The Money Is In The Answer

Zentrales Innovationsprogramm Mittelstand is usually described as a grant for small and mid-sized German companies doing research. That description is accurate and useless. The part that matters is the sub-programme underneath it, the Durchfuehrbarkeitsstudie, called a Vorprojekt in the paperwork, and what it buys is not a product. It buys an answer. The money is small and the arithmetic is odd. Of the roughly EUR 650 million that funds the whole of ZIM, the feasibility slice is about EUR 10 to 15 million. Three hundred to four hundred applications arrive against that slice each year. Between 150 and 200 of them are funded. That is an approval rate near 50 per cent, which is high for a federal innovation programme, and the average award sits between EUR 42,000 and EUR 50,000. In 2024 the published figures were 187 Vorprojekte funded against a total ZIM budget of EUR 620 million, an average Zuschuss of EUR 43,500, a median project duration of 6.5 months and a median 6.2 weeks between submission and the decision letter. Twenty-five per cent of awards landed on the ceiling. Sixty-seven per cent landed between EUR 30,000 and EUR 50,000. The programme is administered by AiF Projekt GmbH in Berlin for the Bundesministerium fuer Wirtschaft und Klimaschutz, under the ZIM 2025 Richtlinie published in the Bundesanzeiger on 26 June 2025, and it sits under Article 25 of the General Block Exemption Regulation. That last detail matters more than it reads. This is not de-minimis money, so it does not consume the EUR 300,000 three-year aid ceiling that a company group has to keep clear. Four to eight weeks is the decision window. Nothing in it is a mystery. All of it is published.

Most applicants size a budget by asking what the work costs. The programme sizes it by asking what the ceiling does to the proportion. The hard numbers are these. A Vorprojekt carries a Zuschuss ceiling of EUR 55,000 that does not move. Maximum eligible costs after the 2025 Richtlinie update are EUR 100,000. The base funding rate is 50 per cent for a Klein company, 45 per cent for a Mittel company and 40 per cent for a consolidated group that still fits the KMU test. On top of that sits a bonus stack: five percentage points if the operating site is in the eastern Laender or part of Berlin, five more if the company is less than ten years old at first application. The combined maximum rate is 60 per cent. Now the trap. At 50 per cent, the EUR 55,000 ceiling binds at EUR 110,000 of eligible costs. At 55 per cent it binds at EUR 100,000. At 60 per cent it binds at EUR 91,667. Budget below that number and you leave programme money unclaimed. Budget above it and you are spending your own cash on work the programme was never going to reimburse. Personnel is the largest line and it is not a negotiation. Annual productive hours are standardised at 1,720, and the reference maximum rates are published by role: project lead up to EUR 100 an hour, senior engineer up to EUR 110, mid-level engineer up to EUR 90, junior up to EUR 70. Overheads run as a flat rate of 15 to 25 per cent of personnel cost, with an absolute cap at 100 per cent of personnel cost. Subcontracts are allowed but only at fixed price, and only up to 35 per cent of total eligible costs. The work of sizing a Vorprojekt is not estimating. It is solving for one number.

One sentence in the Richtlinie costs more applicants their money than any other. The Antragsdatum is the binding stop-the-clock, and any cost incurred before submission is non-eligible. Not partially eligible. Not eligible on appeal. That means the honest reading of the programme is backwards from how most people approach it. You do not build something, then find the funding. You submit, and the clock starts. Every euro of thinking, prototyping and supplier conversation that happens before you press submit is a euro the programme will never see. The applicant tests are narrow and specific. Fewer than 250 employees, and either under EUR 50 million of annual turnover or under EUR 43 million of balance-sheet total, per Recommendation 2003/361/EC. A registered Betriebsstaette in Germany with real operational activity, not a mailbox address. An HRB-registered legal entity. Equity above zero in the latest balance sheet, no insolvency or restructuring proceedings. Independence as a KMU, which means a group subsidiary only qualifies if the consolidated group still passes the thresholds. Then the throughput limit: a single KMU may hold at most four open ZIM applications at one time, counting drafts, submitted files and approved projects that are running. And the plumbing. Since June 2025 the AiF Online-Portal at aif-portal.de is mandatory. There is no paper route and no informal email route. The portal registration has to be made under the applicant's own HRB, with tax identification and signatory details. Free pre-application consultation with the Projekttraeger exists and is worth the afternoon, because the person on that call is the person who will read the file.

