![](\"https:\/\/world.edu\/wp-content\/uploads\/2026\/09\/research-whisperer-diagram.jpg\")
<\/p>\n<\/p>\n
I write European research proposals for a living, and I also evaluate them. Sitting on the assessor\u2019s side changed how I read my own drafts. When you score proposals against a rubric, at speed, alongside a dozen other proposals, you stop looking for brilliance and start looking for reasons to be reassured \u2014 and reasons to hesitate. Most of what an evaluator writes in the margins is the second kind.<\/p>\n
To see how consistent those hesitations are, I read and coded the criticism in 80 real European Union evaluation reports. The single most useful thing I learned is this: evaluators almost never tell you that you are wrong<\/em>. They tell you that you have not said enough<\/em>. The words that recur are not \u201cincorrect\u201d or \u201cfalse\u201d \u2014 they are \u201cnot sufficiently\u2026\u201d, \u201cnot adequately\u2026\u201d, \u201cunclear\u201d, \u201cgeneric\u201d. That is good news, because under-specification is fixable in an afternoon, once you know where to look.<\/p>\n Here are the five objections I saw again and again, and what to do about each one:<\/p>\n 1. \u201cNot sufficiently detailed.\u201d<\/strong> This is the most common comment by a wide margin, and it lands hardest on methodology, feasibility and risk. A sentence like \u201cwe will use established statistical methods to analyse the data\u201d invites it, because it asserts without specifying. The fix is to trade assertions for specifics: name the method, the sample, the threshold, the person responsible and the point in the timeline. \u201cEstablished statistical methods\u201d becomes \u201cmixed-effects models, powered for a 0.3 effect at n=120, run by WP3 in months 7\u201310.\u201d Same claim, no gap to flag.<\/p>\n 2. \u201cGeneric.\u201d<\/strong> Evaluators write this when a passage could be lifted into any proposal in the call without changing a word. This criticism breeds in impact, dissemination and management sections. The test is quick and brutal: could a competitor submit your paragraph unchanged? If yes, it is generic. The fix is to make every sentence specific to this<\/em> consortium, this<\/em> call topic and this<\/em> problem \u2014 the partner who will actually run the workshop, the named body that will actually use the result.<\/p>\n 3. The impact pathway that isn\u2019t there.<\/strong> Proposals routinely assert impact (\u201cthis will transform patient care\u201d) without routing it through specifics. Evaluators are trained to look for the chain, so a claim with no pathway reads as a wish. Draw the chain explicitly: output \u2192 outcome \u2192 impact, and name who takes each step up. \u201cA validated assay (output) adopted by three reference labs (outcome), shortening diagnosis by two weeks (impact), with uptake driven by our clinical partner.\u201d The claim is the same; now it is credible.<\/p>\n 4. Hand-wavy implementation.<\/strong> Work plans attract \u201cnot adequately justified\u201d when effort, dependencies and risks are vague. Person-months appear without rationale; risks are listed as a formality, mitigated by \u201cwe will monitor closely.\u201d Evaluators read the work plan to check that you have actually thought it through. A real risk register \u2014 likelihood, impact, a specific mitigation and an owner \u2014 plus a short justification for the effort on each task will answer most of this before it is asked.<\/p>\n 5. Ambition the evidence doesn\u2019t support.<\/strong> \u201cAmbitious, but\u2026\u201d is a compliment with a knife in it. It means the reviewer wants to believe you and cannot. Bold objectives with no preliminary data, no track record and no feasibility argument read as over-claiming. You do not have to be less ambitious; you have to earn it. Pair every bold claim with the reason it is achievable: your pilot result, your prior grant, the equipment already in place.<\/p>\n Notice what these have in common. None is about being clever. They are about being specific<\/strong><\/em>, about closing the gap between what you know is true and what you have actually put on the page. Reviewers cannot give you credit for what is in your head.<\/p>\n A self-audit you can run the day before submission<\/strong><\/p>\n Read your draft with one job: hunt your own vague words.<\/p>\n \u201cInnovative\u201d, \u201ccutting-edge\u201d, \u201cstate of the art\u201d, \u201cworld-class\u201d, \u201cwill ensure\u201d, \u201crobust\u201d, \u201cseamless\u201d. Highlight every one. Those adjectives are precisely where an evaluator writes \u201cgeneric\u201d or \u201cnot substantiated\u201d, because they claim a quality instead of demonstrating it.<\/p>\n Then, for each highlight, either delete it or replace it with the specific thing that makes it true. If \u201cinnovative\u201d can\u2019t be cashed out into what exactly is new and why it matters, it was decoration, and decoration is where scores leak. You cannot predict your score, and you should distrust anyone or any tool that says they can.<\/p>\n Panels are human and the competition is hidden. But the objections above are not hidden. They are the same handful, cycle after cycle, and every one of them is answerable while you still have time. The most valuable hour before you submit is the one you spend reading your own proposal the way an evaluator will: quickly, against a rubric, looking for the gap. Find it first, and close it.<\/p>\n