The uncomfortable truth about public funding: most rejections are not verdicts on the business. They are verdicts on the file. The same handful of process mistakes accounts for a remarkable share of failed applications, which is good news, because process mistakes are fixable.
These are the seven we see most.
1. Starting the project before applying
The near-universal rule across Luxembourg and EU instruments: the application comes before the work. Signed contracts, ordered equipment, and started development can disqualify the very costs you wanted funded, because the aid must have an incentive effect. Booking the consultant the week before applying for the consulting subsidy is the textbook self-inflicted rejection.
Fix: map the funding landscape before committing money. The application date is a strategic date.
2. Applying to the wrong instrument
A deep-tech research project sent to an SME digitalisation scheme, a commerce business pitched at an innovation programme: mismatched applications do not get charitably rerouted, they get declined. Every programme has a design intent, and files that fight it lose.
Fix: eligibility triage first. Ten minutes of honest mapping beats ten hours of hopeful writing. Start with the funding overview if you have never mapped your options.
3. Financial plans that do not survive contact
Evaluators read financial plans for a living. Hockey sticks without assumptions, costs that ignore the company's actual accounts, and budgets reverse-engineered from the maximum aid amount all signal the same thing: this plan was written for the application, not for the business.
Fix: one financial plan, used for real management, adapted in presentation only. If the numbers embarrass you internally, they are not ready externally.
4. Vague project scope
"Digital transformation of our processes" funds nothing. Programmes fund defined projects: what will be built or studied, by whom, in what period, at what cost, with what outcome. Fuzziness reads as risk.
Fix: write the project as if briefing a contractor: deliverables, milestones, budget lines.
5. Ignoring the formal checklist
Missing annexes, unsigned declarations, absent certificates. It sounds too basic to matter and it decides real files, because incomplete applications stall, miss deadlines, or exhaust evaluator goodwill.
Fix: treat the checklist as the application. Someone whose job is completeness should own it.
6. Silence on the company's own health
Aid goes to companies expected to survive it. Files that dodge questions about existing debts, thin capitalisation or pending issues do not hide them; registries exist. Unexplained weaknesses read worse than explained ones.
Fix: address the obvious question before the evaluator asks it, with context and a plan.
7. Treating the grant as the business model
Evaluators fund engines, not fuel tanks. When the entire plan collapses without the subsidy, the subsidy stops making sense. The strongest files show a business that happens with or without aid, faster with it.
Fix: frame aid as acceleration, and mean it.
What all seven have in common
Every pattern above is preventable before submission and unfixable after. That asymmetry is the entire argument for preparing files properly: the difference between funded and rejected is usually decided in the two weeks before the deadline, not in the committee room.
We build and pressure-test funding files end to end, from instrument choice to submission. Check what you qualify for, and if a rejection already happened, bring it: post-mortems are where the second application gets strong.