The application is a document exercise, and the document list is published in full. Antragsformular B online. A Vorhabensbeschreibung of no more than 15 pages. A cost breakdown on Annex K. A KMU-Erklaerung under the EU definition. A de-minimis declaration if that route is chosen over Article 25. An HRB extract no older than six months. A tax clearance certificate from the Finanzamt. A Bonitaetsauskunft no older than twelve months. Proof that the applicant's own share is funded. Letters of intent from every named subcontractor and research partner. CVs of key personnel. A data-protection consent. And proof of signatory authority if the managing director is not the sole signatory. AiF rejects letters of intent that are described as to be signed. Signed or absent is the choice. The Vorhabensbeschreibung follows a fixed structure and the page count is enforced. A half page of summary. Two to three pages on the state of the art and the degree of innovation, with a national and international comparison and at least five citations. One to two pages of objectives with explicit key performance indicators and go/no-go criteria. Four to five pages of methods and technical concept, which is the weight of the file. Two pages of work packages and a milestone plan. One to two pages of personnel and cost planning that ties back to Annex K. Two pages of the exploitation plan. One page of risks and countermeasures. Half a page to a page on cooperation and subcontracts. And a quarter page declaring the project start, which is the Antragsdatum again, in writing. Every work package in section five needs a corresponding personnel-cost line in Annex K. Reviewers read those two sections side by side, and a package without a cost line reads as padding.

The programme is not buying the technology. It is buying the answer, and one of the answers it accepts is no.

The review runs on a clock. Week one is a formal eligibility check by AiF, which is administrative and rarely fails a complete file. Weeks two to four, two independent technical experts are assigned to the same application. Their reports are consolidated by the case officer. Weeks five to seven, a decision committee meets. Then the Bescheid is issued, and it is one of three things: a Bewilligung, a partial approval, or a refusal. If it is an approval, the project starts on the date written in the letter, usually one to two weeks after it arrives. The money does not come first. AiF does not advance funds. It reimburses against documented spend, with interim claims permitted every three months and the final payment released only after the audit of the Verwendungsnachweis, which lands three to six months after the project ends. So the cash flow is the applicant's problem before it is the programme's. Personnel is claimed at capped hourly rates, project lead up to EUR 100, senior engineer up to EUR 110, mid-level up to EUR 90, junior up to EUR 70, and every one of those hours needs a monthly time sheet showing hours per work package per person. Overheads are a flat rate of 15 to 25 per cent of personnel cost. Subcontractors are paid in full by the applicant and reimbursed at the applicant's funding rate. Mid-flight, budget can be moved between cost categories by an Aenderungsantrag, up to 20 per cent. At the end, the Schlussbericht of no more than 30 pages is due within three months, and the final accounting with it.

Here is the part that the programme states openly and applicants still miss. A Vorprojekt exists to reduce a scientific-technical risk before a larger follow-on project is applied for. The go/no-go criteria are supposed to be written into the objectives section, and they are supposed to be checkable. The programme's own worked example is a study on phosphorus recovery with a criterion that reads: if the Pareto front shows less than 80 per cent recovery, the follow-on cooperation project is not pursued. Read that again. The criterion for success is a number, and the consequence of missing it is not doing the next thing. The probability that the answer comes back no is not zero. In a programme that funds roughly half of the applications it receives and caps the study at six months, a negative answer is a legitimate outcome and a completed deliverable. The refusals at this stage come from a much narrower list: a project too close to the state of the art, an exploitation plan that names no follow-on programme, amount and date, a budget not tied to deliverables, a subcontract share above 35 per cent, costs incurred before the Antragsdatum, an unclear KMU classification, equity below zero in the latest balance sheet, missing letters of intent, no explicit go/no-go criterion, and a cluster of near-identical applications from sister companies in the same group. Every one of those is a document defect, not a technology defect. Which is the whole point. The programme is not buying the technology. It is buying the answer, and one of the answers it accepts is no. A no is a result. It belongs in the report.