Five Signs of Weak Impact in EU Proposals
A proposal can describe a credible technology, a capable consortium and a well-managed work plan, then still lose the competition on Impact. The five signs of weak impact are rarely dramatic errors. More often, they are small breaks between a claimed benefit, the route to achieving it and the evidence offered to support that route. Evaluators find those breaks quickly because they are reading against a defined criterion, not against the team’s intentions.
For Horizon Europe, Impact is commonly assessed alongside Excellence and Quality and Efficiency of Implementation, with thresholds and weighting set out in the applicable evaluation form. Digital Europe and Erasmus+ use their own forms, sub-criteria and score allocations. Do not transfer assumptions from one programme or action type to another. Admissibility and eligibility are separate gates; a proposal can clear both and still receive an award-criterion score that leaves it below the funding line.
The practical question is not whether the proposal contains an impact section. It is whether an evaluator can trace a defensible chain from the project’s outputs to uptake, outcomes and longer-term effects, without filling in missing reasoning themselves.
1. The proposal restates the call, rather than defining a contribution
The first sign of weak impact is a familiar one: the narrative repeats the expected outcomes and destinations from the call text, then treats that repetition as a contribution. “The project will support Europe’s digital transition” may be accurate at a high level. It is not yet an impact case.
An evaluator needs to see what will change, for whom, by what mechanism, and to what extent the project itself contributes. That means distinguishing between an output under consortium control, such as a demonstrator, dataset, training package or policy recommendation, and the outcome that depends on others using it.
What the evaluator is looking for
A convincing account makes the project’s contribution specific. If a call expects improved public-service capacity, identify the service owner, the operational decision that will change, the implementation conditions and the plausible scale. If the expected outcome is market uptake, identify the buyer or adopter, the route to procurement or commercialisation, and the barriers the project can actually reduce.
The repair is not to make larger claims. It is to narrow them until they can be evidenced. Replace generic alignment language with a contribution statement that names the beneficiary group, the baseline, the project output and the expected change. Then test whether the work packages, partner roles and budget provide the means to deliver it.
2. The pathway has a gap where uptake should be
Many impact sections move directly from a completed pilot to Europe-wide benefit. That is not a pathway. It is an unsupported jump.
A pilot can demonstrate feasibility under defined conditions. It does not, by itself, establish that an authority will adopt a method, that a company will integrate a result into a product, or that a standardisation body will take up a recommendation. Each of those steps has an owner, a decision process, a timetable and usually a constraint outside the project.
This matters particularly where the proposal relies on stakeholders who are not beneficiaries or named partners. A letter of support may show interest, but it does not automatically demonstrate commitment to deployment. Equally, a strong exploitation plan cannot compensate for a weak route to validation if the claimed impact depends on performance evidence not produced until the final months.
Map the pathway in sequence: output, validation, uptake decision, implementation, measurable outcome. For every transition, ask who acts, what they need to decide, and what evidence or resource enables the next step. Where the answer is “stakeholders will be engaged”, the proposal has not yet done the work.
3. KPIs measure activity, not change
Weak impact plans are often full of numbers. The problem is that the numbers count dissemination activity rather than the effect that dissemination is meant to produce.
Website visits, event attendance, social media impressions and the number of policy briefs distributed may be useful management indicators. They do not demonstrate adoption, behavioural change, improved capability or economic, social or environmental benefit. An evaluator will recognise the difference, especially where a proposal claims substantial impact while its KPIs stop at communication outputs.
Five signs of weak impact: the KPI test
For each headline claim, identify one indicator that tests whether the relevant change occurred. A training project might measure not only participants trained but competence gained, subsequent application and the institutional conditions for continued delivery. A technology project might track not only prototypes built but performance against a user requirement, validated adopter interest and the progress needed for deployment.
Targets need a baseline, a unit, a data source, a measurement point and a responsible partner. They also need to be proportionate. A precise-looking target without an explanation of how it was derived can be more damaging than a cautious one, because it invites the evaluator to question the entire impact estimate.
There is a trade-off here. Not every long-term effect can be measured during a three-year project. Where that is the case, state the limitation and use credible leading indicators. Do not present a future societal benefit as if it were a project-period deliverable.
4. Exploitation, dissemination and communication are treated as one task
These functions overlap, but they are not interchangeable. A consortium that says it will exploit results through conferences, a project website and newsletters has usually described dissemination and communication. It has not described exploitation.
Exploitation concerns the route by which results are used. Depending on the action, that may involve product development, licensing, open-source maintenance, integration into public systems, service delivery, curriculum adoption, standardisation, replication or policy implementation. The appropriate route depends on the result and its users.
Evaluators will look for ownership, access conditions, intellectual property arrangements where relevant, decision rights and a credible post-project custodian. In a multi-partner consortium, vague collective language is a warning sign. If a result is valuable, someone should be accountable for taking it forward, and the proposal should show that the consortium agreement and exploitation approach will not obstruct that work.
Do not force a commercialisation narrative onto results intended for public or academic uptake. That is not more ambitious; it is often less credible. Instead, show the actual adoption route and the organisations able to carry it beyond the grant period.
5. Risks to impact are absent, generic or outside the work plan
A proposal may acknowledge technical, ethical, regulatory or market risks in a register, yet leave the impact section untouched by them. Evaluators notice when a claimed pathway assumes conditions that the risk analysis says may not hold.
For example, adoption may depend on interoperability, access to a data source, regulatory acceptance, procurement timing, teacher capacity or a local authority’s ability to fund continuation. These are not peripheral matters if they determine whether a result is used.
A useful impact risk is specific enough to change the plan. It names the dependency, assesses its likelihood and consequence, identifies an owner and describes a mitigation that is funded or built into the work packages. “Maintain stakeholder engagement” is not a mitigation. A scheduled co-design process, a fallback validation environment, an alternative adopter segment or an implementation agreement may be.
This is also where internal optimism is most costly. Teams know why a dependency is likely to work out. The proposal must show the evaluator what happens if it does not.
Read Impact as an evaluator will
Before submission, take the Impact narrative away from its authors and test it against the published form for the precise call and action type. Score each sub-criterion using the form’s descriptors and permitted score increments. Then require the reviewer to identify the exact passage supporting every positive judgement.
That last discipline matters. A statement that cannot be tied to text may be knowledge held by the consortium, not evidence available to the panel. Where independent readings differ materially, do not average the disagreement away. Treat it as a locating signal: one evaluator found a pathway convincing, another could not find the proof.
BidShark’s automated assessment is designed around that distinction. It applies the versioned marking scheme for the applicable call type, produces independent criterion readings and shows the proposal passages behind its findings. It is not a human panel or a prediction of funding. Where the stakes justify it, the separate expert Q&A package is the point at which a practising EU evaluator reviews the report and answers written questions.
A late-stage impact review will not create adopter commitment, policy access or a viable deployment model from nothing. It can, however, expose the unsupported claims while the consortium can still assign an owner, adjust a target, fund a mitigation and state the case the evaluator needs to assess.