IP and Schutzrechte in the application: what to claim, what to file, and the university question

Founders often read the IP question as "do you have a patent," and answer it as a yes/no. That is not the question. Reviewers are asking whether the venture can build a defensible position around its core asset and whether the team has the legal right to use what it is building on. A patent is one way to be defensible, but it is not the only one, and for many software ventures it is not even the best one. What reviewers want to see is that you have thought clearly about what protects you from a well-funded competitor copying you - whether that is a patent, proprietary data, a network effect, deep domain integration, speed, or some combination - and that you actually have the rights to your foundational technology.

So the IP section has two jobs. First, articulate the moat honestly: what is your defensibility, why is it durable, and what is your plan to strengthen it during the funded year. Second, clear the rights question: confirm that nothing in your foundation is owned by someone who could pull it out from under you - a former employer, a co-founder who left, or, for spin-offs, the university. The second job is the one that quietly sinks applications, because a brilliant moat you do not have the right to use is worth nothing.

There is a persistent myth in grant applications that you need a patent to look serious. You do not. Patents make sense for genuine technical inventions - a novel material, a hardware mechanism, a specific method that is both inventive and worth the cost and disclosure of patenting. They make much less sense for most software, where the invention is often the product and execution rather than a single patentable method, and where filing a patent forces you to publicly disclose how it works in exchange for protection that is expensive to enforce against a fast-moving competitor. Claiming you will patent software that is not really patentable reads as naive to a reviewer who understands IP.

Be precise about your real moat. If it is proprietary training data nobody else can assemble, say that and explain why it is hard to replicate. If it is a deep integration into a regulated workflow that takes years to build trust into, say that. If you genuinely have a patentable invention, say what it is and where you are in the filing process. The reviewer is testing your judgement about defensibility as much as the defensibility itself - a founder who knows their moat is data and not patents, and can explain why, reads as more sophisticated than one who reflexively promises three patents.

For any venture built on university research - and especially for EXIST Forschungstransfer - the rights question centres on the university. Research results produced at a university, including by employed researchers, are frequently owned or co-owned by the institution rather than the individuals, and inventions made by employees can fall under specific rules on employee inventions. This means the team's right to commercialise the core technology usually has to be formally arranged with the university, through a licence, an assignment, or a negotiated agreement. Reviewers know this, and an application that is vague about who owns the foundational IP and whether the team has the right to use it leaves the biggest possible question mark over the whole venture.

Engage the university's transfer office early. Most German universities have a Technologietransfer or Gründungsbüro function that handles exactly these arrangements, and they have done it many times. Get the conversation started before you submit so you can write, honestly, that you are in active discussions about an assignment or licence on terms you expect to be workable, or that an arrangement is already in place. What you cannot do is leave it unaddressed and hope it sorts itself out after the grant - the rights have to end up clearly with the venture, and reviewers want to see you are on a credible path to that. This is genuinely a legal matter; treat anything here as orientation and get proper legal advice on the actual agreement.

There is a timing trap that catches research founders specifically. Patent protection generally requires that the invention has not been publicly disclosed before filing - and a public disclosure can include a conference talk, a published paper, a thesis, or even a sufficiently detailed public pitch. If your core invention is genuinely patentable and you intend to patent it, you have to manage the sequence: file (at least a priority application) before you disclose, or you may forfeit patentability. This collides directly with the academic incentive to publish, and many a patentable result has been published into the public domain before anyone thought about protecting it.

The grant application is usually confidential to the Projektträger and reviewers, so submitting it does not itself count as a destructive public disclosure - but do not assume, confirm the confidentiality terms, and never post your detailed invention publicly while a patent is still an open option. The practical rule for any patent-relevant venture: talk to a patent attorney about the filing-before-disclosure sequence before you publish anything or give any public talk about the invention. The Coaching budget can fund that first conversation. None of this is legal advice; patentability and disclosure rules are technical and jurisdiction-specific, and the cost of getting the sequence wrong is losing the protection entirely.