A previous video worked through the definition of deep tech as it appears in the EIC Work Programme, taking the official wording apart to see what it actually commits to. This one is more practical. The question here is not what deep tech means on paper but how evaluators decide, in practice, whether the company in front of them is actually deep tech or not.
That distinction matters because the practice is genuinely confusing, and the confusion runs all the way to the top of the system. The Scaleup Europe Fund, which is funded through the European Innovation Council and therefore carries government money at least in part, recently invested in Lovable, a vibe coding platform. So the obvious question: is Lovable deep tech? It is safe to say it is not. It is an excellent business with great execution and a great company behind it, but it does not strike anyone as deep tech. When the institution writing the deep tech rules funds something that is not deep tech, applicants can be forgiven for wondering what the standard actually is.
Deep Tech Smuggling
The same pattern appears further down. There are a lot of companies funded through the EIC Accelerator that simply do not read as deep tech. They sound like simple tools, simple software, or even simple hardware. You look at them and the question forms on its own: is this really a deep tech technology? Frequently what is underneath is some off-the-shelf hardware plus some software that connects a few things, and the whole assembly is deemed deep tech largely because the EIC Accelerator funded it.
This is a question that comes up constantly when reviewing projects: why is this deep tech? Out of that recurring question comes a useful term, which is deep tech smuggling. It describes the practice of packaging something as deep tech that is not really deep tech, and getting it through.
Two caveats need stating immediately, because they pull in opposite directions.
First, you cannot simply trick your way through the EIC Accelerator. It is a very difficult programme with a great many evaluation steps. If you inflate your project heavily in order to advance through the stages, at some point somebody notices. At the end, someone is going to say, wait a second, what you are doing is a very simple software tool. You do not have anything special, you do not really have a moat. That becomes very obvious, at least to the jury members at the final stage.
Second, and against that, there are cases that made it through every step and ended up funded where significant facts were never surfaced. Things the jury simply did not know about. Perhaps nobody asked the question. Perhaps nobody bothered to ask basic things about current funding rounds or similar. There are a lot of points in this process where evaluators and jury members simply make mistakes.
None of this is a suggestion to exploit anything. The useful question is different and entirely legitimate: how do you present yourself as deep tech when you genuinely are?
The First Horseman: Patents, If You Have Hardware
If your company has hardware, a good way to present yourself as deep tech is to have patents. If you have any type of hardware, you should have patents covering it.
Patents are not a magic bullet, and it is worth being honest about their limits. It is always possible for an evaluator to take a patent apart, or to take apart how innovative the patent actually is. A granted patent is not a verdict on novelty in the eyes of a jury. But if you have hardware and no patents at all, you have removed the most straightforward instrument you had for supporting the claim that you are deep tech. That is very important.
The Second Horseman: A Body of Work, If You Have Software
If your company is software, you need to be able to describe a body of work. Whatever software you are building, you still have to identify what the moat is.
If the software itself is very simple, then the innovation has to lie somewhere else. And it is worth being realistic about where it does not lie. The innovation cannot be that you have a very special back end, because the back end is almost certainly using standard server and database logic like everyone else. The architecture is probably not going to be unique. Evaluators know this.
So what is genuinely new about you? It might be a different type of language you are using. It might be a particular knowledge graph or a data repository. Whatever form it takes, there has to be a body of work: something you can point at, that of course does not have to be physical, but that you can digitally touch and say, that thing, nobody has it, and nobody even knows how to create it. That is mine.
So the pairing so far is simple. Hardware means there should be patents. Software means there should be a body of work, something digitally tangible you can hold on to. Those two things are the first place to look when trying to establish the deep tech aspect of a company.
The Third Horseman: The Science Angle
The third element is science, and it applies to software and hardware alike.
The strongest version is that you are a university spin-off, or that there is genuine science around what you do, and ideally that you participated in it. That participation can come from either direction. Either you were the object of study, meaning university researchers studied the thing you are building, or you were on the inside of the research, perhaps as the scientist who invented it.
Where there is science involved through a university, this is a big plus for deep tech. Note that this sits on top of the previous two rather than replacing them: on top of the patents for the hardware, and on top of the body of work for the software.
That gives three things, the first three horsemen of deep tech.
The Fourth Horseman: The Industry You Operate In
There is a fourth element, and it is the one applicants tend to underestimate because it has nothing to do with the technology itself. It is the industry you operate in.
If you are in a very technical manufacturing, industrial, chemical or medtech industry, it almost does not matter exactly what you are doing. It will always seem as though there is science involved, or that there is deep tech present, simply because you are in a technical industry. Perception attaches to the sector.
The clearest way to see this is to hold the technology constant and change only the application. If you have a technology and you apply it to shoes, it will not read as deep tech, or as technical, to nearly the same degree as if you applied the exact same technology to pharmaceuticals. Pharmaceuticals as an industry simply appears far more deep tech. The underlying invention did not change at all. Only the context did.
The Four Horsemen Together
That completes the set. The four horsemen of deep tech are patents for hardware, a body of work for software, the science angle, which can come from a number of different directions, and finally the industry itself.
The last one reduces to two questions worth asking honestly about your own company before an evaluator asks them for you. Are you operating in an industry that seems technical? And are you operating in an industry that seems aligned to science and engineering?
