Business Finland
Business Finland R&D application budget: how to structure it
How to structure a Business Finland R&D budget that survives assessment: work packages, cost allocation, subcontracting share, and realistic staff loading.
Evaluators read the budget as a lie detector for the rest of the application. A tight, bottom-up budget signals that you have actually thought through the work. A round-number, top-down budget signals that you picked a funding target and reverse-engineered a project to justify it. The numbers themselves matter less than what their structure reveals about your thinking.
Here is how to build one that holds up.
Start from work packages, not a total
The single most important move: do not start with a total and divide it. Start with the work.
Break the project into a handful of work packages — coherent chunks of R&D, each with an objective, the activities inside it, and a clear deliverable or milestone. For an AI project, packages often follow the arc of the uncertainty: data and baseline, method development, evaluation and hardening, integration and piloting. Then cost each package from the bottom up. The total is an output of this process, not an input.
This matters because assessment tests whether costs map to work. When every euro traces to a work package that traces to a milestone, the budget is defensible line by line.
Staff the packages realistically
Personnel is the largest line for most AI R&D, and it is where credibility is won or lost.
- Load real people at real allocations. Name roles, estimate the hours each will genuinely put into each package, and make sure the totals are humanly possible. A plan that quietly assumes three people working at 150% capacity falls apart on inspection.
- Match seniority to the work. R&D that requires senior researchers should be staffed and costed as such, not padded with junior hours to lower the average.
- Tie hours to milestones. If a work package claims a large personnel budget but produces a thin deliverable, the mismatch is obvious.
Remember that personnel is claimed on actual booked hours at reporting time, so the plan should reflect how the team will really work — because that is what you’ll have to substantiate later.
Get the subcontracting share right
Subcontracting is useful but politically loaded. Two rules:
- Keep it a minority of the project. The R&D should be substantially your company’s own work. When subcontracting dominates, the project reads as procurement, and it may breach caps.
- Justify each external engagement. Say what specialist capability you’re buying and why it can’t reasonably be done in-house. Unexplained subcontracting invites questions; related-party subcontracting invites scrutiny.
Handle overhead correctly
Indirect costs come in as a percentage allowance on top of personnel, not as itemised lines. Don’t separately budget for the office, licenses, and general expenses the overhead already covers — that’s double-counting, and it gets corrected downward. Apply the allowance and move on.
Match the budget to the timeline
Spread costs across the project period in a way that matches the plan. A budget that front-loads all spending into month one, or bunches it implausibly, contradicts the milestone structure. Phasing the budget in line with the work packages reinforces that the plan is real.
The credibility checklist
Before you submit, the budget should pass all of these:
- Every cost traces to a named work package.
- Every work package produces a milestone or deliverable.
- Personnel hours are humanly possible and matched to seniority.
- Subcontracting is a justified minority of the total.
- Overhead is applied as an allowance, not itemised twice.
- The totals are specific, not suspiciously round.
- Spending is phased to match the timeline.
The mindset
A budget is not a request for the most money you can plausibly ask for. It is a model of the work you are actually going to do. Build it as an honest model and it does three jobs at once: it survives assessment, it protects you at reporting time, and it makes the whole application read as the work of a team that knows exactly what it is taking on.
Related: Eligible costs in a Business Finland R&D application · How much Business Finland R&D funding can an AI project receive? · Business Finland R&D project reporting requirements