Annual Tax Act 2026: What the Draft Could Mean for the Forschungszulage
The German cabinet has approved the draft Annual Tax Act 2026. For the Forschungszulage it contains a higher funding cap per project and more legal certainty on deadlines.
There are essentially three ways to bill Forschungszulage consulting:
So the pricing question is really a risk question. The Forschungszulage runs in two stages: first the certification body (Bescheinigungsstelle Forschungszulage, BSFZ) reviews the technical eligibility of your project, then the tax office assesses the allowance as part of your tax return. Between signing the contract and receiving the money there are two authorities that can say no, or say "less". Which fee model you choose decides who carries that risk.
| Model | Payment if rejected | Consultant's incentive | Your risk |
|---|---|---|---|
| Hourly fees | Yes | Effort | High |
| Fixed price per application | Yes | Keep effort low | High |
| Success-based | No | Maximize funding | Low |
Hourly billing is the classic model in funding consulting. It is transparent in the sense that you can see what time is spent on. And it has a legitimate use case: a tightly scoped second opinion, a review of an application you wrote yourself, a single question about how to draw the line between routine work and R&D.
It becomes a problem as soon as the consultant owns the whole application. Three effects then reinforce each other:
The last point is the most expensive one, because it never shows up on an invoice. It shows up in the decision.
A fixed price feels predictable, and it is. It just makes the wrong thing predictable. With the Forschungszulage, the difference between a good and a mediocre application is rarely whether it gets through: it is how much funding it unlocks. Are all eligible projects actually identified, including those from earlier years? Are employees cleanly assigned by percentage? Are contractors and external development partners claimed correctly?
That work takes time, and a fixed price does not reward it. Quite the opposite: the faster the provider is done, the more profitable the engagement. The incentive points at the smallest possible effort, not at your largest possible result. An application covering one project is cheaper to produce than one covering four fiscal years and six projects, and it earns the provider the same money.
How strongly the assessment base drives the funding amount is something you can work through in the guide to calculating the Forschungszulage. That is where the leverage sits, not in the form.
"Success-based" sounds unambiguous. It isn't. "Success" can refer to three very different amounts:
The same percentage means a completely different amount depending on the reference. If you compare two success-based offers without knowing the reference amount, you are comparing nothing. Ask for it in writing, in one sentence, in the contract.
The second thing to watch: what counts as success when only part of the application goes through? With several projects across several years, partial certification is the norm, not the exception. A clean contract states that the commission applies to the part actually funded, not to what was originally claimed.
Which types of provider exist on the market, and which compensation model typically goes with which type, is broken down in our overview of Forschungszulage provider types.
The fee is only one part of the bill. The other is internal hours, and they appear in no proposal.
The BSFZ reviewers are engineers and computer scientists. An application convinces them only if it is technically concrete: state of the art, specific technical target values, comprehensible uncertainties. That knowledge sits with exactly the people in your company who have the least time. If a consultant cannot follow you technically, the work shifts back to your CTO: first as explanation loops, then as correction loops on the draft.
Two questions make this block visible before you sign:
With us, your team's effort is under five hours, spread across four calls. We run the interviews, draft the technical descriptions as proposals and file the application; you review.
This is the clause that moves the most money and gets read the least.
The BSFZ certificate is neither a payout nor a final decision on your funding amount. It establishes that your project is eligible. Only after that does your tax advisor claim the allowance in the tax return, and only the tax office decides on the costs. Reductions can happen at that point, for instance if cost assignments or supporting evidence do not hold up.
So if a fee falls due with the certificate, it can be based on a sum you never actually receive in that amount. Conversely, a provider who ties the due date to the payout has a direct interest in a cost breakdown that holds up at the tax office, not just at the BSFZ.
The same logic applies to rejections. In the BSFZ procedure a rejection is not the end: you can improve and resubmit a revised project, which makes it all the more important to settle this question in the contract. What to do in that situation is covered in Forschungszulage rejected: how to respond. Clarify in advance whether revisions and a resubmission are included in the price or count as a new engagement.
The fifth question is worth asking because the baseline is known: industry-wide, the BSFZ rejects around 25% of projects. Across the applications we support, the success rate is over 92%. That difference does not come from tricks but from the technical drafting, written the way the BSFZ reviewers assess it.
We work 100% success-based. In concrete terms:
That has a consequence which matters more to you than any percentage: we only earn on approved applications. So we will tell you honestly if an application is not worth it for you, rather than taking on an engagement that predictably will not hold. What is covered in detail is on the page about our Forschungszulage consulting.
Which leaves the number you actually came for. We will tell you directly, with no proposal process and no documents required. Book a first call: 15 minutes, and within the first five of them you will know how we bill and whether an application is worth it for you at all.
It depends entirely on the fee model. There are three on the market: hourly fees, a fixed price per application, and a success-based commission. With hourly and fixed-price models you pay regardless of the outcome; with a success-based model you only pay if funding actually flows. We work 100% success-based: a commission on the funding actually paid out, and no payment if the application is rejected. You get the concrete number in a first call, with no proposal process required.
With hourly fees and fixed prices, yes: the fee is tied to effort, not to the outcome. Not with us. If the application is rejected you pay nothing, not even for work already done.
This is the single most important question in the contract, because there are three very different reference amounts: the assessment base (your eligible expenses), the amount certified in the BSFZ procedure, and the Forschungszulage actually paid out by the tax office. The assessment base is a multiple of the payout, so an identical-sounding percentage means a completely different amount depending on the reference. Always get it in writing. Our commission refers to the funding actually paid out.
That is negotiable and belongs in the contract. What matters is this: the BSFZ certificate is only the first stage. The tax office then assesses the Forschungszulage as part of your tax return, and it can reduce costs at that point. A fee that falls due with the certificate can therefore exceed what actually reaches you. With us, the commission only falls due once the money is in your account.
Your team's time, above all your developers'. If you write the technical project descriptions yourself, or supply them through many rounds of revisions, you tie up exactly the people who should be working on the product. With us, your team's total effort is under five hours across four calls; we do the writing and the filing.
On paper a commission can be larger than a flat fee. What matters, though, is what you keep net. A fixed-price provider has no economic reason to look for further eligible projects or cost positions; a success-based provider earns precisely by doing so. So don't compare fee against fee, compare expected funding minus fee.
No. In the BSFZ procedure you can improve and resubmit a revised project; a rejection always refers to the application in the form it was submitted. Only the tax office's decisions are final. For your fee model that means: clarify up front whether revisions and a resubmission are included in the agreed price or billed separately.
Drop us a line about what's on your mind — we'll get back to you personally and without obligation.
The German cabinet has approved the draft Annual Tax Act 2026. For the Forschungszulage it contains a higher funding cap per project and more legal certainty on deadlines.
As of 7 July 2026, the BMWE is no longer accepting new ZIM applications. What the freeze means, who is affected and why the Forschungszulage is now the better baseline funding.
Scaneca secured 280,000€ in Forschungszulage for developing their mobile 3D body scanners. Why the team hesitated for years — and what changed.
What is changing about the Forschungszulage in 2024 and 2026? KMU funding rate 35%, overhead flat rate 20%, higher assessment basis. Wachstumschancengesetz explained.