

The useful question is not whether the proposal is good in broad terms. It is whether an evaluator can locate the evidence needed to award the next score band, under the actual form for that action.

Erasmus+ proposal evaluation starts with the right test

Applicants should separate admissibility and eligibility from award criteria. A proposal can be technically admissible and eligible yet score too weakly for funding. Conversely, a polished response cannot compensate for an ineligible partnership, an ineligible applicant or a missing mandatory attachment.

The applicable call documentation governs both the thresholds and the weighting. Do not import a marking scheme from a previous Erasmus+ action merely because the headings look familiar. Cooperation Partnerships, for example, commonly assess relevance, quality of project design and implementation, quality of the partnership and cooperation arrangements, and impact. The maximum score is often 100, with an overall threshold and minimum scores under individual criteria. Other actions use different sub-criteria, score ranges or conditions.

This is not administrative detail. Weighting determines where a weak section is most expensive. A two-point loss in a heavily weighted criterion can be harder to recover than a superficial improvement across several lower-value passages. Before revising, transcribe the current evaluation form into a working scorecard and map every sub-criterion to a page, table or section in the draft.

What loses points under each criterion

Relevance: claims that do not narrow to the call

Relevance is often weakened by a familiar pattern: the proposal describes a real societal or educational problem, cites broad European priorities, then moves straight to planned activities. What is missing is the causal and contextual bridge.

An evaluator needs to see why this target group, in these participating countries, faces the stated barrier; why the selected priorities are central rather than decorative; and why transnational cooperation produces more value than parallel national activity. A list of programme priorities is not alignment. It is a list.

The repair is usually specific rather than expansive. State the evidence base, identify the gap in current provision, and show how each objective answers that gap. Where needs differ between countries, say how the design changes accordingly. A generic needs section may sound inclusive while giving the evaluator no basis to judge fit.

Quality of design and implementation: activity is not method

Work packages tend to lose points when they describe what will happen without showing how the work will be controlled. “Develop a toolkit”, “run pilots” and “hold dissemination events” are activities. They do not establish methodology, quality assurance or credible sequencing.

Evaluators look for dependencies. Can piloting occur before the material has passed a defined review? Who validates the outputs? What selection criteria apply to participants? How will feedback alter the next iteration? Which deliverables are decision points rather than documents produced for their own sake?

A credible work plan makes the chain visible: input, method, output, quality check, revision and use. It also distinguishes outputs from outcomes. A multilingual platform is an output. Sustained use by practitioners, supported by training and measurable adoption data, is closer to an outcome. Confusing the two makes the impact case look inflated.

Risk registers are another common weak point. A table populated with “low”, “medium” and “high” is not risk management unless it identifies an owner, trigger, mitigation and contingency. The real test is whether the consortium has considered what it will do if recruitment falls short, a national policy changes, a key partner cannot deliver or pilot findings contradict the original assumption.

Partnership quality: named organisations are not capacity evidence

A multi-country consortium can look impressive and still score modestly for cooperation arrangements. Evaluators are not scoring the number of logos. They are assessing whether the partnership is necessary, complementary and governable.

Each partner should have a role that connects its competence to a defined task, output and decision right. Saying that a university will support research, an NGO will reach users and a company will provide technical expertise is usually too broad. Show the relevant prior capability, who supplies the personnel, what they will produce, and how the coordinator will verify quality and resolve disagreement.

The same applies to management. Monthly meetings alone are not governance. Define the operational forum, escalation route, financial control, reporting cycle and authority to approve changes. If one partner owns a critical technical component or access to a priority target group, the proposal should acknowledge the dependency and show the fallback.

Impact: aspiration without a route to use

Impact sections often contain the most optimistic language and the least testable evidence. Numbers of downloads, event attendees and social media impressions may be useful dissemination indicators, but they do not prove uptake or change.

A stronger impact pathway identifies who will use each result, what changes in their practice or provision, what prevents adoption, and what mechanism addresses that barrier. For example, a curriculum resource may require translation, accreditation alignment, training for educators and a host organisation willing to maintain it. Without those conditions, claims of European-scale use are speculative.

KPIs should be attributable where possible. “Reach 10,000 people” says little without a baseline, collection method, target group definition and reason why the figure matters. Use a mixture of delivery, quality, uptake and outcome indicators, but do not create a measurement system the consortium cannot operate.

Read the draft as an evaluator would

A disciplined pre-submission review works in a sequence. First, test compliance separately from merit. Second, score each criterion against the published descriptors rather than against the team’s overall impression. Third, require a citation from the proposal for every positive finding and every weakness.

That last discipline changes the discussion. “The impact is clear” becomes “the pathway from pilot results to post-grant adoption is not evidenced after page 34”. “The partnership is strong” becomes “partner four’s contribution is described, but its accountable deliverable and decision authority are absent”. These are fixable findings. General reassurance is not.

Independent readings are valuable because evaluators do not always interpret an ambiguous passage in the same way. If one reader gives a criterion 4 and another gives it 2.5, averaging the figures hides the problem. The disagreement should be examined: what evidence led one reader to infer a strength that the other could not find? That is often the exact sentence, table or missing cross-reference requiring attention.

BidShark applies this logic through independent criterion readings, a red-team reading for unsupported claims and adjudication where scores materially diverge. The automated scoring is tied to the applicable published form; it is not a human panel verdict. Where applicants choose the Expert Q&A package, the subsequent written answers are provided by a person who evaluates EU proposals and has read the report.

Prioritise revisions by score movement, not annoyance

Not every comment deserves equal drafting time. Fixing a typo is cheap, but it rarely moves an award-criterion score. Rebuilding an impact pathway may take partner input, but it can change how the evaluator understands the proposal as a whole.

Start with significant weaknesses under high-weighted criteria and with any sub-criterion close to its minimum threshold. Then address contradictions between sections: a target group defined differently in the needs analysis and work plan, a KPI unsupported by the budget, or an output promised after the activity that depends on it. Finally, improve navigability. Evaluators work to a brief and under time pressure. Signposting evidence is not cosmetic when it prevents a valid strength from being missed.

You only get one chance to submit the version that will be marked. Before the deadline, ask a reader with no investment in the drafting history to identify what the form requires, what the proposal actually proves, and where the gap remains. That is where the next point is usually lost or recovered.